Description
Reports Study on Understanding State Goal Orientation: Leadership and Work-group Climate as Key Antecedents :- Goal orientation is a "disposition toward developing or demonstrating ability in achievement situations". Previous research has examined goal orientation as a motivation variable useful for recruitment, climate and culture, performance appraisal, and selection.
REPORTS STUDY ON UNDERSTANDING STATE GOAL ORIENTATION: LEADERSHIP AND WORKGROUP CLIMATE AS KEY ANTECEDENTS
ABSTRACT This research attends to a broad range of practically significant employee achievement goals and provides insight into how to enhance individual-level performance by examining the antecedents to individual-level state goal orientation in organizational work groups. State goal orientation is defined here as a temporary achievement goal, and it is theorized that leadership and work group climate processes parallel each dimension of state goal orientation to cue and ultimately induce the corresponding achievement goal among individual work group members. The leader's achievement priority is argued to drive the formation of work group climate consistent with this priority. The resulting work group climate signals and compels group members to adopt the ascribed form of state goal orientation. The quality of the leader-member exchange (LMX) relationship is viewed as a means to internalize cues from the work group climate in the emergence of state goal orientation. Results from experimental and field studies provide evidence that (1) leadership and climate perceptions are related to their parallel form of state goal orientation, (2) the relation between individual perceptions of a climate for learning and state learning goal orientation is stronger when group members enjoy higher quality
exchange relationships with their leader, and (3) state goal orientation may be validly and reliably assessed using the measure created especially for this research.
TABLE OF CONTENTS List of Tables????????????????????????????... vii List of Figures????????????????????????????.. viii Chapter1: PROBLEM STATEMENT?????????????..??????.1 Chapter 2: MODEL DEVELOPMENT AND HYPOTHESES???...??????10 Model Overview????????????????????????.. 10 Key Construct Definitions????????????????????... 13 Leader Priority??????????????????????. 13 Work-group Climate...???????????????????. 22 Goal Orientation?????????????????????.. 32 Individual Outcomes????????????????????37 Leadermember exchange relationship????????????? 39 Hypothesized Relationships????????????????????.41 Relation between leadership and work-group climate???????.42 Relation between work-group climate and state goal orientation???45 Relation between state goal orientation and outcomes??????? 51 Moderating role of the quality of LMX relationship???????... 57 Chapter 3: RESEARCH METHODS???????????..???????... 62 Research Design: Phase I—Qualitative Stage???????????...... 62 Sample and design????????????????????... 62 Key findings??????????????????????? 63 Research Design: Phase II—Measurement Study??????????? 65 Sample?????????????????????????. 65 Design?????????????????????????. 65 Data collection procedures?????????????????.. 66 Measures????????????????????????.. 67 Control and theoretically-relevant measures???????.. 67 Dependent measures????????????????? 68 State goal orientation????????????????... 68 Development of measure????????????. 68 Method for gathering validity evidence??????.. 69 Convergent and discriminant validity evidence???.. 72 Summary of key findings???????????? 82 Research Design: Phase III—Field Study?????????????? 83 Sample?????????????????????????. 83 Design and data collection procedures????????????? 84 Measures???????????????????????..... 85 Analytic procedures???????????????????? 91
v
Issues of control?????????????????????.. 92 Chapter 4: RESULTS FROM THE FIELD STUDY?????????????. 94 Psychometric evidence of the state goal orientation measure??????.. 94 Findings from hypothesis testing?????????????????. 96 Findings from tests for mediation?????????????????. 106 Summary??????????????????????????... 108 Chapter 5: DISCUSSION???????????????????????... 110 Study limitations???????????????????????.. 120 Research implications?????????????????????.. 122 Managerial implications????????????????????... 124 Conclusion?????????????????????????? 125 APPENDICES????????????????????????????. 126 A: Interview Protocol?????????????????????...126 B: Focus Group Protocol????????????????????. 129 C: Experimental Conditions??????????????????? 131 D: Time 1 Measurement Study Items???????????????.. 134 E: Time 2 Measurement Study Items???????????????.. 136 F: Field Sample Descriptive Statistics by Branch??????????...140 G: Time 1 Field Study Items???????????????????149 H: Time 2 Field Study Items (collected from employees)???????...151 I: Time 2 Field Study Items (collected from managers)???????? 155 REFERENCES?????????????????????????..??.. 160
vi
List of Tables Theoretical Development 1. Behavioral Indications of Leader Priority???????????????.21 2. Description of Three Prototypical Work-group Climates??????..??.. 32 3. Construct Definitions???????????????????.???... 41 Measurement Study 4. Final Results from Confirmatory Factor Analysis: Factor Loading?????. 72 5. Multi-trait, Multi-Method Results: Model Fit Statistics?????????.. 75 6. Variance Components due to Trait, Method, and Error for Model 1????... 75 7. Measurement Study Variable Means, SDs, and Intercorrelations?????? 79 8. Mean Differences in State Goal Orientation Across Experimental Conditions... 81 9. Hierarchical Regression Results: State Goal Orientation Predicting Outcomes. 82 Field Study 10. Analytic Strategy Used for Testing Hypotheses????????????.92 11. Final Results from Confirmatory Factor Analysis: Factor Loadings????. 95 12. Branch-level Main Study Variable Means, SDs, and Intercorrelations???..96 13. Individual-level Main Study Variable Means, SDs, and Intercorrelations??.97 14. Random Coefficient Modeling Results: Leader Priority Predicting Psychological Climate???????????. 100 15. Random Coefficient Modeling Results: Psychological Climate Predicting State Goal Orientation????????. 102 16. Random Coefficient Modeling Results: State Goal Orientation Predicting Outcomes?????????????..103 17. Random Coefficient Modeling Results: Interactive Effect of Psychological Climate and LMX in Predicting State Goal Orientation??????????????????????.105
vii
List of Figures 1. Proposed Model of Contextual Influences on Individual State Goal Orientation and Associated Outcomes???????????12 2. Data Sources and sequencing???????????????????85 3. Interactive Effect of Psychological Learning Climate and LMX on State Goal Orientation?????????????????????...106
viii
CHAPTER 1 Problem Statement To remain competitive in an environment characterized by continuous change, high levels of complexity, and tremendous opportunity, contemporary organizations need to remain adaptive, innovative, and offer exceptional high products and services to customers. To achieve this end, organizations must motivate their members to
continuously learn and adapt to changing job demands and discover novel and innovative methods for solving complex problems (Hall & Mirvis, 1996). At the same time,
however, organizational members need to maintain high levels of performance to maintain high levels of satisfaction among customers, their colleagues, and organizational leaders. Research has shown leaders are instrumental in encouraging the motivation to learn and perform at high levels (e.g., Locke & Latham, 1990; VanVelsor, McCauley, & Moxley, 1998). In addition to these more positive outcomes, leadership can deter
organizational members from owning up to mistakes, thereby sabotaging unique learning opportunities and promoting "face saving" types of behavior (e.g., Edmondson, 1996). Building on this work, this research explores how organizational leaders encourage unique motivational responses for learning, performance, and avoiding failure among their employees and the outcomes associated with these different motivational states. Academic research on goal orientation provides a useful means to understanding the influence of contextual factors, such as leadership, on individual motivation and outcomes. Goal orientation is defined as one's goal preference in achievement settings (Dweck, 1986) and refers to one's desire to develop, realize, or exhibit capability to perform a specific activity (Dweck & Leggett, 1988). Initial research identified two
1
broad classes of underlying goals that individuals pursue: mastery and performance (Dweck, 1986; Dweck & Leggett, 1988). Individuals displaying a mastery or learning orientation focus on building their competence and/or improving their abilities. Those favoring a performance orientation seek to demonstrate their competence by seeking favorable judgments or avoiding negative judgments regarding their capabilities. While often treated as a dispositional trait, theory and research suggests that situational demands may induce an orientation contrary to one's preference (e.g., Button, Mathieu, & Zajac 1996; Elliott & Dweck, 1988; Nicholls, 1984). Substantial empirical evidence suggests that even subtle differences in experimental instructions in research settings can encourage individuals to adopt an orientation consistent with situational cues (e.g., Butler, 1993; Elliot & Harackiewicz, 1996; Harackiewicz & Elliot, 1993; Kozlowski, Gully, Brown, Salas, Smith, & Nason, 2001; Mangos & Steele-Johnson, 2001; SteeleJohnson, Beauregard, Hoover, & Schmidt, 2001; Stevens & Gist, 1997). Goal orientation has been shown to be a potent predictor of individual outcomes. For example, researchers have documented its relation to other motivational factors such as self-efficacy and intrinsic motivation (e.g., Chen, Gully, Whiteman & Kilcullen, 2000; Elliot & Church, 1997), behavioral outcomes such as performance (e.g., Brett & VandeWalle, 1999; VandeWalle, Brown, Cron, & Slocum, 1999) and attitudinal and affective responses (e.g., Ames & Archer, 1988; Elliot & McGregor, 1999). In the context of learning and development, goal orientation has been shown to be related to the acquisition of knowledge and transference of skills learned in training to the work context (Brown, 2001; Fisher & Ford, 1998; Ford, Smith, Weissbin, Gully & Salas, 1998;
2
Kozlowski et al, 2001). This body of research demonstrates the profound impact of goal orientation on behavioral, motivational and attitudinal outcomes. The wide-reaching impact of the construct of goal orientation and evidence of its malleability sets the stage to pose an intriguing question: if it is possible to promote different types of goal orientations, can organizational leaders facilitate adoption of different forms of goal orientation among their employees, and in turn, affect their performance? And, if so, how do leaders impact this motivational process and its associated outcomes? This line of inquiry provides practitioners and scholars value by uncovering insights into how leaders can better motivate employees to learn and perform in order to achieve superior results. Pursuit of understanding this phenomenon is
evidenced by the long tradition in the organizational behavior field that dates back to the Hawthorne Studies (as cited in Wren, 1979), in which Elton Mayo posited that supervisor's attention to workers enhanced employee morale, and in turn, affected performance levels. More contemporary scholars have studied this question by exploring leadership impact on employee self-efficacy and goal setting and performance (e.g., Bandura, 1997; Eden, 1992; Locke & Latham, 1990). While organizational behavior scholars generally support the contention that studying leader impact on employee motivation and performance is essential, what value does examining leadership's relationship to one specific type of motivation, i.e., goal orientation, add to our understanding? Much of the existing research has focused on leadership's impact on motivating employees as a way to increase their productivity. Yet, with the emergence of the knowledge economy and the shift in the nature of work to becoming more complex, information- and skill-intensive, and fluid (Howard, 1995), the
3
importance of studying factors affecting employees' motivation to learn becomes vital. The changing organizational landscape and continuously shifting job requirements are redefining "employee effectiveness," demanding that researchers broaden their focus to include a motivation to learn, as well as more task-specific, productivity-related motivational concepts. Moreover, recent research suggests in some work group settings, contextual influences promote a motivation to "save face" among employees, dissuading employees from acknowledging their mistakes, continuously learning, adapting to new job requirements, and performing effectively (e.g., Edmondson, 1996; 1999). These organizational-, job- and research-related trends challenge researchers to examine a broader range of employee motivations. In this way, the concept of goal orientation provides a necessary vehicle to studying this variance in employee motivations. Some researchers have begun to explore how leadership and other contextual factors relate to employee goal orientation and performance (Ames & Archer, 1988; Potosky & Ramakrishna, 2002; Smith-Jentsch, Van Duyne, & Reynolds, 2001). Most directly, Smith-Jentsch et al (2001) investigated team leader goals as an inducement of team members' adoption of specific forms of goal orientation. Other studies have
explored broader contextual influences. For example, Ames & Archer (1988) explored how student perceptions of their classroom experiences affected their adoption of specific goal orientations. Potosky & Ramakrishna (2002) found that the work environment's emphasis on updating skills moderated the relationship between learning goal orientation and job performance. While these studies begin to document how contextual influences may impact employees' goal orientation, our understanding is limited by the scarcity of research in
4
this area, as well as the conceptual and methodological shortcomings of existing research. Conceptually, researchers have taken a myopic view of either leadership and/or context. Smith-Jentsch et al (2001) incorporate leader goals as the only contextual influence into their research. While leader goals have been shown to be important (Bass, 1985; Bennis & Nanus, 1985; Kotter, 1982), they represent one leadership process and confine our understanding only to the impact of leader goals rather than the larger influence of leadership on employee goal orientation. Methodologically, these studies have struggled to adequately measure goal orientations as induced by situational influences. Ames and Archer (1988) assess goal orientation by measuring classroom environment, thus confounding context with student motivation. The measure used by Smith-Jenstch et al (2001) includes thoughts and general assessments of the situation rather than a drive toward learning and/or performance. Lastly, Potosky and Ramakrishna's (2002) use of a data-analytic technique that does not properly accounts for the multi-level nature (i.e., non-independent) of their data raises questions regarding the validity of their results. The result of the paucity of research in this area and methodological and conceptual shortcomings of existing studies is that very little is known about how contextual influences, such as leadership, can facilitate the adoption of different forms of goal orientation among employees. In particular, this limited conceptualization of leadership influences is indicative of a larger shortcoming of the leadership literature. Graen and Uhl-Bien (1995) argue that most leadership research has employed a stove-pipe approach, choosing to focus on only one aspect of leadership process (e.g., leader, follower, or leader-follower relationship). More specifically, Graen and Uhl-Bien (1995) assert that incorporation of
5
variables representing leader-focused, follower-focused, and leader-follower relationshipbased perspectives will ensure against overlooking critical leadership aspects, thereby increasing predictive power of leadership models. In addition, a multi-perspective
approach allows scholars to better understand how each of these leadership domains complement one another in affecting outcomes (Graen & Uhl-Bien, 1995) and how relationships among these domains may vary depending on the setting and type of outcome examined (Vecchio, 1982). Despite these predictive advantages, few leadership studies have examined multiple leadership perspectives (Basu & Green, 1997; Howell & Hall-Merenda, 1999; Scott & Bruce, 1994; Tierney, Farmer, & Graen 1999 are notable exceptions). These studies begin to uncover the additive and interactive effects of leader characteristics and style and the quality of relationship between leaders and their followers in predicting outcomes. However, due to the limited number of studies taking this approach,
researchers continue to be challenged to better understand how leader-, follower-, and relationship-oriented processes complement one another in affecting outcomes. This research augments existing knowledge on leadership by examining multiple leadership perspectives and their impact on individual-level outcomes. Consistent with the average-leadership style approach (i.e., leaders affect followers equally), early leadership theory (e.g., Ohio State and University of Michigan studies), and the work of early climate researchers (e.g., Lewin, Lippitt, & White, 1939; McGregor, 1960), leader priorities are examined as antecedents to work-group climate and represent an important aspect of the leader-focused perspective. Work-group climate is defined here as
"incumbents' perceptions of events, practices, and procedures and the kinds of behaviors
6
that get rewarded, supported, and expected in a setting" (Schneider, 1990: 384). I argue that leaders, through their consistent emphasis on specific priorities, provide cues to employees regarding valued behaviors and these cues are meaningfully interpreted by work-group members and can impact their goal orientation and associated performance. Furthermore, this research incorporates the leader-follower relationship-based approach by examining how the quality of leader-employee exchange (LMX) relationships affects group-members' goal orientation and performance outcomes. Toward this end, contemporary leader-member exchange theory and research will be used to hypothesize the effects of this more interpersonal, one-to-one relational leadership process. I employ the concept of leader priority to capture the achievement focus of the leader (i.e., the "what") and present the concept of LMX as a relational medium through which leaders transmit and emphasize their priority (i.e., the "how"). Taken together, these two approaches to leadership processes are expected to yield a richer, more comprehensive view of the impact of leadership on employee motivation and performance. In addition, this research offers a further improvement over much of the existing leadership and motivation literature. One underlying assumption of leadership theory is that leaders influence follower motivation, which in turn, affects follower performance. While some researchers have directly measured each component of this causal chain (e.g., Durham, Knight & Locke, 1997), this practice has not been consistently employed. For example, Shamir, House and Arthur (1996) noted that charismatic leadership theory and research is limited in its ability to fully explain the motivational process by which charismatic leaders affect follower attitudes and performance. Further, empowering
7
leadership approaches assume that leaders build intrinsic motivation among followers to lead themselves. However, notable studies in this area simply examine the relationship of empowering leadership to perceptions of leadership effectiveness by group members, neglecting effects on group member motivation (e.g., Manz & Sims, 1987). This direct leadership—outcome approach is problematic because of its inability to rule out alternative explanations and more completely test underlying theoretical assumptions. It is plausible, for example, that higher employee performance results may be due to higher performers being attracted and selected to a particular type of leader, as posited by the attraction-selection-attrition framework (Schneider, 1987) instead of the leader's behaviors. Or, a certain type of leader may be more likely to provide
challenging assignment to employees. The resulting performance gains in this case would not necessarily be due leadership per se, but rather the motivational stretch inherent in these types of assignments (McCauley, Ruderman, Ohlott & Morrow, 1994). The current research extends current theorizing and testing by clearly delineating the causal chain that links leadership processes to employee performance outcomes. More specifically, this research examines how leadership affects the adoption of specific types of employee goal orientation, and in turn, impacts performance. In sum, the research question addressed here is: can organizational leaders facilitate adoption of different forms of goal orientation among their employees, and in turn, affect their performance? And, if so, how do leaders impact this motivational process and its associated outcomes? It has been argued that this research contributes to our understanding of leadership, motivation, and performance in three distinct ways. First, by incorporating goal orientation as its central motivational variable, this study is
8
better able to examine the antecedents and consequences of a broader range of individual motivations that are consistent with contemporary organizational dynamics and demands. Secondly, this research takes a multi-perspective approach to leadership in order to provide richer, more comprehensive perspective of the effects of leadership. Lastly, it incorporates motivational processes into the proposed causal chain in order to more fully test underlying theoretical assumptions and adequately rule out alternative explanations.
9
CHAPTER 2 Model Development and Hypotheses Model Overview Kozlowski and Farr (1988) recommend that future research examine the underlying psychological mechanisms that influence individual learning and
performance. Building on their suggestion, the model proposed herein posits group members' state goal orientation as the key motivational mechanism impacting individual outcomes.1 Consistent with the socio-cognitive approach employed by early goal
orientation researchers (e.g., Dweck, 1986), the central mechanism underlying these relationships center on how unique goal orientation dimensions create different mental frameworks of the task situation that serve to structure cognitive, affective, and behavioral responses. In addition, the proposed model predicts that group members adopt a state goal orientation consistent with the types of behaviors that are rewarded, supported and expected, as communicated by the work-group climate. The logic
underlying this linkage is consistent with climate researchers who demonstrate a connection between climate and its effects on group members, such as shaping attributions of events, attitudes, and behavior (e.g., Hofmann & Stetzer, 1998).
1
The terms "group" and "team" are often used interchangeably in the academic literature and are defined as
"a collection of individuals who are interdependent in their tasks, who share responsibility for outcomes, who see themselves and who are seen by others as an intact social entity embedded in one or more larger social systems" (Cohen & Bailey, 1997, p. 241). However, some note that groups may differ from teams in the level of interdependence that exists among the members (e.g., Cohen & Bailey; Katzenbach & Smith, 1993). Here, I use the term "group" exclusively to convey that interdependence among members is not a necessary condition of this theoretical model.
10
Moreover, as argued in early work on early climate (e.g., Lewin et al, 1939), organizational behavior (e.g., McGregor, 1960), and leadership (i.e., Schein, 1992, the Ohio State and University of Michigan studies), the model presented here argues that climate is established through the leader's priority. Lastly, a moderating role of leadermember exchange is offered to better understand the combined effects of climate and leadership on group-member motivation. These hypotheses are grounded in social
exchange theory (Blau, 1964) and leader-member exchange research (e.g., Dansereau, Cashman, & Graen, 1973). The proposed model centers on three key themes: learning, performance, and avoiding failure. It is argued that these themes pervade work-group and individual dynamics to differentially impact group-member motivations and outcomes. The logic underlying this research suggests that a consistency exists across leadership, work-group climate, group-member state goal orientation, and associated outcomes. This consistency creates unique work group "worlds" that revolve around a particular theme (i.e., learning, performance, avoiding failure), in which leaders and their members act and perceive work-related events in congruence with the ascribed work-group emphasis. It is argued that emergence of a particular theme is largely initiated by the leader's priority on learning, performance, and/or avoiding failure, which drives the formation of a workgroup climate that embodies this priority. It is argued that each of these climates induce group members' adoption of the parallel form of goal orientation, and in turn, differentially impact individual-level outcomes. The quality of the exchange relationship between the leader and employee is viewed as a more relational, proximal leadership
11
process, serving to uniquely complement the existing backdrop established by the leader's priority and associated climate. These relationships are depicted in Figure 1. Figure 1. Proposed Model of Contextual Influences on Individual State Goal Orientation and Associated Outcomes
Leader Priority
Work-group Climate
• Climate for • Employee
development H1 learning H2
Group-member State Goal Orientation
Group-member Outcomes
H3
• Learning • Proving • Avoid
• Task
performance
• Climate for
performance
• Learning
strategies
• Goal
accomplishment
• Climate for
avoiding failure
• Defensive
behaviors
Member Perception of the Quality of Leader-Member Exchange
H4
The following discussion is organized into two sections.
The first section
discusses each of the core constructs depicted in the model to provide a construct definition and the proposed investigated level of analysis. The second section theorizes about each of the relationships depicted above. First, the relation between each specific leader priority and its respective, prototypical work-group climate will be specified. Second, the hypothesized mechanisms linking work-group climate and group-member state goal orientation will be detailed. Third, group members' state goal orientation and its relation to individual-level outcomes will be discussed. Lastly, an interaction between
12
of leader-member exchange and work-group climate will be proposed in affecting individual state goal orientation. Before discussion of these hypothesized linkages, I define each of the core constructs of the proposed model. Construct Definitions Leader Priority. A priority is defined as "something given special or prior
attention" (Random House, 1998, p. 570). Here, leader priority refers to the special attention leaders devote to employee development, goal accomplishment and/or avoiding failure. Dating back to the leadership studies conducted by scholars at Ohio State
University and the University of Michigan (as cited by Bass, 1990; Porter, Steers, & Bigley, 1996), researchers have long acknowledged that leaders "initiate structure," that is, they structure their own work and the work of group members. Setting priorities is one way leaders structure the work group's attention, work, and resources. Here, priorities and goals are viewed as distinct, but interrelated, concepts. First, the concept of priority captures the relative attention paid to a variety of goals, while goals can be examined in isolation of other demands. This notion of relativity is
particularly important in organizational settings where leaders and their members need to manage multiple goals but do not have adequate resources to attend to every goal immediately. Consequently, leaders and their members pay more attention to some goals than others. Secondly, Locke and Latham (1990) purposively choose the word "goal" for their goal-setting theory to connotate emphasis on the end result. Here, I use the concept of priority to include the ongoing process of emphasizing a particular outcome as well as the outcome itself. For this purposes of this study, this more inclusive concept better
13
captures the fuller leadership process surrounding the structuring of group members' attention on a particular focus. This approach of examining leadership with a particular strategic focus or emphasis is relatively new. However, a recent paper by Schneider, Ehrhart, Mayer, and Saltz (2002) examines service leadership, that is, a leader priority on delivering high quality service, as an antecedent to climate. Implicit in their logic is that leaders who prioritize service will be more likely to communicate their commitment through their behaviors. I adopt a similar approach here and propose that relatively consistent actions taken by a leader indicates his/her priority on employee development, goal accomplishment, and avoiding failure. While perhaps not explicit, much of existing leadership literature indicates the existence of these three priorities (i.e., learning, goal accomplishment, and the avoidance of failure). For example, Bass (1985; 1990) identifies individualized consideration and intellectual stimulation as dimensions of transformational leadership. In part,
individualized consideration refers to leaders' provision of developmental support to group members through coaching and/or delegating developmental assignments. Transformational leaders provide intellectual stimulation when they challenge group members to think in novel ways about work challenges, the organization, themselves, and/or the leader. In addition, Manz and Sims (1987; 1992) describe empowering
leadership as the process of leading others to become self-leaders. According to Manz and Sims (1987, 1992), this process involves coaching and developing group members to become leaders. In addition, team leadership researchers assert the need for leaders to emphasize learning and development of their members (e.g., Kozlowski, Gully, Salas, &
14
Cannon-Bowers, 1996; Zaccaro & Marks, 1999).
These examples drawn from
transformational, empowering, and team leadership theory may be viewed as behavioral representations of a leader's priority on employee development. Leader priority on goal accomplishment is evidenced both in academic and practitioner leadership literatures. For example, House's (1971) path-goal leadership model describes a contingency approach in which leaders adopt the best style to maximize follower motivation, and in turn, follower effort and performance based on characteristics of employees and environmental circumstances. In addition, many
scholars have discussed the importance of the establishment of a vision statement to communicate the overall direction and motivate organizational members to accomplishing objectives (e.g., Bass, 1990; Bennis & Nanus, 1985), such a statement communicates the priority of the leader on particular goals. Lastly, leader use of popular management techniques such as Management by Objectives that focus on setting objectives, developing actions plans, monitoring progress, and evaluating performance. These examples illustrate ways in which leaders communicate their priority on accomplishing set goals through their behavior. Behavioral representations of a leader's priority on avoiding failure can be seen in leadership, teams and politics research. Ancona and Caldwell (1992) found that leaders and their teams carefully selected what types of information would be shared with upper management as a way to avoid appearing incompetent. Lastly, Longnecker, Sims & Gioia (1987) found that leaders purposively misrepresented their employees' performance in the appraisal process, in some cases, to promote the appearance of competence of themselves or group members. Edmondson (1996) found that some nurse
15
managers blamed nurses for making mistakes, making them feel as though they were "on trial;" Edmondson (1996) theorized that these units reported less mistakes than other units to avoid the appearance of "failure." These examples demonstrate some of the ways in which a leader's priority to avoid failure to appear competent are expressed through their behavior. These examples illustrate how leader priorities are communicated through behavior and suggest a more complex, subtle process of how leaders transmit their beliefs about what is important. In Schein's (1992) classic work on leadership and culture, he details six mechanisms through which this transmission occurs. These six mechanisms serve as a conduit through which leaders express their priority to group members. More specifically, group members interpret leader behavior revolving around these six areas in ongoing leader-group interactions as reflective of an overall priority on learning, goal accomplishment, or avoidance of failure. Consequently, if relatively consistent, the
patterns of leader behavior converge to elucidate the leader's priority for the group and communicate his/her focus. Differing from the fundamental assumptions of leader-
member exchange theory, leader priorities are assumed to apply to the group as a whole. For this reason, the construct of leader priority is conceptualized as a group-level phenomenon. To better specify the leader behaviors indicative of the leader's priority, each of Schein's (1992) six mechanisms is described below and used to propose a range of leader behaviors indicative of each of the three priorities of interest: employee development, goal accomplishment, and avoiding failure.
16
What leaders pay attention to, measure, and control on a regular basis. This mechanism most directly communicates the leader's priority to group members. Their attention may take the form of questions, casual comments, rewards, priorities on meeting agendas, and measurement of specific behaviors. Conversely, powerful signals are sent to group members when a leader routinely ignores specific stimuli and indicates a lack of importance. Lastly, leaders' strong emotional reactions to specific events may be indicative of their priority on a particular aspect of group functioning. How leaders react to critical incidents and organizational crises. Leaders who label a set of circumstances as a "crisis" convey a level of urgency of the situation to the group members. Moreover, their endorsed method for addressing critical incidents (e.g., through learning, demonstrating ability, avoiding the appearance of incompetence) communicate their assumptions and beliefs about effective crisis management. Signaling a particular way to handle a problem helps to clarify the leader's priority on employee development, goal accomplishment and/or impression management. Criteria for resource allocations. Because resources are limited in organizations, decisions on how to allocate time, money, and other resources can indicate the leader's priority. Resource allocation decisions may, in turn, guide goal setting and identification of strategies for achieving objectives, serving to limit the range of alternative courses of action a leader can endorse for his/her group and elucidating a particular leadership priority. Role modeling, teaching, and coaching. Leaders often role model the behavior they value, expect and believe effective. According to Schein (1992), these informal messages serve as powerful teaching lessons for group members. In addition, a central
17
part of teaching and coaching involves providing feedback to employees.
During
feedback sessions, both the content (i.e., what the leader notices) and focus (i.e., what values are implicit in the leader's message) embed and reinforce the leader's priority within the work-group climate. Criteria for allocation of rewards and status. promoting their ascribed priority through rewards. Leaders reinforce behaviors Rewarding valued behavior
continuously emphasizes their priority. By observing promotion cycles, participating in performance appraisals, discussing performance expectations with their manager, and observing what types of behaviors receive positive and negative acknowledgements from the leader, employees learn what the leader values. Criteria for recruitment, selection, retirement, and excommunication. Schein's (1992) basic assertion is that leaders base their hiring and firing decisions based on the level of congruence between the leader's values and priorities and those expressed by the job candidate or incumbent. These decisions cue group members as to what is important to the leader and facilitate a longer-term shared value system among the group members. Because I hope to sample relatively mature, intact work groups, it is more instructive to examine the criteria leaders use to determine job assignments. Just as with hiring and firing decisions, these criteria signal to members the emphasis the leader places on employee development, goal accomplishment, and/or avoiding failure. Behavioral Indications: Leader Priority on Employee Development. Leaders indicate a priority for employee development when they consistently inquire about how well group members are developing skills and knowledge, consistently discuss how different job assignments may contribute to employees' long-term growth, encourage
18
employees to set developmental goals for themselves, and/or hold regular on-one-one meetings to discuss developmental issues with group members. They will often provide resources (i.e., coaching, training, time off to engage in developmental activities) to facilitate employee development. Moreover, they view job assignments as potential opportunities for development, and make selection decisions based on how the assignment will contribute to the group member's growth. In crisis situations, they react by encouraging group members to experiment with different strategies and remain open to learning from these challenging situations; and, they often model this behavior for their employees. Leaders who prioritize employee development offer more feedback to group members and focus their coaching on skill improvement, rather than solely providing knowledge about the outcome of their performance. They consistently reward group
member effort and a dedication to development and growth. Behavioral Indications: Leader Priority on Goal Accomplishment. express their priority on goal accomplishment when they consistently measure performance against a set standard, pay almost exclusive attention to meeting task standards, measure success only in terms of accomplishing a set performance goal, and routinely ignore why objectives may not have been met, instead choosing to classify the experience as a failure. They allocate resources only to those activities that promote getting the work done. In crisis situations, leaders emphasizing goal accomplishment routinely call on group members who have a demonstrated track record in handling similar circumstances to resolve the problem. If necessary, the leader will step in so that accomplishment of task goals is not compromised. Similarly, in making staffing Leaders
decisions, leaders who prioritize goal accomplishment assign jobs and tasks based on
19
demonstrated ability. Moreover, they consider who will have the greatest likelihood of meeting the standards associated with the task, and therefore, allow the leader and group to receive positive judgments regarding their competence. Feedback from leaders
focused on goal accomplishment simply assesses group member performance against set standards and does not provide insight into how members may improve. Lastly, leaders emphasizing goal accomplishment consistently rewards employees for high levels of performance through recognition, promotion, and other extrinsic rewards. Behavioral Indications: Leader Priority on Avoiding Failure. Leaders who
prioritize the avoidance of failure consistently focus on how competent they or their group members appear to others, measure success in terms of number of errors made, and ignore opportunities to experiment and learn. Their focus is on the appearance of
competence, even if they or their members are not capable. They define a crisis as someone making a mistake and this error becoming known; they deal with critical situations by either avoiding discussion of the error or by "spinning" organizational events so that the leader and his/her employees avoid being characterized as incompetent. Considerable resources, particularly time, are devoted to best positioning ideas and contributions to organizational stakeholders in the hopes of avoiding negative judgments regarding group members' ability. These leaders delegate "safe" assignments to group members and approach assignment decisions with the intent of minimizing errors. They encourage group members to avoid challenging situations that may threaten the appearance of being capable and role model how to appropriately manage impressions so as to appear skilled. These leaders rewards group members for "not screwing up" with promotions, awards, and verbal recognition.
20
In sum, a leader priority refers to the special attention leaders devote to a particular aspect of group functioning. In this study, I focus on three leader priorities: employee development, goal accomplishment and avoidance of failure. While not
explicitly acknowledged in the literature, behavioral expressions of each of the proposed leader priorities have been studied previously. To augment existing research and provide further elaboration, I have employed Schein's (1992) work to articulate the types of behaviors that convey a leader's priority on employee development, goal
accomplishment, and/or avoiding failure. These behaviors are summarized in Table 1 below. Table 1. Behavioral Indications of Leader Priority
Embedding Mechanism (Schein, 1992) What leaders pay attention to and measure How leaders react to crises Leader Priority on Employee Development Employee development Leader Priority on Goal Accomplishment Goal accomplishment Leader Priority Avoiding Failure on
Experiment; learn from mistakes
Send in the person with demonstrated expertise to resolve problem
Leader criteria for resource allocation Leader coaching, feedback teaching, and
Allocate resources to facilitate group member development Focused on skill improvement
Allocate resources to further progress toward operational objectives Focused on providing assessments of performance relative to objectives/ standards Outstanding performance "Who has demonstrated they can do the job?"
Leader criteria for awards Leader criteria for job assignment
Improved performance and level of effort "Who needs development in this area?"
Others' perceptions of leader and group member capabilities Avoid discussions of errors or manage impressions to avoid negative competence judgments Dedicate resources (i.e., time) to managing impressions Focused on avoiding challenging situations and managing the appearance of competence For "not screwing up" "Who will most likely not make errors?" (i.e., placement into "safe" assignments)
21
Work-group Climate: Construct Definition.
Schneider defined climate as
"incumbents' perceptions of events, practices, and procedures and the kinds of behaviors that get rewarded, supported, and expected in a setting" (1990: 384). This definition suggests two broad classes of perceptions: routines (i.e., the events, practices and
procedures), and rewards (i.e., behaviors that are rewarded, supported, and expected) (Schneider, 1990). In this way, climate is a perceptual interpretation of environmental cues offered by the routines and rewards of the particular setting through which employees may understand what is important and valued. With the exception of a recent trend to explore the effects of varying degrees of shared perceptions (e.g., Gonzalez-Roma et al, 2002; Lindell & Brandt, 2000; Schneider, Salvaggio, & Subirats, 2002), climate scholars consistently focus on the shared perceptual nature of climate in their research. Implicit in the definition of climate is the understanding that relatively little variance in organizational member perceptions exist; otherwise, shared interpretations of the setting would not emerge. Consistent with this tradition, theoretical discussions of climate in this proposal assume a high level of perceptual agreement among members and a relatively consistent view of the workrelated events and practices across individuals in a particular work group. Because early climate research examined such a variety of dimensions, the boundaries of what constituted climate became blurred (Schneider, 1975). For this
reason, Schneider (1975) recommended conceptualizing climate with respect to a specific referent, such as service, quality, safety. In other words, climate is best defined, and in turn, understood when considering a central strategic focus (Schneider, 1990). Examples of climates with a strategic focus include a climate for safety (e.g., Zohar, 1980), a
22
climate for service (e.g., Schneider, 1990), a climate for updating skills (e.g., Kozlowski & Hults, 1987), transfer of training climate (e.g., Tracey, Tannenbaum, & Kavanagh, 1995) and in the present case, a climate for learning, a climate for performance, and a climate for avoiding failure. In addition, climate has been examined at multiple levels of analysis including the organization (e.g., Pritchard & Karasick, 1973), the sub-unit, such as a branch of a bank or a plant (e.g., Joyce & Slocum, 1984; Schneider, White & Paul, 1998), the work-group (e.g., Hofmann & Stetzer, 1996), and the individual, also referred to as psychological climate (e.g., Jones & James, 1979), levels. Here, the focus is on the work-group level of analysis, and the climate construct at this level of analysis has received considerable empirical support for its validity (Gonzalez-Roma et al, 2002; Howe, 1977; Powell & Butterfield, 1978; Rentsch, 1990; Zohar, 2000). Work-Group Climate: Dimensionality. In my review of approximately twentyfive empirical studies on climate, close to fifty distinct dimensions have been assessed and measured. Despite the wide variance, however, several dimensions emerge as being the most commonly used in research: 1. Structure of the work environment—formality of the work environment, value placed on protocol; used in 29% of the studies reviewed. 2. Rewards—use of various organizational tangible and intangible rewards; used in 29% of the studies reviewed. 3. Support—availability of necessary resources and social support and encouragement; used in 50% of the studies reviewed.
23
4.
Teamwork and cooperation—mutual support among team members; used in 33% of the studies reviewed.
5.
Participation and involvement in decision making—level of influence in work activities and decisions; used in 29% of the studies reviewed.
6.
Goal emphasis—types of outcomes and standards expected; used in 29% of the studies reviewed.
7.
Communication flow—the frequency of communication; used in 29% of the studies reviewed.
8.
Supervisory/managerial values and practices—the actions, values, expectations, and level of commitment demonstrated by the supervisor and/or higher-level manager; used in 33% of the studies reviewed. While these dimensions surfaced as being the most commonly used, some are
more appropriate for inclusion in this study than others. For example, the dimension of teamwork and cooperation is defined as "the extent to which employees perceive that a friendly cooperative work environment exists" (Kozlowski & Doherty, 1989, p. 549). This dimension describes the general environment and is too generic for use in this research. More specifically, this dimension does not help to define distinguishing
characteristics across the three climates of interest in this dissertation. For example, would a climate that had a high level of teamwork and cooperation be more indicative of a climate for learning, a climate for performance, or a climate for avoiding failure? In attempting to tailor this dimension more specifically to highlight distinguishing features across these three settings, the "teamwork and cooperation" dimension becomes redundant with the "support dimension."
24
In addition, studying supervisory practices as a part of climate blurs the distinction between leadership and climate. Perhaps, the frequency with which
researchers measure managerial practices as a part of climate is in part due to the relative rarity of explicitly testing leadership as an antecedent to climate (Barling, Loughlin & Kelloway, 2002; Kozlowski & Doherty, 1989; Schneider et al, 2002 are notable exceptions). However, most recently, researchers have distinguished between these two constructs by specifying leadership as the guiding force or the influence that sets a direction for group members and the work-group climate as the perceptual medium that interprets and implements this direction through a series of ongoing, complex, social interactions among group members, other organizational stakeholders, customers, and the leader (personal communication: B. Schneider, November 20 2002). Furthermore, the dimensions of work-group climate serve to characterize the patterns of these interactions perceived by group members. Seen in this light, it becomes imprecise, and therefore inappropriate, to consider supervisory practices as a part of work-group climate. The other six dimensions align well with the construct definition of climate and are not too generic, nor redundant. They have been adapted so as to properly describe the three climates of interest in this study (i.e., learning, performance, and avoiding failure). Adjustments to the dimensions described above primarily center on more directly assessing the content and focus of rewards and routines, rather than their frequency or level. Incorporation of these changes yield the following climate dimensions and
definitions: 1. Structure of the work environment—extent to which group members perceive that protocols, standardized practices, and means of accomplishing learning,
25
performance of tasks, and/or avoiding failure are clearly specified (adapted from Abbey & Dickson, 1983; Burke, Borucki & Hurley, 1992; Kopelman, Brief & Guzzo, 1990; Kozlowski & Doherty, 1989; Ostroff, 1993; Solomon, 1986; Tagiuri & Litwin, 1968). 2. Rewards—the extent to which group members perceive that various rewards are used to encourage and acknowledge learning, performance, and/or avoiding failure (adapted from Abbey & Dickson, 1983; Burke et al, 1992; Joyce & Slocum, 1984; Kopelman et al, 1990; Ostroff, 1993; Solomon, 1986; Tagiuri & Litwin, 1968). 3. Support—the extent to which group members perceive that the necessary resources and social support, including feedback from colleagues, is available to promote learning, performance, and/or avoiding failure (adapted from Abbey & Dickson, 1983; Burke et al, 1992; Gonzalez-Roma et al, 2002; Jones & James, 1979; Kopelman et al, 1990; Kozlowski & Doherty, 1989; Kozlowski & Farr, 1988; Kozlowski & Hults, 1988; Rouiller & Goldstein, 1993; Solomon, 1986; Tagiuri & Litwin, 1968; Tracey et al, 1995). To reduce redundancy with the leadership construct, this dimension only focuses on support provided by the organization and colleagues, not work-group leaders. 4. Participation and involvement in decision making—the extent to which the group members perceive their involvement in decision making regarding learning, performance, and/or avoiding failure issues (adapted from Abbey & Dickson, 1983; Drexler, 1977; Kozlowski & Doherty, 1989; Kozlowski & Farr, 1988; Ostroff, 1993; Solomon, 1986).
26
5. Goal emphasis—the extent to which group members perceive learning, performance, and/or avoiding failure as valued and expected outcomes (adapted from Burke et al, 1992; Gonzalez-Roma et al, 2002; Kopelman et al, 1990; Kozlowski & Doherty, 1989; Lindell & Brandt, 2000; Rouiller & Goldstein, 1993; Tracey et al, 1995). 6. Communication—the extent to which group members perceive that work-group communication centers on learning, performance, and/or avoiding failure (adapted from Drexler, 1977; Griffin & Neal, 2000; Jones & James, 1979; Kozlowski & Doherty, 1989; Kozlowski & Hults, 1987; Neal, Griffin, & Hart, 2000; Rouiller & Goldstein, 1993; Tracey et al, 1995). Adequate measurement of work-group climate will hinge on the applicability of these dimensions to the context of the sampled organization. For this reason, these dimensions will be further validated for their appropriateness, as well as the identification of additional dimensions, during the qualitative stage of data collection (please see Chapter 3 for more details). Work-group Climate: Description of Three Prototypical Climates. These six
dimensions will be used in describing three prototypical work-group climates (i.e., climate for learning, a climate for performance, a climate for avoiding failure). These climates are not mutually exclusive; instead, each of these climates may be present within the same work group to varying degrees. Because limited organizational resources
demand focused attention on a limited set of valued behaviors and outcomes, it is likely that one, perhaps two, particular climate(s) is/are more dominant than the others as a way to help work-group members focus on the more important priorities.
27
Climate for learning. The concept of a climate for learning builds on previous work in the employee developmental literature and has been referred to as a "continuous learning culture" (Tracey et al, 1995) and "a climate for technical updating" (Kozlowski & Farr, 1988; Kozlowski & Hults, 1987). Training researchers have explored a similar concept, namely a "transfer of training climate" (Burke & Baldwin, 1999; Rouiller & Goldstein, 1993). Consistent with this prior research, a climate for learning emphasizes continuous development of knowledge and skills. Employees perceive that continuous learning and ongoing engagement in developmental activities is valued, supported, and the expected means for achieving individual and work-group goals (Rosow & Zager, 1988; Tracey et al, 1995). In this type of climate, informal practices and formal systems provide opportunities for development and reinforce learning (Dubin, 1990). Intrinsic rewards such as satisfaction gained from continuous improvement and learning are stressed as important rewards; extrinsic rewards, such as public acknowledgement for learning and promotion based on demonstrated dedication to continuous improvement of one's skills, are also emphasized. Co-workers in climates for learning provide the necessary social support, challenge, encouragement to learn, and feedback to assist in developing and maintaining motivation in the face of challenges (Dubin, 1990; Kram, 1985; Noe & Wilk, 1993; VanVelsor et al, 1998); it is perceived that resources are available, such as training, time to engage in development, and access to other developmental activities (Dubin, 1990; Maurer & Tarulli, 1994; Morrison & Brantner, 1992; Noe & Wilk, 1993). Moreover, in this type of climate, incumbents perceive themselves as being an active participant in
28
deciding when, how, and which skills they need to develop. Lastly, given this description of the structure, rewards, support structure, and decision-making, it is probably not surprising that issues of employee development, learning, feedback, and developmental activities are perceived as being commonly discussed within the group. Climate for performance. The construct of a climate for performance is similar to what safety climate researchers have referred to as "high reliability organizations." High reliability organizations are defined as "organizations in which complex technology must be controlled and complex processes carried out in an error free manner" (Hofmann, Jacobs & Landy, 1996, p. 139). Moreover, within the oil industry, Wright (1986)
observed an organizational pressure to complete the work as quickly as possible. The central theme here is on high levels of productivity, without error. A climate for
performance is a bit broader than observations from the safety climate literature in that performance is not limited to productivity levels and/or efficiency. Performance, as in many knowledge-based organizations, also refers to effectiveness in completion of complex tasks or projects. In a climate for performance, employees perceive achieving the performance standard as the most valued outcome. It is expected that these performance standards will be attained efficiently, effectively, and accurately through applying team members' current skills, knowledge, and abilities. Informal practices and policies support the
accomplishment of work tasks, specifying clear methods or task approaches. Extrinsic rewards, such as promotion and salary, are consistently provided for individuals who are viewed as having high ability in meeting or exceeding performance standards. Group
29
members perceive satisfaction (i.e., intrinsic rewards) when particularly difficult, challenging performance standards are met (Locke & Latham, 1990). Perceived support focuses on removing possible barriers to achieving performance goals, such as a lack of materials, information, staff, or time. In addition, group members support one another to accomplish the task by providing advice on difficult issue and encouragement to meet performance standards. In this type of climate, group members perceive themselves as involved in decisions regarding performance issues, such as goal-setting and performance strategies and approaches, and work-group communication primarily centers around goals, goal accomplishment, and performance strategies. Many of these points complement Locke and Latham's (1990) high performance cycle in which they detail how performance goals and standards motivate high levels of performance. Consistent with their view, rewards in this climate are seen as contingent upon performance, and situational constraints are seen as barriers to performance, and therefore, need to be removed. While the concept of a climate for performance
emphasizes more of the environmental conditions, Locke and Latham (1990) work parallels the central emphasis of this particular climate and supports some of its major distinctions. Climate for avoiding failure. In this context, avoiding failure (i.e., making errors) is extremely valued because it is perceived to be instrumental in maintaining the appearance of competence. Being labeled as "incompetent" risks future opportunities (e.g., promotion), limits access to resources (e.g., information, salary increases), and respect and acceptance from colleagues and organizational leaders. The teams and
30
climate literatures documents how some work groups foster a climate that is not supportive of admitting errors, discussing mistakes, or offering dissenting opinions (e.g., Edmondson, 1996; 1999; Hofmann & Stetzer, 1998). In a climate for avoiding failure, the common practice is to engage in "face saving" behaviors (e.g., avoid discussions of errors and mistakes, asking for help, shifting blame). Typically, group members perceive that they are encouraged to avoid blame, resist taking on assignments that risk failure, and/or avoid taking action so as to evade revealing low ability of a particular group member. Group members perceive being rewarded for "not screwing up" with verbal acknowledgements, promotion, and salary increases. It is perceived that committing errors carry significant punishment such as constant reminders of one's mistake and brutal, demeaning confrontations by others (Edmondson, 1996). Because this environment is unforgiving of even the slightest error, group members do not perceive the same type of social support from their colleagues as in a climate for learning or performance. Here, support comes in the form of covering up for one another. This phenomenon of attributing mistakes to external events to shift blame away from group members has been documented in Hofmann and Stetzer's (1998) work on safety climate. Moreover, because errors are not discussed, group members in this climate are not involved in decisions regarding how to improve group functioning. Work-group discussions center around anything but errors and mistakes, potentially at the expense of addressing key issues preventing members from effectively performing their jobs.
31
In summary, work-group climate has been defined as work-group members' "perceptions of events, practices, and procedures and the kinds of behaviors that get rewarded, supported, and expected in a setting" (Schneider, 1990: 384). In this
dissertation, three prototypical climates will be examined: a climate for learning, a climate for performance, and a climate for impression management. These climates have been described using standard dimensions derived from review of existing literature. A summary of these descriptions can be found in Table 2. Table 2. Descriptions of Three Prototypical Work-group Climates
Climate Dimensions Goal emphasis Climate for Learning Continuously developing and learning new skills, knowledge, and capabilities Formal systems and practices provide opportunities for development and reinforce learning For learning Climate for Performance Achieving performance goals and standards Climate for Avoiding Failure Avoid committing and admitting to mistakes
Structure
Formal systems and practices specify how best to achieve performance goals For high performance
Informal practices endorse avoiding discussion of errors
Rewards
Support
Availability of developmental activities; encouragement and feedback on how to develop from peers Involved in decisions pertaining to one's development Focuses on learning and development
Involvement in decision making Communication
Availability of resources necessary to meet performance goals; encouragement and coaching from peers on how to accomplish specific tasks Involved in decisions pertaining to goal accomplishment Focuses on meeting performance goals
For "not screwing up;" severely punished for making mistakes Group members covering up for one another
Not involved in decisions regarding improvement issues Focuses on anything but mistakes and errors
Goal orientation: construct definition. As previously discussed, goal orientation is defined as one's goal preference in achievement settings (Dweck, 1986) and refers to one's desire to develop, realize, or exhibit the capability to perform a specific activity
32
(Dweck & Leggett, 1988). Initial research identified two broad classes of underlying goals that individuals pursue: mastery and performance (Dweck, 1986; Dweck & Leggett, 1988). Individuals displaying a mastery or learning orientation focus on Those favoring a
building one's competence and/or improving their abilities.
performance orientation seek to demonstrate their competence by seeking favorable judgments and/or avoiding negative judgments regarding their capabilities. Goal orientation provides a mental framework that affects how individuals view challenging situations and shapes their responses in these situations (Dweck, 1986; Dweck, 2000; Dweck & Leggett, 1988). The causal mechanism underlying these two broad classes of goals is one's implicit beliefs regarding intellectual ability and effort. Individuals with a learning orientation hold an incremental view of ability and believe their competencies may be developed through effort and experience. In this view, effort and improved task strategies lead to success. In contrast, individuals with a performance orientation ascribe to an entity view of ability, believing that ability is a fixed personal attribute that cannot be changed. According to this mental framework, higher effort is an indication of lack of ability because individuals who are naturally talented should not have to try so hard to be successful. Goal orientation: dimensionality. Our understanding of the dimensionality of goal orientation has evolved. Initially, Dweck and her colleagues left the question of dimensionality unclear, although some scholars have suggested early work posited goal orientation as uni-dimensional (Button et al, 1996; VandeWalle, 1997). The first
measure developed specifically for adult populations was developed and validated by Button and his colleagues (1996). In this research, they provided support for two, distinct
33
dimensions of goal orientation: learning and performance. Their empirical approach was consistent with the goal orientation theory and research of the time and rigorously compared various models consisting of different combinations of learning and performance orientations. No conceptual work at that time had suggested the potential for a three-factor model, and therefore, they did not test this possibility. More recent empirical evidence supports further partitioning performance orientation into two components, i.e., prove and avoid, arguing that "seeking favorable judgments" and "avoiding negative judgments" are two distinct goals (Elliot & Harackiewicz, 1996; Middleton & Midgley, 1997; VandeWalle, 1997). Accordingly, individuals with a proving orientation prefer tasks that allow them the opportunity to demonstrate their capabilities and gain favorable feedback. Their avoiding-oriented
counterparts avoid responsibilities that may reveal their shortcomings to others. It is important to note that both individuals with a proving orientation and individuals with an avoiding orientation subscribe to the entity theory of intellectual ability. Goal orientation: malleability. Researchers consider goal orientation to be a stable, dispositional construct that is susceptible to situational influences. Researchers treating goal orientation as a dispositional trait, measure goal orientation, rather than induce it (Bell & Kozlowski, 2002; Brett & VandeWalle, 1999; Brown, 2001; Chen et al, 2000; Colquitt & Simmering, 1998; Elliot & Church, 1997; Elliot & McGregor, 1999; Fisher & Ford, 1998; Ford et al, 1998; Middleton & Midgley, 1997; Potosky & Ramakrishna, 2002; Phillips & Gully, 1997; Towler & Dipboye, 2001; VandeWalle et al, 1999; VandeWalle, Cron, & Slocum, 2001; VandeWalle & Cummings, 1997). This body
34
of research predicts specific behavioral and motivational outcomes based on individual differences in goal orientations. Researchers treating goal orientation as more malleable induce specific forms of goal orientation through situational influences and often refer to this construct as state goal orientation. A state orientation is a temporary preference for specific achievement goals. Dweck and Leggett (1988) suggest that when the situation offers few cues as to which orientation to adopt, individuals will rely on their dispositional goal orientation. If, on the other hand, the circumstances increase the value or saliency of a specific orientation, "predispositions should be overridden and greater homogeneity among individuals will result" (1988: 269-270). These arguments are consistent with Mischel (1977) who asserts that situational strength impacts the extent to which dispositional variables become observable through behavior. According to Mischel (1977), strong situations encourage individuals to
construe the situation similarly, induce conformity in expectations regarding appropriate behaviors, reward expected behaviors, and require skills individuals possess. On the other hand, weak situations allow for variation in perceptions and behavioral responses. Personality researchers suggest that situations such as experimental settings create strong situations through clearly articulated expectations and requirements and structured tasks (Schneider, 1983; Weiss & Adler, 1984). Consequently, in strong situations, personality influences do not emerge as powerful predictor of behavior. Empirical evidence documents that trait goal orientation has a state counterpart. Through confirmatory factor analysis, Button et al (1996) showed that dispositional and state goal orientations are distinct dimensions, and situational influences may cause
35
individuals to adopt an orientation different than their dispositional one. Additional research has documented that the strength of specific situational cues, such as framing of the task, competition, extrinsic rewards, and evaluations standards, cause individuals to adopt a state goal orientation (Ames & Archer, 1988; Butler, 1993; Elliot & Harackiewicz, 1996; Elliott & Dweck, 1988; Harackiewicz & Elliot, 1993; Nicholls, 1984; Stevens & Gist, 1997). In studies by Butler (1993), Elliot and Harackiewicz (1996), Harackiewicz and Elliot (1993), and Stevens and Gist (1997), researchers induced state goal orientations through experimental manipulations. Adult participants were
randomly assigned to an experimental condition in which a different goal (i.e., performance or mastery) for the experimental task was emphasized. For other studies, state goal orientation was induced through task instructions that emphasized the fixed or incremental nature of ability (Mangos & Steele-Johnson, 2001; Steele-Johnson et al, 2000). Even though the framing of the task was subtle, it proved effective in inducing state goal orientations among participants. Studies conducted with samples of children demonstrate a similar pattern of results. With a sample of fifth graders, Elliott and Dweck (1988) randomly provided each student with a learning or performance goal. Students with the performance goal consistently demonstrated behavior consistent with a performance orientation. Likewise, students with the learning goal exhibited behavior consistent with a learning orientation. Similarly, Ames and Archer (1988) examined how the teacher's emphasis on performance or learning goals in the classroom affected junior high and high school students. Perceptions of learning or performance goals led to similar behavioral patterns
36
found by Elliott and Dweck (1988). These studies highlight the potency of goals and the situational influence of an authority figure on state goal orientation. Given the most recent empirical evidence on the dimensionality and malleability of goal orientation, three dimensions of state goal orientation will be investigated at the individual level of analysis in this dissertation. The current study extends this previous work by examining state goal orientation in an organizational setting with a working adult population, rather than in an experimental setting. This methodology will allow for greater understanding of how naturally occurring contextual factors, such as leadership and work-group climate, affect individual state goal orientation and associated outcomes. Individual outcomes. To better capture the multi-faceted construct of
performance, three primary categories of outcome variables are investigated in this dissertation: task performance, learning strategies, and defensive behaviors (e.g., avoiding challenge, blaming, excuse making). The more specific nature of the outcome variables examined in this study, in part, will be driven by the context of the sampled organization, which has yet to be determined. Therefore, I provide below specific
definitions of these constructs as well as examples from previous research as a way to clarify the meaning of the outcome variables of interest. Task performance refers to the quality, accuracy, and quantity of performance associated with execution of the task and will be investigated at the individual-level of analysis. Examples of task performance from previous studies include exam performance (e.g., Chen et al, 2000; Colquitt & Simmering, 1998; Elliot & McGregor, 1999; Phillips & Gully, 1997), quality of presentation (e.g., Brett & VandeWalle, 1999), performance of skills learned in training and transferred to another context (Ford et al, 1998; Kozlowski
37
et al, 2001), quality of performance on problem-solving tasks (e.g., Butler, 1993), sales volume (e.g., VandeWalle et al, 1999), and accuracy on problem solving tasks (e.g., Steele-Johnson et al, 2001; Mangos & Steele-Johnson, 2001). As argued by other scholars (e.g., Hall, 2002), individual learning and adapting are necessary for employee effectiveness given changing business, organizational and career demands. Learning refers to "a relatively permanent change in knowledge or skill produced by experience" (Weiss, 1990, p. 172). This definition suggests the need to document a relatively stable change through a comparison of pre- and post-assessments of knowledge and/or skill and is difficult to study in cross-sectional designs, such as the one proposed here. For this reason, this dissertation adopts the same approach taken by others and examines precursors of learning, that is, learning strategies (e.g., Fisher & Ford, 1998; Ford et al, 1998; Kozlowski et al, 2001). Learning strategies refer to "an internal process by which learners select and modify their ways of attending, learning, remembering, and thinking" (Gagne, Briggs, & Wager, 1992, p. 66). These strategies become apparent through exhibited behavior. Examples of these types of strategies studied by previous researchers include feedback seeking (e.g., Butler, 1993; Middleton & Midgley, 1997; VandeWalle & Cummings, 1997), metacognition (e.g., Ford et al, 1998), and other strategies, such as practice, self-maintenance activities, organizing knowledge into coherent structures, and elaboration (e.g., Brown, 2001; Fisher & Ford, 1998; Kozlowski et al, 2001; Stevens & Gist, 1997). Consistent with the research cited above, these types of behaviors are viewed as more proximal indicators of learning and will be investigated at the individual-level of analysis.
38
According to some scholars, one barrier to achieve the learning and adaptation necessary in today's business environment is defensive behaviors (e.g., Edmondson, 1996). Defensive behaviors are defined as "reactive and protective actions intended to reduce a perceived threat to or avoid an unwanted demand of an individual or group" (Ashforth & Lee, 1990, p. 622) and may include avoiding delivering "bad news," avoiding blame, and avoiding taking action that might implicate oneself. behaviors have been studied as a part of political behavior (Valle, 1997), means to managing mentor-protégé relationships (Tepper, 1995), and as a barrier to team learning and effectiveness (Ancona & Caldwell, 1992; Edmondson, 1996). Conventional These
wisdom and the extant literature, albeit it limited, suggests that these behaviors are largely dysfunctional and negatively impact group performance and functioning. Here, they are particularly relevant to understanding how different dimensions of goal orientation may affect the frequency in which individuals engage in these potentially dysfunctional behaviors, an area of study that has received no research attention to date. Quality of leader-member exchange. The construct of leader-member exchange relationships, also commonly referred to as LMX, is built on two fundamental and interrelated theoretical perspectives: role making and social exchange theory (Blau, 1964; Graen, 1975). Graen (1975) posited that dyads of interdependent organizational
members engage in the role making process to define how each will behave and agree on the general nature of their relationship. In particular, the relationship between the leader and subordinate, also referred to as vertical dyad linkages in early LMX research, becomes especially important in shaping role expectations of the member (Graen, 1975; Graen & Cashman, 1975; Dienesch & Liden, 1986), and therefore, is the central
39
relationship of interest. Moreover, through the role-making process, differentiated role definitions develop between leaders and each of their members, resulting in varied leader-member exchange relationships within a particular group. Drawing on social exchange theory (Blau, 1964), LMX researchers have argued that the basis for exchange relationships lie in the valued resources each party may provide (e.g., Liden, Sparrowe & Wayne, 1997). For example, leaders may provide challenging assignments, mentoring, and/or informational resources. In exchange,
members offer valued resources such as greater levels of effort and/or loyalty. Based on this theoretical background, the most commonly-used working definition of leader-member exchange is the "quality of the exchange relationship between leader and subordinate" (Schriesheim, Castro, & Cogliser, 1999, p. 77). Often times, LMX is operationalized by having leaders, their members, and/or both parties respond to a set of items assessing the quality of the relationship. The analysis is then conducted at the individual level using the perceptions from the leader, member or both. Recently, however, researchers have questioned this practice because of the logical inconsistency of theoretically defining the construct's level of analysis at the dyadic or relationship level and testing the hypothesized relationships at the individual level (Schriesheim et al, 1999; Schriesheim, Castro, & Yammarino, 2000). While this is a valid criticism, for the purposes of this dissertation, the compelling force believed to drive whether a group member adopts a particular state goal orientation is the member's perception of the quality of the relationship, not necessarily the dyad's assessment of the quality of this relationship. Focusing on member perceptions is logically consistent with examining the role of work-group climate, another perceptual medium. For these
40
reasons, this dissertation will focus on the member's perception of LMX, and it will be investigated at the individual level of analysis. Summary of Construct Definitions and their Level of Analysis. In sum, this dissertation will focus on five key constructs: leader priority, work-group climate, group member state goal orientation, individual outcomes, and member's perception of the quality of leader-member exchange relationship. Each of these constructs has been defined in the preceding discussion, and these definitions are summarized in Table 3 below. Table 3. Construct Definitions
Construct Definition Proposed Investigated of Analysis Group Group Level
Leader priority Work-group climate
State goal orientation Individual Outcomes -Task performance Learning strategies -Defensive behaviors
Member perception of the quality of LMX relationship
An aspect of the group functioning given special attention by the leader (adapted from Random House Dictionary). Group members' "perceptions of events, practices, and procedures and the kinds of behaviors that get rewarded, supported, and expected in a setting" (Schneider, 1990, 384). A temporary preference for specific achievement goals. Task performance: the quality, accuracy, and quantity of performance associated with execution of the task Learning strategy: "an internal process by which learners select and modify their ways of attending, learning, remembering, and thinking" (Gagne, Briggs, & Wager, 1992, p. 66). Defensive behaviors: "reactive and protective actions intended to reduce a perceived threat to or avoid an unwanted demand of an individual or group" (Ashforth & Lee, 1990, p. 622) "quality of the exchange relationship between leader and subordinate" (Schriesheim, Castro, & Cogliser, 1999, p. 77).
Individual Individual
Individual
Hypothesized Relationships As highlighted previously, the model proposed herein posits group leaders' priorities as the key drivers of three unique climates described in the preceding discussion. It is predicted that group members adopt a state goal orientation consistent
41
with the types of behavior that are rewarded, supported and expected, as communicated by the work-group climate. In turn, group members' state goal orientation is posited as the key motivational mechanism impacting individual outcomes. Lastly, it is theorized that the effects of climate and the role of the exchange relationship interact to impact group member state goal orientation. In the following section, the rationale for each of these hypotheses is detailed in the order highlighted above. Relation between leader priority and work-group climate. There is a strong
historical tradition documenting the key role of leadership in establishing work-group climate. In their classic works, Likert (1967) and McGregor (1960) asserted that leaders transmit their beliefs and expectations through the climate they create. McGregor (1960) suggests that managers convey their attitudes about workers, consistent with Theory X or Theory Y, through their behavior. In turn, this behavior establishes the work-group climate. Lewin et al (1939) empirically tested the role of leadership in establishing workgroup climate. Across three leadership conditions (i.e., democratic, authoritarian, and laissez-faire), Lewin et al (1939) found notable differences in followers' attitudes and behaviors between the groups. More recently, Eden and his colleagues' research on the Pygmalion Leadership Effect (2000) shows how leaders transmit their expectations and beliefs regarding their group member ability to create a supportive climate. These
findings suggest a group-level phenomenon that is shaped by the leader and, in turn, affects the behaviors and attitudes of group members. Schneider and Reichers (1983) provide three theoretical explanations for why climates emerge, two of which are relevant to this discussion. First, citing Payne and Pugh (1976), they suggest one way in which climate is influenced is through
42
organizational members' perceptions of objective characteristics of the work setting, such as the extent of centralization of decision making and the degree to which rules and policies constrain behavior. According to Payne and Pugh (1976), this structure shape how organizational members perceive organizational events. This structural argument complements early and contemporary leadership theory. According to researchers involved in the Ohio State and University of Michigan studies (as cited in Bass, 1990; Porter, Steers, & Bigley, 1996), leaders structure the work of their employees through rules, policies and goal setting. This basic idea can be seen in more recent theorizing of leadership in climate. For example, Kopelman et al (1990) and Ostroff and Bowen (2000) argue that human resource practices give rise to climate. In work-group settings, human resource practices are executed in large part by group leaders (Zohar, 2000) and are used to organize and guide the work of group members. In turn, this imposed structure influences group member perceptions of group events and rewards. As an example, Smith-Jentsch, Salas, and Brannick (2001) document that team leadership as exercised through specific job assignment practices, such as delegation of opportunities to perform trained tasks (e.g., Ford, Quinones & Sego, & Sorra, 1992), lead to a transfer of training climate. An additional argument presented by Schneider and Reichers (1983) posits an interactionist perspective, stating that "climates emerge out of the interactions that members of the work group have with each other" (p. 30). Through these interactions, shared meaning and perceptions of objective aspects of the work environment emerge. Drawing on social comparison and social conformity theory, Ashforth (1986) argues
43
greater conformity of perceptions and meaning converge on a particular "fulcrum" when a "compelling referent" exists in the work group. A leader acts as a "compelling referent" in the work group. Their formal status in the organization provides them a heightened degree of saliency and value within the group. Drawing from the safety climate literature, Hofmann et al (1995) echo Ashforth's (1986) assertions by arguing that workers motivation and attitudes mirror the leaders' priorities on safety, thereby impacting the efficacy of safety climate. Leader-group communications have been shown to impact how members perceive the openness of the climate, and in turn, shape their attributions of work events and willingness to admit mistakes (Edmondson, 1996; Hofmann & Stetzer, 1998). In addition, Scott and Bruce (1994) find that leader expectations affect a climate for innovation, and in turn, innovative behavior. Lastly, Barling et al (2002) show that transformational leaders influence climate through expressing idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration. These aspects of leadership provide opportunities for interaction between leaders and their group members. These interactions convey the leader priorities, thereby shaping climate perceptions. In sum, building from Payne and Pugh (1976) and Schneider and Reichers' (1983) structural arguments, it has been argued that, through communication of their priorities, leaders provide the necessary structure to shape group members perceptions of group routines and rewards. A second complementary argument adopted from Schneider and Reichers (1983) and Ashforth (1986) asserts that specific forms of climate emerge through interactions among group members and their leaders. During these interactions, leaders represent a "compelling referent," driving conformity of perceptions to converge
44
on their central priorities. Moreover, more recent empirical evidence substantiates the link between leadership and climate. For these reasons, I offer the following general proposition, accompanied by more specific hypotheses: Proposition 1: The leader's priority directly impacts work-group climate. Hypothesis 1a: The leader's priority on employee development is positively related to a climate for learning Hypothesis 1b: The leader's priority on goal accomplishment is positively related to a climate for performance Hypothesis 1c: The leader's priority on avoiding failure is positively related to a climate for avoiding failure. Relation between work-group climate and group-member state goal orientation. In part, the relationship between work-group climate and group-member state goal orientation hinges on the question of whether one's perceptions of the environment can affect his/her motivation. The broader motivational literature demonstrates a relationship between individual perceptions and one's motivation. For example, equity theory
(Adams, 1965) posits that motivation is a function of how an individual perceives his/her ratio of efforts to rewards as compared to that of another party. These perceptions of oneself in comparison to another drive individual motivation. Similarly, expectancy theory predicts effort will be a function of an individual's perceptions regarding the utility of effort, value of the reward, and probability an achieved outcome will result in receipt of reward (Vroom, 1964). Lastly, Bandura's (1997) triadic reciprocal causation model stipulates that perceptions of factors relating to the environment, individual and his/her behavior continuously interact to affect one another. Moreover, individual
45
factors, such as motivational constructs (e.g., self-efficacy), are products and drivers of the perceived environment and one's behavior. Consistent with this logic, the reasoning underlying the model proposed herein suggests that individual perceptions of their work environments, i.e., work-group climate and their relationship with the leader, affect one's motivation, and in turn, their behavior. Consistent with the broader literature on motivation, research support the logic that climate impacts individuals motivationally, and in turn, their performance outcomes (Griffin & Neal, 2000; Kopelman et al, 1990; Neal et al, 2000). Kopelman et al (1990) posit that climate affect motivation by helping to clarify outcome expectancies through rewards, influencing individuals' self-efficacy through task support, and affecting the valences of particular behaviors and outcomes through rewards and group structures. Empirically, Griffin and Neal (2000) and Neal et al (2000) demonstrate that climate influences individual motivation levels, and in turn, performance. However, as Bandura (1997) points out, person-factors, such as personality, affect causal relationships between environmental factors, motivational states, and behavioral outcomes. Consequently, the relationship of work-group climate to state goal orientation also depends on the extent to which this contextual factor is compelling enough to override personal preferences. Mischel (1977) theorized that situational strength impacts the extent to which personality factors are dominated by contextual influences. He describes strong situations as ones that lead situation participants to hold uniform perceptions of events and expectations of appropriate behaviors and receive rewards for behaving consistent with expectations. On the other hand, weak situations are not
sufficiently compelling to cause situation incumbents to share perceptions regarding
46
events and expected behaviors. Mischel's theory (1977) predicts that in strong situations, individuals are more likely to behave consistent with situational demands. Conversely, in weak situations, individual behavior will be more reflective of individual difference factors. Like other strong situations, a work-group climate provides clear situational cues as to what behaviors are expected, valued, and rewarded. However, why might
individuals conform to the demands of a strong situation, such as work-group climate? Schneider (1975) provides two theoretical arguments that explain why individuals align their attitudes, motivations, and behavior with the existing climate. First, Gestalt
Psychologists posit that the duty of every individual is to understand the order that objectively exists, as represented by situational cues, and to behave consistently with this apprehended order. In this way, Gestalt Psychologists assert that individuals are driven to make sense of their patterned context and to act in accordance with the demands of this environment. Consequently, acting consistent with the demands of the climate, fulfill
individuals' need for order and understanding of their environment. Secondly, Schneider (1975) describes the Functionalist school of thought as an additional rationale for why individuals conform to the existing climate. Similar to Gestalt Psychology, Functionalists assert that individuals perceive individual environmental cues as representative of a larger, patterned order of their environment; however, Functionalists believe that individuals apprehend this order in order to effectively adapt to their environment, not because they are driven to without choice. Moreover, this school of thought asserts that people behave consistently with this perceived order to achieve homeostatis and harmony with their environments. This logic
47
implies that individuals would be greatly challenged to resist conforming to the climate they perceive (Schneider, 1975). Taken together, Gestalt Psychology and the
Functionalist tradition extends the suggestion that work-group climate provides a strong situation to provide rationales for why individuals are driven to adapt their perceptions, motivations, and behavior in accordance with the expectations conveyed by the climate. Recent research and theory supports these theoretical assertions that workgroup/sub-unit climate affect individual outcomes (Joyce & Slocum, 1984; Hofmann, Morgeson, & Gerras, in press; Hofmann & Stetzer, 1996; Hofmann & Stetzer, 1998; Zohar, 2000). In their investigation of three plants within a duty truck manufacturer, Joyce and Slocum (1984) found that unique plant climates were differentially related to individual performance and job satisfaction. Hofmann and Stetzer (1996, 1998) found that a climate for safety shaped attributions for the causes of workplace accidents and impacted the frequency of unsafe behaviors and accidents. Similarly, Zohar (2000) found that a work-group climate that encouraged and valued safe behaviors led to a decrease in injuries on the job. Lastly, Hofmann et al (in press) found that together with leadership, work-group safety climate impacted individual perceptions of their organizational role, and in turn, affected safe behaviors. Taken together, these results suggest that workgroup climate is a strong situation that influences individual-level phenomena. While research has not explored the specific question of whether work-group climate affects individual state goal orientation, theory and empirical evidence regarding goal orientation suggests the existence of such a relationship. Dweck and Leggett (1986) assert that when strong situational cues are present, as in the case of work-group climate, individual dispositional goal orientations may be dominated by a state goal orientation.
48
Nicholls (1984) provides additional clarification regarding the specific conditions that impact state goal orientation. According to Nicholls (1984), individuals are more
inclined to hold a mastery or learning orientation if there is a moderate amount of task challenge and task-related extrinsic rewards are not salient. Because in a climate for learning, growth and development are emphasized, completion of easy tasks is not valued. Consequently, in this climate, on average, employees tend to be faced with moderate to high task challenge. In addition, intrinsic satisfaction, gained through
professional growth, is emphasized. These points suggest that individuals will adopt a state learning orientation when working in a climate for learning. Nicholls (1984) asserts that certain conditions encourage individuals to adopt a performance orientation: (1) when individuals are cognizant that their performance on a task is being evaluated, (2) when competition among group members is fostered, and (3) when task performance is highly visible. The saliency of task evaluation and level of competition are particularly relevant here. In both a climate for performance and avoiding failure, employees receive subtle and explicit messages that their performance is being evaluated. In the climate for performance, employee performance is consistently gauged against set standards. Performance and achievement are emphasized consistently, heightening the saliency of performance evaluation. In addition, task and socio-emotional support are around
attaining performance goals, reinforcing the notion that performance is being monitored and evaluated continuously. In a climate for avoiding failure, the consistent evaluation of performance is subtler. Employees are expected to avoid committing and admitting mistakes in order to
49
appear competent. Consequently, group members are consistently monitoring potential errors that may need to be covered up, thereby heightening the saliency of sub-par performance. Moreover, in these types of climates, the focus is on avoiding negative competence assessments. Consequently, attention is focused almost exclusively on the evaluation of performance rather than on possible task strategies, such as experimentation and learning from mistakes. Typically, organizations have a fixed amount of resources, including rewards. For this reason, emphasis on extrinsic rewards fosters competition among team members. However, in a climate for learning, this "fixed pie" perception is offset by the emphasis on intrinsic rewards such as intrinsic satisfaction for improving and preserving in the face of challenge. In climates for performance and avoiding failure, intrinsic rewards are not salient; therefore, employees more actively compete for extrinsic rewards. Consequently, the perception of a fixed quantity of rewards breeds competition within these climates, encouraging the emergence of state performance orientations. Lastly, researchers have consistently shown that individuals adopt state goal orientations consistent with situational demands. As reviewed earlier, Butler (1993), Elliot and Harackiewicz (1996), Harackiewicz and Elliot (1993), Mangos & SteeleJohnson (2000), Steele-Johnson et al (2000), and Stevens and Gist (1997) induced state goal orientations among respondents that were consistent with the experimental task instructions. Studies conducted with samples of children demonstrate a similar pattern— participants conform their state goal orientation to situational demands (Ames & Archer, 1988; Elliott & Dweck, 1988).
50
Based on motivational theory, climate research and theory, arguments for climate as a strong situation, goal orientation theory and related empirical evidence, I propose the following proposition and hypotheses: Proposition 2: Work-group climate is directly related to team members' state goal orientation. Hypothesis 2a: A climate for learning is positively related to team members' state learning orientation. Hypothesis 2b: A climate for performance is positively related to team members' state prove performance orientation. Hypothesis 2c: A climate for avoiding failure is positively related to team members' state avoid performance orientation. Relation between state goal orientation and outcomes. Consistent with social cognitive theory (Bandura, 1986), which stipulates that learning and behavior is motivated and regulated by one's cognitions, the theory underlying goal orientation centers on a socio-cognitive approach to motivation. Early goal orientation scholars theorized that one's goal orientation, derived from their implicit beliefs of ability, created cognitive frameworks that shape their interpretations of achievement settings (Dweck, 1986; Dweck & Leggett, 1988). More specifically, two sets of underlying cognitions drive behavioral differences: one set center on interpretations of the task while the other set deal with cognitive responses during task performance. Research shows that when state goal orientation is induced, these cognitions emerge in parallel form to their trait goal orientation counterpart (Ames & Archer, 1988; Elliott & Dweck, 1988; Mangos & Steele-Johnson, 2001). For example, Elliott and
51
Dweck (1988) found that children given specific achievement goals (i.e., learning, performance) mirrored the same cognitive, affective, and performance patterns as children relying on their own trait goal orientation. Ames and Archer (1984) found similar results: perceptions of task challenge and causes of success and failure shaped by the type of achievement goal the children's classroom promoted paralleled their respective trait goal orientation cognitions. Lastly, consistent with trait explanations, Mangos and Steele-Johnson (2001) found that a state goal orientation was related to perceptions of task complexity. Because similar sets of cognitions accompany state goal orientations as their trait counterparts and these cognitions are reasoned to drive behavioral differences, I rely on trait and state goal orientation evidence in theorizing about the relationships between state goal orientation and outcomes. For the sake of clarity, however, I note those studies that employed a state approach. Goal orientation theory suggests that learning-oriented individuals view achievement settings as opportunities to develop their competence (Dweck, 1986; Dweck & Leggett, 1988). Toward this end, individuals holding a learning orientation will devote the necessary effort to persevere in the face of failure to achieve higher levels of performance. Empirical evidence supports this theoretical claim and has shown that learning orientation predicts higher levels of task performance (Brett & VandeWalle, 1999; Butler, 1993; Chen et al 2000; Ford et al, 1998; Phillips & Gully, 1997). Brett and VandeWalle (1999) demonstrated that MBA students with a mastery orientation attained higher ratings for the quality of their in-class presentations. In adopting a state approach, Butler (1993) found that individuals subjected to a mastery orientation experimental condition asked more task-related questions and, in turn, received higher scores on a
52
problem-solving task. Ford et al (1998) demonstrated that individuals with a learning orientation made more correct decisions after receiving training on a PC-based, dynamic decision-making simulation program. In addition, Chen et al (2000), Phillips and Gully (1997), and VandeWalle et al (1999) found consistent results indicating a positive, indirect relationship between learning orientation and graded exam performance. Dweck and her colleagues (1986; 1988; 2000) further suggest that learningoriented individuals will employ effective learning strategies to self-regulate their responses to challenging settings. Organizational research provides convincing evidence substantiating this link. VandeWalle and Cummings (1997) document that learningoriented individuals perceive higher value in feedback, and as a result, engage in more feedback seeking behavior. Brett and VandeWalle (1999) found that individuals holding a learning orientation were more likely to set learning goals. Ford et al (1998) found that trainees with a learning orientation monitored their own thought processes so as to pinpoint ways to improve and maintain active mental engagement in the training. Fisher and Ford (1998) found that the effect of mastery orientation is positively related to effort and negatively related to off-task attention. Research on state goal orientation offers similar findings, further substantiating a link between state learning goal orientation and use of learning strategies. Stevens and Gist (1997) found that individuals in a mastery condition engaged in more skillmaintenance activities after receiving negotiation training. Kozlowski et al (2001) and Schmidt, Chambers, Kozlowski & DeShon (2001) found that individuals with a learning state goal orientation engaged in sophisticated cognitive strategies, such as organizing their knowledge into consistent patterns and meta-cognitive activities. Lastly, Ames &
53
Archer (1988) found that students engaged in higher levels of learning strategies, such as self-planning and monitoring, when operating in a state mastery condition. Together, these studies provide a sound empirical justification to predict that state learning orientation is positively related to use of learning strategies. Because a state and trait learning orientation serves to regulate cognitive functioning by focusing individual attention on the task, rather than on preserving one's ego (e.g., Dweck, 1986; Nicholls, 1984), it is unlikely that learning-oriented individuals would engage in defensive behaviors. By their very nature, defensive behaviors require individuals to shift their attention from the task to preserving their appearance of being competent. In this way, these behaviors would require increased off-task attention, a condition that contradicts Fisher and Ford's (1998) finding that mastery orientation is negatively related to off-task attention. While limited empirical evidence exists, theory and logic make it reasonable to expect a negative relationship between state learning goal orientation and defensive behaviors. Theory suggests that individuals with a performance orientation interpret challenging situations as potential threats to their confidence (Dweck, 1986; Dweck & Leggett, 1988. As a result, these individual withdraw from the task and experience significant declines in performance in the face of obstacles (Dweck, 1986, 2000; Dweck & Leggett, 1988). Consistent with this assertion, empirical research has shown a
negative relationship between individuals holding an avoid orientation and task performance. Elliot and Church (1997) and Elliot and McGregor (1999) found that individuals who held avoid performance orientations demonstrated lower levels of performance on exams. VandeWalle et al (1999) found similar results: students high on
54
avoid performance orientation exhibited lower levels of effort and self-efficacy, which in turn, produced lower levels of performance on exams. Elliot and Harackiewicz (1996) showed that a state approach-avoidance, i.e., state avoid orientation, was negatively related to task involvement. While they did not measure task performance, it seems likely that if individuals are disengaged in a task, their task performance would suffer as a result. Theory and evidence suggests that an avoid orientation is negatively related to task performance. Because performance-oriented individuals believe that ability cannot be augmented or developed, theory predicts that they would not engage in strategies to learn new competencies (Dweck, 1986, 2000; Dweck & Leggett, 1988). Holding true to this assertion, research documents that performance orientation was not related to metacognition (Ford et al, 1998), rehearsal of new skills, organization of new material, using examples to better learn (Fisher & Ford, 1998), nor practice level (Brown, 2001). Adopting a state approach, researchers have found similar results: state performance orientation is unrelated to learning strategies (Ames & Archer, 1988) and organizing coherent knowledge structures (Kozlowski et al, 2001). It is important to note that these studies did not tease apart the effects of avoid and prove goal orientation. A few studies provide more fine-grained insights into how avoid and prove orientation might be related to use of learning strategies. VandeWalle and Cummings (1997) found that avoidant individuals perceive higher cost associated with feedback and engaged in less feedback seeking behavior. However, prove orientation was unrelated to perceive cost of feedback and feedback seeking behavior. Elliot and Harackiewicz's (1996) finding that state avoid orientation is negatively related to task involvement
55
suggests the potential for decreased use of learning strategies, particularly given that other researchers have shown a significant relationship between effort and involvement and learning strategies (Fisher & Ford, 1998). These findings do not provide convincing evidence of a relationship between prove orientation and learning strategies; therefore, no formal hypothesis is offered. However, this research does suggest a negative relationship between avoid orientation and learning strategies. Additional research and logic further justifies a hypothesized, negative relationship between avoid orientation and learning strategies. Learning from errors and experimentation have been shown to be effective learning strategies for organizations as well as individuals (e.g., McCall, Lombardo & Morrison, 1988; Sitkin, 1992). However, engaging in these types of behaviors is perceived as particularly threatening to those with an avoid goal orientation. To them, making errors, asking for assistance, and/or
experimenting only invites others to perceive them as incompetent, an outcome they are actively motivated to avoid. Moreover, it seems reasonable to expect that these
individuals will engage in more defensive behaviors when errors do occur (e.g., excuse making, blaming others) in order to preserve the façade of being competent. Taken together, this research and logic suggests that avoid orientation and learning strategies are negatively related, and avoid orientation and defensive behaviors are positively related. Dweck and Leggett (1988) and Heyman and Dweck (1992) note that an emphasis on performance goals may be adaptive in certain circumstances and have a positive effect on task performance. Dweck and Leggett assert: "it is often important for individuals to evaluate their abilities or to gain positive judgments of their competence" (1988, 260). The emphasis on gaining positive competence judgments is consistent with a proving
56
orientation, and research has shown that a prove performance orientation is positively related to task performance. Brett and VandeWalle's (1999) study of MBA students demonstrated that individuals with a proving orientation adopted goals to refine their skills and to perform better than others to positively impact the quality of their presentations. Elliot and McGregor (1999) and Elliot and Church (1997) found a direct, positive relationship between prove performance orientation and exam performance, while VandeWalle et al (1999) found an indirect, positive relationship between prove orientation and exam performance. In sum, guided by extant theory and research, I propose the following proposition and hypotheses: Proposition 3: individual outcomes. Hypothesis 3a: Group members' state learning goal orientation will be positively related to task performance and use of learning strategies and negatively related to defensive behaviors. Hypothesis 3b: Group members' state prove performance goal orientation will be positively related to task performance. Hypothesis 3c: Group members' state avoid goal performance orientation will be negatively related to task performance and use of learning strategies and positively related to defensive behaviors. Role of the quality of leader-member exchange relationship. Leader-member exchange (LMX) describes a relational leadership process. Because leaders develop different types of relationships with each of their members (Graen & Cashman, 1975), Group members' state goal orientation is directly related to
57
work group members experience different levels of access to the leader and varying degrees of a sense of obligation to meet leader expectations (Dienesch & Liden, 1986; Graen & Cashman, 1975; Graen & Uhl-Bien, 1995; Liden et al, 1997). These two mechanisms, level of access and sense of obligation, characterize the quality of the leader-member exchange relationship. It is argued that leader-member relationships
provide a conduit for transmission, receipt, and internalization of the messages being conveyed through a particular climate. Together, work-group climate and the quality of leader-member exchange relationship are expected to affect state goal orientation. The nature of this interaction is detailed below. Because leaders create unique relationships with each of his/her employees, some members enjoy greater levels of access, communication, and interaction with their leader, while others interact with the leader strictly on the basis of their formal employment agreement (Dienesch & Liden, 1986; Graen & Cashman, 1975; Graen & Uhl-Bien, 1995; Liden et al, 1997). Members who perceive themselves as having a better quality
exchange relationship with their leader have an advantage in deciphering expectations of their role and performance as communicated by the work-group climate over members who do not have this same quality of relationship. This advantage has been documented as positively influencing performance (cf., Liden et al, 1997), reducing role conflict, and improving role clarity (Gerstner & Day, 1997). Research demonstrates that leaders and members with high quality LMX relationships are more likely to be in perceptual agreement regarding work issues and messages conveyed by the climate (Graen & Schiemann, 1978; Kozlowski & Doherty, 1989). Holding uniform expectations and perceptions is one crucial precondition to
58
situational strength (Mischel, 1977). Consequently, sharing similar perceptions with one's leader augments the existing situational strength provided by the work-group climate and supplies an additional vehicle through which one can further clarify expectations and heighten one's sense of obligation to conform to these expectations. In this case, group members with high quality LMX relationships are more likely to perceive an even stronger situation than those with low quality LMX relationships. In turn, the perceived augmented strength of the situation for members enjoying highquality LMX is expected produce greater clarity regarding which form of state goal orientation is favored and greater pressure to conform to the ascribed motivational state. This sense of obligation arises out of the convention of reciprocity established by the leader-member exchange relationship (Dienesch & Liden, 1986; Graen & Cashman, 1975; Liden et al, 1997; Sparrowe & Liden, 1997). Built on the notion from social exchange theory (Blau, 1964) that social exchanges between two parties "engender feelings of obligation, gratitude, and trust" (p. 94), LMX researchers have argued that members with higher quality relationships with their leaders are given greater amounts of resources, and in return, offer higher levels of loyalty, behavioral consistency with expectations, effort, performance, and positive work-related attitudes (Gerstner & Day, 1997; Graen & Cashman, 1975; Graen, Novak & Sommerkamp, 1982; Liden et al, 1997; Scandura & Schriesheim, 1994; Scott & Bruce, 1994; Tierney et al, 1999). Employing this logic, Hofmann et al (in press) argue that climate cues members as to which behaviors are emphasized and valued, and higher levels of LMX create a heightened sense of obligation to redefine their roles consistent with these cues. Their empirical findings support this theoretical contention that high-quality LMX relationships produce
59
greater levels of conformance with the expectations conveyed through the work-group climate as compared to lower-quality LMX. The work by Hofmann et al (in press) supports the logic endorsed here. It has been argued that leader-member exchange relationships provide a channel for further transmission, clarification, and internalization of cues conveyed by the work-group climate. In this way, the work-group climate signals the preferred motivational
orientation while the leader-member exchange relationship provide a means for reinforcing climate cues, clarifying expectations conveyed through the climate, and engendering a sense of obligation to adopt the state goal orientation ascribed by the climate. This logic suggests the following proposition and hypotheses: Proposition 4: Work-group climate and group member perceptions of their
leader-member exchange relationship interact to affect member state goal orientation. Hypothesis 4a: A climate for learning is more strongly related to state learning orientation when members perceive a higher-quality LMX relationship than when members perceive LMX to be of low quality. Hypothesis 4b: A climate for performance is more strongly related to state proving orientation when members perceive a higher-quality LMX relationship than when members perceive LMX to be of low quality. Hypothesis 4c: A climate for avoiding failure is more strongly related to state avoid orientation when members perceive a higher-quality LMX relationship than when members perceive LMX to be of low quality. To summarize, it is argued that leader priorities shape three prototypical workgroup climates: a climate for learning, a climate for performance, and a climate for
60
avoiding failure (Proposition 1). It is predicted that group members adopt a state goal orientation consistent with the types of behavior that are perceived to be rewarded, supported and expected, as communicated by the work-group climate (Proposition 2). Further, the type of state goal orientation will be differentially related to individual task performance, use of learning strategies, and defensive behaviors (Proposition 3). Finally, the effects of leader-member exchange are predicted to interact with those of work-group climate to further explain group members' adoption of the state goal orientation (Proposition 4).
61
CHAPTER 3 Research Methods To test the proposed relationships, a multi-phase, multi-method cross-sectional empirical study was conducted. The first phase utilized a qualitative approach to gather data to refine construct measurement and develop a measure of state goal orientation. The second phase involved development and measurement testing of the newly-created state goal orientation measure and is referred to here as the measurement study. The final stage involved surveying respondents via questionnaire in a field setting to test the hypothesized relationships. Each of these phases is discussed in more detail below. Because the findings from the first two phases informed the final stage of research, i.e., the field study, findings from the qualitative and measurement studies are reported here. Research Design: Phase I—Qualitative Stage The purpose of this stage was to gather data to (1) tailor existing measures to be used in the field study to the particular research setting (2) augment and refine the leadership and climate measures, and (3) develop a context-appropriate, valid state goal orientation measure. Sample and design. Six one-on-one interviews were conducted with managers from the regional retail bank that opted to participate in the final phase of this dissertation. This particular bank has a unique reporting structure —financial center managers (FCM) are responsible for overseeing the performance of the branch and directly supervise the personal banking staff (e.g., relationship bankers, loan officers). The area operations managers (AOM) manage the operational aspects of 3-5 different branches, have their office in only one of these branches, and directly supervise the tellers
62
and other administrative staff within those branches. Consequently, I conducted three interviews with FCMs and the other three with AOMs. I worked with senior managers to identify one FCM and one AOM that best represented each of the leadership styles of interest (i.e., places a priority on employee development, goal accomplishment, avoiding failure). Each interview lasted approximately 60-90 minutes and were used primarily to identify the most relevant performance measures for retail bank employees and a range of leader behaviors that might be appropriate for inclusion in the leader priority measures (see Appendix A for the interviewing protocol). In addition, six three-hour focus groups with 4-6 retail bank employees were conducted. Focus group participants were selected by senior managers to represent a diversity of job functions and company tenure. A completely crossed design was used with two primary experimental factors: (1) type of leader that focus group participants report to (e.g., learning oriented, performance-oriented, sensitive to failure) and (2) whether the AOM was co-located with the participant. This approach enabled me to determine the extent to which the physical proximity of the AOM influenced the core processes investigated here (it was found to have a minimal effect). The agenda for the focus groups included gathering data on climate dimensions and sample task situations that required unique motivational orientations, which ultimately provided the basis for the measure of state goal orientation (see Appendix B for the focus group protocol). Key findings. Several important findings emerged in the qualitative portion of this dissertation that guided subsequent data collection and analysis. First, there was high agreement among managers and employees that the Financial Center Manager/Leader is
63
responsible for setting the tone and priorities within the branch.2 The Area Operation Manager was often described as having a supporting role in this regard. This finding is important because it identified the "leader" in this particular research context. Second, focus group data on branch climate revealed that a strong, consistent infrastructure exists in this particular organization that standardizes training and rewards. Focus group
participants did still note that there was variation in the extent to which branch leaders make opportunities and rewards available to their employees. With respect to measure development, data collected during the interviews with the managers indicated that tellers and administrative personnel are assessed on different performance measures than members of the personal banking staff. As a result, using a broad based measure of individual-level performance seemed appropriate. Focus group data on the behaviors that leaders use to communicate their priorities validated Schein's (1992) work and suggested that what leaders pay attention to and measure and the type and focus of their coaching are particularly important in conveying implicit priorities in this particular research context. Lastly, output from the focus group contained a variety of task situations that employees in this organization routinely face. Three task situations were identified as highly relevant to all participants who would be involved in the final study of this dissertation, regardless of job class and branch—generating new business, managing difficult, irate clients, and participating in meetings. In addition, focus group respondents indicated that bank employees adopt a wide range of motivational orientations in these three task situations.
2
Financial Center Leaders differ from Financial Center Managers in that they supervise the operations of
more than one branch.
64
Research Design: Phase II—Measurement Study Sample. A total of 273 part-time MBA students participated in this study
(average response rate: 95.39%). Sixty-four percent of participants are White, 16% are Asian/Pacific Islander, 5% are Latino or Hispanic, 4% are African American, and 4% are of other ethnic backgrounds. Sixty-three percent of this sample is male, and on average, participants are 28 years old (SD=3.51 years). Participants have a wide variety of work experience (overall work experience: mean=5.49 years; SD=3.38) in various industries. Approximately, half of participants do not currently supervise others, 23% of respondents are supervisors, 14% are middle managers, and 3% are executives. Those who manage others have, on average, 1.94 years of experience (SD=2.61). Design. The design of the measurement study was a 3X3 between- and withinsubjects design. The between-subject factor was the achievement orientation of the work-group context (i.e., leader and climate) and was communicated to participants through a description of a hypothetical leader and work group (see Appendix C for experimental conditions). Repeated measures of the state goal orientation measure were captured by asking respondents to assess their motivational orientation under three different task situations—generating new business, handling difficult clients, and participating in meetings (i.e., within-subject factor). A situation-based approach to measure state goal orientation was employed to (1) allow me to test the potential effects of objective task demands on state goal orientation, (2) minimize measurement error by providing respondents the same set of task conditions, and (3) increase reliability by capturing repeated assessments.
65
Data Collection Procedures. Data were collected at two points in time. First, during their core HR course, 260 MBA students were asked to sign a consent form and complete a survey that collected various background (i.e., gender, ethnic background, age, managerial experience, total work experience) and personality measures (i.e., trait goal orientation, implicit theory of ability, core self evaluations, self monitoring). Of these 260 students, 253 provided data for a response rate of 97.24%. Approximately two weeks later, these same students were randomly assigned to one of three different experimental conditions and were asked about their state goal orientation in three task situations and the likelihood they would engage in a variety of behaviors in these situations (e.g., feedback seeking, self promotion, and avoiding blame). In addition, it
should be noted that the ordering of the task situations were randomized so as to remove any possible priming effects. A total of 262 students were asked to participate in the Time 2 data collection, and 244 provided data (response rate: 93.53%). Matching
responses from Time 1 and Time 2 yielded a total matched sample of 227. Surveys administered at Time 1 and 2 appear as Appendices D and E, respectively. To ensure that participants understood the experimental condition in which they participated, responses to the following three items were captured on a 5-point scale ("I responded to questions asked on this survey as if I worked for a boss who emphasizes?(a) the importance of continuous learning, (b) the importance of proving one's ability to others inside and outside the branch, and (c) the avoidance of committing and admitting mistakes"). One-way analysis of variance conducted on these
manipulation checks revealed significant differences across the three experimental conditions (learn check F=136.42, p
Reports Study on Understanding State Goal Orientation: Leadership and Work-group Climate as Key Antecedents :- Goal orientation is a "disposition toward developing or demonstrating ability in achievement situations". Previous research has examined goal orientation as a motivation variable useful for recruitment, climate and culture, performance appraisal, and selection.
REPORTS STUDY ON UNDERSTANDING STATE GOAL ORIENTATION: LEADERSHIP AND WORKGROUP CLIMATE AS KEY ANTECEDENTS
ABSTRACT This research attends to a broad range of practically significant employee achievement goals and provides insight into how to enhance individual-level performance by examining the antecedents to individual-level state goal orientation in organizational work groups. State goal orientation is defined here as a temporary achievement goal, and it is theorized that leadership and work group climate processes parallel each dimension of state goal orientation to cue and ultimately induce the corresponding achievement goal among individual work group members. The leader's achievement priority is argued to drive the formation of work group climate consistent with this priority. The resulting work group climate signals and compels group members to adopt the ascribed form of state goal orientation. The quality of the leader-member exchange (LMX) relationship is viewed as a means to internalize cues from the work group climate in the emergence of state goal orientation. Results from experimental and field studies provide evidence that (1) leadership and climate perceptions are related to their parallel form of state goal orientation, (2) the relation between individual perceptions of a climate for learning and state learning goal orientation is stronger when group members enjoy higher quality
exchange relationships with their leader, and (3) state goal orientation may be validly and reliably assessed using the measure created especially for this research.
TABLE OF CONTENTS List of Tables????????????????????????????... vii List of Figures????????????????????????????.. viii Chapter1: PROBLEM STATEMENT?????????????..??????.1 Chapter 2: MODEL DEVELOPMENT AND HYPOTHESES???...??????10 Model Overview????????????????????????.. 10 Key Construct Definitions????????????????????... 13 Leader Priority??????????????????????. 13 Work-group Climate...???????????????????. 22 Goal Orientation?????????????????????.. 32 Individual Outcomes????????????????????37 Leadermember exchange relationship????????????? 39 Hypothesized Relationships????????????????????.41 Relation between leadership and work-group climate???????.42 Relation between work-group climate and state goal orientation???45 Relation between state goal orientation and outcomes??????? 51 Moderating role of the quality of LMX relationship???????... 57 Chapter 3: RESEARCH METHODS???????????..???????... 62 Research Design: Phase I—Qualitative Stage???????????...... 62 Sample and design????????????????????... 62 Key findings??????????????????????? 63 Research Design: Phase II—Measurement Study??????????? 65 Sample?????????????????????????. 65 Design?????????????????????????. 65 Data collection procedures?????????????????.. 66 Measures????????????????????????.. 67 Control and theoretically-relevant measures???????.. 67 Dependent measures????????????????? 68 State goal orientation????????????????... 68 Development of measure????????????. 68 Method for gathering validity evidence??????.. 69 Convergent and discriminant validity evidence???.. 72 Summary of key findings???????????? 82 Research Design: Phase III—Field Study?????????????? 83 Sample?????????????????????????. 83 Design and data collection procedures????????????? 84 Measures???????????????????????..... 85 Analytic procedures???????????????????? 91
v
Issues of control?????????????????????.. 92 Chapter 4: RESULTS FROM THE FIELD STUDY?????????????. 94 Psychometric evidence of the state goal orientation measure??????.. 94 Findings from hypothesis testing?????????????????. 96 Findings from tests for mediation?????????????????. 106 Summary??????????????????????????... 108 Chapter 5: DISCUSSION???????????????????????... 110 Study limitations???????????????????????.. 120 Research implications?????????????????????.. 122 Managerial implications????????????????????... 124 Conclusion?????????????????????????? 125 APPENDICES????????????????????????????. 126 A: Interview Protocol?????????????????????...126 B: Focus Group Protocol????????????????????. 129 C: Experimental Conditions??????????????????? 131 D: Time 1 Measurement Study Items???????????????.. 134 E: Time 2 Measurement Study Items???????????????.. 136 F: Field Sample Descriptive Statistics by Branch??????????...140 G: Time 1 Field Study Items???????????????????149 H: Time 2 Field Study Items (collected from employees)???????...151 I: Time 2 Field Study Items (collected from managers)???????? 155 REFERENCES?????????????????????????..??.. 160
vi
List of Tables Theoretical Development 1. Behavioral Indications of Leader Priority???????????????.21 2. Description of Three Prototypical Work-group Climates??????..??.. 32 3. Construct Definitions???????????????????.???... 41 Measurement Study 4. Final Results from Confirmatory Factor Analysis: Factor Loading?????. 72 5. Multi-trait, Multi-Method Results: Model Fit Statistics?????????.. 75 6. Variance Components due to Trait, Method, and Error for Model 1????... 75 7. Measurement Study Variable Means, SDs, and Intercorrelations?????? 79 8. Mean Differences in State Goal Orientation Across Experimental Conditions... 81 9. Hierarchical Regression Results: State Goal Orientation Predicting Outcomes. 82 Field Study 10. Analytic Strategy Used for Testing Hypotheses????????????.92 11. Final Results from Confirmatory Factor Analysis: Factor Loadings????. 95 12. Branch-level Main Study Variable Means, SDs, and Intercorrelations???..96 13. Individual-level Main Study Variable Means, SDs, and Intercorrelations??.97 14. Random Coefficient Modeling Results: Leader Priority Predicting Psychological Climate???????????. 100 15. Random Coefficient Modeling Results: Psychological Climate Predicting State Goal Orientation????????. 102 16. Random Coefficient Modeling Results: State Goal Orientation Predicting Outcomes?????????????..103 17. Random Coefficient Modeling Results: Interactive Effect of Psychological Climate and LMX in Predicting State Goal Orientation??????????????????????.105
vii
List of Figures 1. Proposed Model of Contextual Influences on Individual State Goal Orientation and Associated Outcomes???????????12 2. Data Sources and sequencing???????????????????85 3. Interactive Effect of Psychological Learning Climate and LMX on State Goal Orientation?????????????????????...106
viii
CHAPTER 1 Problem Statement To remain competitive in an environment characterized by continuous change, high levels of complexity, and tremendous opportunity, contemporary organizations need to remain adaptive, innovative, and offer exceptional high products and services to customers. To achieve this end, organizations must motivate their members to
continuously learn and adapt to changing job demands and discover novel and innovative methods for solving complex problems (Hall & Mirvis, 1996). At the same time,
however, organizational members need to maintain high levels of performance to maintain high levels of satisfaction among customers, their colleagues, and organizational leaders. Research has shown leaders are instrumental in encouraging the motivation to learn and perform at high levels (e.g., Locke & Latham, 1990; VanVelsor, McCauley, & Moxley, 1998). In addition to these more positive outcomes, leadership can deter
organizational members from owning up to mistakes, thereby sabotaging unique learning opportunities and promoting "face saving" types of behavior (e.g., Edmondson, 1996). Building on this work, this research explores how organizational leaders encourage unique motivational responses for learning, performance, and avoiding failure among their employees and the outcomes associated with these different motivational states. Academic research on goal orientation provides a useful means to understanding the influence of contextual factors, such as leadership, on individual motivation and outcomes. Goal orientation is defined as one's goal preference in achievement settings (Dweck, 1986) and refers to one's desire to develop, realize, or exhibit capability to perform a specific activity (Dweck & Leggett, 1988). Initial research identified two
1
broad classes of underlying goals that individuals pursue: mastery and performance (Dweck, 1986; Dweck & Leggett, 1988). Individuals displaying a mastery or learning orientation focus on building their competence and/or improving their abilities. Those favoring a performance orientation seek to demonstrate their competence by seeking favorable judgments or avoiding negative judgments regarding their capabilities. While often treated as a dispositional trait, theory and research suggests that situational demands may induce an orientation contrary to one's preference (e.g., Button, Mathieu, & Zajac 1996; Elliott & Dweck, 1988; Nicholls, 1984). Substantial empirical evidence suggests that even subtle differences in experimental instructions in research settings can encourage individuals to adopt an orientation consistent with situational cues (e.g., Butler, 1993; Elliot & Harackiewicz, 1996; Harackiewicz & Elliot, 1993; Kozlowski, Gully, Brown, Salas, Smith, & Nason, 2001; Mangos & Steele-Johnson, 2001; SteeleJohnson, Beauregard, Hoover, & Schmidt, 2001; Stevens & Gist, 1997). Goal orientation has been shown to be a potent predictor of individual outcomes. For example, researchers have documented its relation to other motivational factors such as self-efficacy and intrinsic motivation (e.g., Chen, Gully, Whiteman & Kilcullen, 2000; Elliot & Church, 1997), behavioral outcomes such as performance (e.g., Brett & VandeWalle, 1999; VandeWalle, Brown, Cron, & Slocum, 1999) and attitudinal and affective responses (e.g., Ames & Archer, 1988; Elliot & McGregor, 1999). In the context of learning and development, goal orientation has been shown to be related to the acquisition of knowledge and transference of skills learned in training to the work context (Brown, 2001; Fisher & Ford, 1998; Ford, Smith, Weissbin, Gully & Salas, 1998;
2
Kozlowski et al, 2001). This body of research demonstrates the profound impact of goal orientation on behavioral, motivational and attitudinal outcomes. The wide-reaching impact of the construct of goal orientation and evidence of its malleability sets the stage to pose an intriguing question: if it is possible to promote different types of goal orientations, can organizational leaders facilitate adoption of different forms of goal orientation among their employees, and in turn, affect their performance? And, if so, how do leaders impact this motivational process and its associated outcomes? This line of inquiry provides practitioners and scholars value by uncovering insights into how leaders can better motivate employees to learn and perform in order to achieve superior results. Pursuit of understanding this phenomenon is
evidenced by the long tradition in the organizational behavior field that dates back to the Hawthorne Studies (as cited in Wren, 1979), in which Elton Mayo posited that supervisor's attention to workers enhanced employee morale, and in turn, affected performance levels. More contemporary scholars have studied this question by exploring leadership impact on employee self-efficacy and goal setting and performance (e.g., Bandura, 1997; Eden, 1992; Locke & Latham, 1990). While organizational behavior scholars generally support the contention that studying leader impact on employee motivation and performance is essential, what value does examining leadership's relationship to one specific type of motivation, i.e., goal orientation, add to our understanding? Much of the existing research has focused on leadership's impact on motivating employees as a way to increase their productivity. Yet, with the emergence of the knowledge economy and the shift in the nature of work to becoming more complex, information- and skill-intensive, and fluid (Howard, 1995), the
3
importance of studying factors affecting employees' motivation to learn becomes vital. The changing organizational landscape and continuously shifting job requirements are redefining "employee effectiveness," demanding that researchers broaden their focus to include a motivation to learn, as well as more task-specific, productivity-related motivational concepts. Moreover, recent research suggests in some work group settings, contextual influences promote a motivation to "save face" among employees, dissuading employees from acknowledging their mistakes, continuously learning, adapting to new job requirements, and performing effectively (e.g., Edmondson, 1996; 1999). These organizational-, job- and research-related trends challenge researchers to examine a broader range of employee motivations. In this way, the concept of goal orientation provides a necessary vehicle to studying this variance in employee motivations. Some researchers have begun to explore how leadership and other contextual factors relate to employee goal orientation and performance (Ames & Archer, 1988; Potosky & Ramakrishna, 2002; Smith-Jentsch, Van Duyne, & Reynolds, 2001). Most directly, Smith-Jentsch et al (2001) investigated team leader goals as an inducement of team members' adoption of specific forms of goal orientation. Other studies have
explored broader contextual influences. For example, Ames & Archer (1988) explored how student perceptions of their classroom experiences affected their adoption of specific goal orientations. Potosky & Ramakrishna (2002) found that the work environment's emphasis on updating skills moderated the relationship between learning goal orientation and job performance. While these studies begin to document how contextual influences may impact employees' goal orientation, our understanding is limited by the scarcity of research in
4
this area, as well as the conceptual and methodological shortcomings of existing research. Conceptually, researchers have taken a myopic view of either leadership and/or context. Smith-Jentsch et al (2001) incorporate leader goals as the only contextual influence into their research. While leader goals have been shown to be important (Bass, 1985; Bennis & Nanus, 1985; Kotter, 1982), they represent one leadership process and confine our understanding only to the impact of leader goals rather than the larger influence of leadership on employee goal orientation. Methodologically, these studies have struggled to adequately measure goal orientations as induced by situational influences. Ames and Archer (1988) assess goal orientation by measuring classroom environment, thus confounding context with student motivation. The measure used by Smith-Jenstch et al (2001) includes thoughts and general assessments of the situation rather than a drive toward learning and/or performance. Lastly, Potosky and Ramakrishna's (2002) use of a data-analytic technique that does not properly accounts for the multi-level nature (i.e., non-independent) of their data raises questions regarding the validity of their results. The result of the paucity of research in this area and methodological and conceptual shortcomings of existing studies is that very little is known about how contextual influences, such as leadership, can facilitate the adoption of different forms of goal orientation among employees. In particular, this limited conceptualization of leadership influences is indicative of a larger shortcoming of the leadership literature. Graen and Uhl-Bien (1995) argue that most leadership research has employed a stove-pipe approach, choosing to focus on only one aspect of leadership process (e.g., leader, follower, or leader-follower relationship). More specifically, Graen and Uhl-Bien (1995) assert that incorporation of
5
variables representing leader-focused, follower-focused, and leader-follower relationshipbased perspectives will ensure against overlooking critical leadership aspects, thereby increasing predictive power of leadership models. In addition, a multi-perspective
approach allows scholars to better understand how each of these leadership domains complement one another in affecting outcomes (Graen & Uhl-Bien, 1995) and how relationships among these domains may vary depending on the setting and type of outcome examined (Vecchio, 1982). Despite these predictive advantages, few leadership studies have examined multiple leadership perspectives (Basu & Green, 1997; Howell & Hall-Merenda, 1999; Scott & Bruce, 1994; Tierney, Farmer, & Graen 1999 are notable exceptions). These studies begin to uncover the additive and interactive effects of leader characteristics and style and the quality of relationship between leaders and their followers in predicting outcomes. However, due to the limited number of studies taking this approach,
researchers continue to be challenged to better understand how leader-, follower-, and relationship-oriented processes complement one another in affecting outcomes. This research augments existing knowledge on leadership by examining multiple leadership perspectives and their impact on individual-level outcomes. Consistent with the average-leadership style approach (i.e., leaders affect followers equally), early leadership theory (e.g., Ohio State and University of Michigan studies), and the work of early climate researchers (e.g., Lewin, Lippitt, & White, 1939; McGregor, 1960), leader priorities are examined as antecedents to work-group climate and represent an important aspect of the leader-focused perspective. Work-group climate is defined here as
"incumbents' perceptions of events, practices, and procedures and the kinds of behaviors
6
that get rewarded, supported, and expected in a setting" (Schneider, 1990: 384). I argue that leaders, through their consistent emphasis on specific priorities, provide cues to employees regarding valued behaviors and these cues are meaningfully interpreted by work-group members and can impact their goal orientation and associated performance. Furthermore, this research incorporates the leader-follower relationship-based approach by examining how the quality of leader-employee exchange (LMX) relationships affects group-members' goal orientation and performance outcomes. Toward this end, contemporary leader-member exchange theory and research will be used to hypothesize the effects of this more interpersonal, one-to-one relational leadership process. I employ the concept of leader priority to capture the achievement focus of the leader (i.e., the "what") and present the concept of LMX as a relational medium through which leaders transmit and emphasize their priority (i.e., the "how"). Taken together, these two approaches to leadership processes are expected to yield a richer, more comprehensive view of the impact of leadership on employee motivation and performance. In addition, this research offers a further improvement over much of the existing leadership and motivation literature. One underlying assumption of leadership theory is that leaders influence follower motivation, which in turn, affects follower performance. While some researchers have directly measured each component of this causal chain (e.g., Durham, Knight & Locke, 1997), this practice has not been consistently employed. For example, Shamir, House and Arthur (1996) noted that charismatic leadership theory and research is limited in its ability to fully explain the motivational process by which charismatic leaders affect follower attitudes and performance. Further, empowering
7
leadership approaches assume that leaders build intrinsic motivation among followers to lead themselves. However, notable studies in this area simply examine the relationship of empowering leadership to perceptions of leadership effectiveness by group members, neglecting effects on group member motivation (e.g., Manz & Sims, 1987). This direct leadership—outcome approach is problematic because of its inability to rule out alternative explanations and more completely test underlying theoretical assumptions. It is plausible, for example, that higher employee performance results may be due to higher performers being attracted and selected to a particular type of leader, as posited by the attraction-selection-attrition framework (Schneider, 1987) instead of the leader's behaviors. Or, a certain type of leader may be more likely to provide
challenging assignment to employees. The resulting performance gains in this case would not necessarily be due leadership per se, but rather the motivational stretch inherent in these types of assignments (McCauley, Ruderman, Ohlott & Morrow, 1994). The current research extends current theorizing and testing by clearly delineating the causal chain that links leadership processes to employee performance outcomes. More specifically, this research examines how leadership affects the adoption of specific types of employee goal orientation, and in turn, impacts performance. In sum, the research question addressed here is: can organizational leaders facilitate adoption of different forms of goal orientation among their employees, and in turn, affect their performance? And, if so, how do leaders impact this motivational process and its associated outcomes? It has been argued that this research contributes to our understanding of leadership, motivation, and performance in three distinct ways. First, by incorporating goal orientation as its central motivational variable, this study is
8
better able to examine the antecedents and consequences of a broader range of individual motivations that are consistent with contemporary organizational dynamics and demands. Secondly, this research takes a multi-perspective approach to leadership in order to provide richer, more comprehensive perspective of the effects of leadership. Lastly, it incorporates motivational processes into the proposed causal chain in order to more fully test underlying theoretical assumptions and adequately rule out alternative explanations.
9
CHAPTER 2 Model Development and Hypotheses Model Overview Kozlowski and Farr (1988) recommend that future research examine the underlying psychological mechanisms that influence individual learning and
performance. Building on their suggestion, the model proposed herein posits group members' state goal orientation as the key motivational mechanism impacting individual outcomes.1 Consistent with the socio-cognitive approach employed by early goal
orientation researchers (e.g., Dweck, 1986), the central mechanism underlying these relationships center on how unique goal orientation dimensions create different mental frameworks of the task situation that serve to structure cognitive, affective, and behavioral responses. In addition, the proposed model predicts that group members adopt a state goal orientation consistent with the types of behaviors that are rewarded, supported and expected, as communicated by the work-group climate. The logic
underlying this linkage is consistent with climate researchers who demonstrate a connection between climate and its effects on group members, such as shaping attributions of events, attitudes, and behavior (e.g., Hofmann & Stetzer, 1998).
1
The terms "group" and "team" are often used interchangeably in the academic literature and are defined as
"a collection of individuals who are interdependent in their tasks, who share responsibility for outcomes, who see themselves and who are seen by others as an intact social entity embedded in one or more larger social systems" (Cohen & Bailey, 1997, p. 241). However, some note that groups may differ from teams in the level of interdependence that exists among the members (e.g., Cohen & Bailey; Katzenbach & Smith, 1993). Here, I use the term "group" exclusively to convey that interdependence among members is not a necessary condition of this theoretical model.
10
Moreover, as argued in early work on early climate (e.g., Lewin et al, 1939), organizational behavior (e.g., McGregor, 1960), and leadership (i.e., Schein, 1992, the Ohio State and University of Michigan studies), the model presented here argues that climate is established through the leader's priority. Lastly, a moderating role of leadermember exchange is offered to better understand the combined effects of climate and leadership on group-member motivation. These hypotheses are grounded in social
exchange theory (Blau, 1964) and leader-member exchange research (e.g., Dansereau, Cashman, & Graen, 1973). The proposed model centers on three key themes: learning, performance, and avoiding failure. It is argued that these themes pervade work-group and individual dynamics to differentially impact group-member motivations and outcomes. The logic underlying this research suggests that a consistency exists across leadership, work-group climate, group-member state goal orientation, and associated outcomes. This consistency creates unique work group "worlds" that revolve around a particular theme (i.e., learning, performance, avoiding failure), in which leaders and their members act and perceive work-related events in congruence with the ascribed work-group emphasis. It is argued that emergence of a particular theme is largely initiated by the leader's priority on learning, performance, and/or avoiding failure, which drives the formation of a workgroup climate that embodies this priority. It is argued that each of these climates induce group members' adoption of the parallel form of goal orientation, and in turn, differentially impact individual-level outcomes. The quality of the exchange relationship between the leader and employee is viewed as a more relational, proximal leadership
11
process, serving to uniquely complement the existing backdrop established by the leader's priority and associated climate. These relationships are depicted in Figure 1. Figure 1. Proposed Model of Contextual Influences on Individual State Goal Orientation and Associated Outcomes
Leader Priority
Work-group Climate
• Climate for • Employee
development H1 learning H2
Group-member State Goal Orientation
Group-member Outcomes
H3
• Learning • Proving • Avoid
• Task
performance
• Climate for
performance
• Learning
strategies
• Goal
accomplishment
• Climate for
avoiding failure
• Defensive
behaviors
Member Perception of the Quality of Leader-Member Exchange
H4
The following discussion is organized into two sections.
The first section
discusses each of the core constructs depicted in the model to provide a construct definition and the proposed investigated level of analysis. The second section theorizes about each of the relationships depicted above. First, the relation between each specific leader priority and its respective, prototypical work-group climate will be specified. Second, the hypothesized mechanisms linking work-group climate and group-member state goal orientation will be detailed. Third, group members' state goal orientation and its relation to individual-level outcomes will be discussed. Lastly, an interaction between
12
of leader-member exchange and work-group climate will be proposed in affecting individual state goal orientation. Before discussion of these hypothesized linkages, I define each of the core constructs of the proposed model. Construct Definitions Leader Priority. A priority is defined as "something given special or prior
attention" (Random House, 1998, p. 570). Here, leader priority refers to the special attention leaders devote to employee development, goal accomplishment and/or avoiding failure. Dating back to the leadership studies conducted by scholars at Ohio State
University and the University of Michigan (as cited by Bass, 1990; Porter, Steers, & Bigley, 1996), researchers have long acknowledged that leaders "initiate structure," that is, they structure their own work and the work of group members. Setting priorities is one way leaders structure the work group's attention, work, and resources. Here, priorities and goals are viewed as distinct, but interrelated, concepts. First, the concept of priority captures the relative attention paid to a variety of goals, while goals can be examined in isolation of other demands. This notion of relativity is
particularly important in organizational settings where leaders and their members need to manage multiple goals but do not have adequate resources to attend to every goal immediately. Consequently, leaders and their members pay more attention to some goals than others. Secondly, Locke and Latham (1990) purposively choose the word "goal" for their goal-setting theory to connotate emphasis on the end result. Here, I use the concept of priority to include the ongoing process of emphasizing a particular outcome as well as the outcome itself. For this purposes of this study, this more inclusive concept better
13
captures the fuller leadership process surrounding the structuring of group members' attention on a particular focus. This approach of examining leadership with a particular strategic focus or emphasis is relatively new. However, a recent paper by Schneider, Ehrhart, Mayer, and Saltz (2002) examines service leadership, that is, a leader priority on delivering high quality service, as an antecedent to climate. Implicit in their logic is that leaders who prioritize service will be more likely to communicate their commitment through their behaviors. I adopt a similar approach here and propose that relatively consistent actions taken by a leader indicates his/her priority on employee development, goal accomplishment, and avoiding failure. While perhaps not explicit, much of existing leadership literature indicates the existence of these three priorities (i.e., learning, goal accomplishment, and the avoidance of failure). For example, Bass (1985; 1990) identifies individualized consideration and intellectual stimulation as dimensions of transformational leadership. In part,
individualized consideration refers to leaders' provision of developmental support to group members through coaching and/or delegating developmental assignments. Transformational leaders provide intellectual stimulation when they challenge group members to think in novel ways about work challenges, the organization, themselves, and/or the leader. In addition, Manz and Sims (1987; 1992) describe empowering
leadership as the process of leading others to become self-leaders. According to Manz and Sims (1987, 1992), this process involves coaching and developing group members to become leaders. In addition, team leadership researchers assert the need for leaders to emphasize learning and development of their members (e.g., Kozlowski, Gully, Salas, &
14
Cannon-Bowers, 1996; Zaccaro & Marks, 1999).
These examples drawn from
transformational, empowering, and team leadership theory may be viewed as behavioral representations of a leader's priority on employee development. Leader priority on goal accomplishment is evidenced both in academic and practitioner leadership literatures. For example, House's (1971) path-goal leadership model describes a contingency approach in which leaders adopt the best style to maximize follower motivation, and in turn, follower effort and performance based on characteristics of employees and environmental circumstances. In addition, many
scholars have discussed the importance of the establishment of a vision statement to communicate the overall direction and motivate organizational members to accomplishing objectives (e.g., Bass, 1990; Bennis & Nanus, 1985), such a statement communicates the priority of the leader on particular goals. Lastly, leader use of popular management techniques such as Management by Objectives that focus on setting objectives, developing actions plans, monitoring progress, and evaluating performance. These examples illustrate ways in which leaders communicate their priority on accomplishing set goals through their behavior. Behavioral representations of a leader's priority on avoiding failure can be seen in leadership, teams and politics research. Ancona and Caldwell (1992) found that leaders and their teams carefully selected what types of information would be shared with upper management as a way to avoid appearing incompetent. Lastly, Longnecker, Sims & Gioia (1987) found that leaders purposively misrepresented their employees' performance in the appraisal process, in some cases, to promote the appearance of competence of themselves or group members. Edmondson (1996) found that some nurse
15
managers blamed nurses for making mistakes, making them feel as though they were "on trial;" Edmondson (1996) theorized that these units reported less mistakes than other units to avoid the appearance of "failure." These examples demonstrate some of the ways in which a leader's priority to avoid failure to appear competent are expressed through their behavior. These examples illustrate how leader priorities are communicated through behavior and suggest a more complex, subtle process of how leaders transmit their beliefs about what is important. In Schein's (1992) classic work on leadership and culture, he details six mechanisms through which this transmission occurs. These six mechanisms serve as a conduit through which leaders express their priority to group members. More specifically, group members interpret leader behavior revolving around these six areas in ongoing leader-group interactions as reflective of an overall priority on learning, goal accomplishment, or avoidance of failure. Consequently, if relatively consistent, the
patterns of leader behavior converge to elucidate the leader's priority for the group and communicate his/her focus. Differing from the fundamental assumptions of leader-
member exchange theory, leader priorities are assumed to apply to the group as a whole. For this reason, the construct of leader priority is conceptualized as a group-level phenomenon. To better specify the leader behaviors indicative of the leader's priority, each of Schein's (1992) six mechanisms is described below and used to propose a range of leader behaviors indicative of each of the three priorities of interest: employee development, goal accomplishment, and avoiding failure.
16
What leaders pay attention to, measure, and control on a regular basis. This mechanism most directly communicates the leader's priority to group members. Their attention may take the form of questions, casual comments, rewards, priorities on meeting agendas, and measurement of specific behaviors. Conversely, powerful signals are sent to group members when a leader routinely ignores specific stimuli and indicates a lack of importance. Lastly, leaders' strong emotional reactions to specific events may be indicative of their priority on a particular aspect of group functioning. How leaders react to critical incidents and organizational crises. Leaders who label a set of circumstances as a "crisis" convey a level of urgency of the situation to the group members. Moreover, their endorsed method for addressing critical incidents (e.g., through learning, demonstrating ability, avoiding the appearance of incompetence) communicate their assumptions and beliefs about effective crisis management. Signaling a particular way to handle a problem helps to clarify the leader's priority on employee development, goal accomplishment and/or impression management. Criteria for resource allocations. Because resources are limited in organizations, decisions on how to allocate time, money, and other resources can indicate the leader's priority. Resource allocation decisions may, in turn, guide goal setting and identification of strategies for achieving objectives, serving to limit the range of alternative courses of action a leader can endorse for his/her group and elucidating a particular leadership priority. Role modeling, teaching, and coaching. Leaders often role model the behavior they value, expect and believe effective. According to Schein (1992), these informal messages serve as powerful teaching lessons for group members. In addition, a central
17
part of teaching and coaching involves providing feedback to employees.
During
feedback sessions, both the content (i.e., what the leader notices) and focus (i.e., what values are implicit in the leader's message) embed and reinforce the leader's priority within the work-group climate. Criteria for allocation of rewards and status. promoting their ascribed priority through rewards. Leaders reinforce behaviors Rewarding valued behavior
continuously emphasizes their priority. By observing promotion cycles, participating in performance appraisals, discussing performance expectations with their manager, and observing what types of behaviors receive positive and negative acknowledgements from the leader, employees learn what the leader values. Criteria for recruitment, selection, retirement, and excommunication. Schein's (1992) basic assertion is that leaders base their hiring and firing decisions based on the level of congruence between the leader's values and priorities and those expressed by the job candidate or incumbent. These decisions cue group members as to what is important to the leader and facilitate a longer-term shared value system among the group members. Because I hope to sample relatively mature, intact work groups, it is more instructive to examine the criteria leaders use to determine job assignments. Just as with hiring and firing decisions, these criteria signal to members the emphasis the leader places on employee development, goal accomplishment, and/or avoiding failure. Behavioral Indications: Leader Priority on Employee Development. Leaders indicate a priority for employee development when they consistently inquire about how well group members are developing skills and knowledge, consistently discuss how different job assignments may contribute to employees' long-term growth, encourage
18
employees to set developmental goals for themselves, and/or hold regular on-one-one meetings to discuss developmental issues with group members. They will often provide resources (i.e., coaching, training, time off to engage in developmental activities) to facilitate employee development. Moreover, they view job assignments as potential opportunities for development, and make selection decisions based on how the assignment will contribute to the group member's growth. In crisis situations, they react by encouraging group members to experiment with different strategies and remain open to learning from these challenging situations; and, they often model this behavior for their employees. Leaders who prioritize employee development offer more feedback to group members and focus their coaching on skill improvement, rather than solely providing knowledge about the outcome of their performance. They consistently reward group
member effort and a dedication to development and growth. Behavioral Indications: Leader Priority on Goal Accomplishment. express their priority on goal accomplishment when they consistently measure performance against a set standard, pay almost exclusive attention to meeting task standards, measure success only in terms of accomplishing a set performance goal, and routinely ignore why objectives may not have been met, instead choosing to classify the experience as a failure. They allocate resources only to those activities that promote getting the work done. In crisis situations, leaders emphasizing goal accomplishment routinely call on group members who have a demonstrated track record in handling similar circumstances to resolve the problem. If necessary, the leader will step in so that accomplishment of task goals is not compromised. Similarly, in making staffing Leaders
decisions, leaders who prioritize goal accomplishment assign jobs and tasks based on
19
demonstrated ability. Moreover, they consider who will have the greatest likelihood of meeting the standards associated with the task, and therefore, allow the leader and group to receive positive judgments regarding their competence. Feedback from leaders
focused on goal accomplishment simply assesses group member performance against set standards and does not provide insight into how members may improve. Lastly, leaders emphasizing goal accomplishment consistently rewards employees for high levels of performance through recognition, promotion, and other extrinsic rewards. Behavioral Indications: Leader Priority on Avoiding Failure. Leaders who
prioritize the avoidance of failure consistently focus on how competent they or their group members appear to others, measure success in terms of number of errors made, and ignore opportunities to experiment and learn. Their focus is on the appearance of
competence, even if they or their members are not capable. They define a crisis as someone making a mistake and this error becoming known; they deal with critical situations by either avoiding discussion of the error or by "spinning" organizational events so that the leader and his/her employees avoid being characterized as incompetent. Considerable resources, particularly time, are devoted to best positioning ideas and contributions to organizational stakeholders in the hopes of avoiding negative judgments regarding group members' ability. These leaders delegate "safe" assignments to group members and approach assignment decisions with the intent of minimizing errors. They encourage group members to avoid challenging situations that may threaten the appearance of being capable and role model how to appropriately manage impressions so as to appear skilled. These leaders rewards group members for "not screwing up" with promotions, awards, and verbal recognition.
20
In sum, a leader priority refers to the special attention leaders devote to a particular aspect of group functioning. In this study, I focus on three leader priorities: employee development, goal accomplishment and avoidance of failure. While not
explicitly acknowledged in the literature, behavioral expressions of each of the proposed leader priorities have been studied previously. To augment existing research and provide further elaboration, I have employed Schein's (1992) work to articulate the types of behaviors that convey a leader's priority on employee development, goal
accomplishment, and/or avoiding failure. These behaviors are summarized in Table 1 below. Table 1. Behavioral Indications of Leader Priority
Embedding Mechanism (Schein, 1992) What leaders pay attention to and measure How leaders react to crises Leader Priority on Employee Development Employee development Leader Priority on Goal Accomplishment Goal accomplishment Leader Priority Avoiding Failure on
Experiment; learn from mistakes
Send in the person with demonstrated expertise to resolve problem
Leader criteria for resource allocation Leader coaching, feedback teaching, and
Allocate resources to facilitate group member development Focused on skill improvement
Allocate resources to further progress toward operational objectives Focused on providing assessments of performance relative to objectives/ standards Outstanding performance "Who has demonstrated they can do the job?"
Leader criteria for awards Leader criteria for job assignment
Improved performance and level of effort "Who needs development in this area?"
Others' perceptions of leader and group member capabilities Avoid discussions of errors or manage impressions to avoid negative competence judgments Dedicate resources (i.e., time) to managing impressions Focused on avoiding challenging situations and managing the appearance of competence For "not screwing up" "Who will most likely not make errors?" (i.e., placement into "safe" assignments)
21
Work-group Climate: Construct Definition.
Schneider defined climate as
"incumbents' perceptions of events, practices, and procedures and the kinds of behaviors that get rewarded, supported, and expected in a setting" (1990: 384). This definition suggests two broad classes of perceptions: routines (i.e., the events, practices and
procedures), and rewards (i.e., behaviors that are rewarded, supported, and expected) (Schneider, 1990). In this way, climate is a perceptual interpretation of environmental cues offered by the routines and rewards of the particular setting through which employees may understand what is important and valued. With the exception of a recent trend to explore the effects of varying degrees of shared perceptions (e.g., Gonzalez-Roma et al, 2002; Lindell & Brandt, 2000; Schneider, Salvaggio, & Subirats, 2002), climate scholars consistently focus on the shared perceptual nature of climate in their research. Implicit in the definition of climate is the understanding that relatively little variance in organizational member perceptions exist; otherwise, shared interpretations of the setting would not emerge. Consistent with this tradition, theoretical discussions of climate in this proposal assume a high level of perceptual agreement among members and a relatively consistent view of the workrelated events and practices across individuals in a particular work group. Because early climate research examined such a variety of dimensions, the boundaries of what constituted climate became blurred (Schneider, 1975). For this
reason, Schneider (1975) recommended conceptualizing climate with respect to a specific referent, such as service, quality, safety. In other words, climate is best defined, and in turn, understood when considering a central strategic focus (Schneider, 1990). Examples of climates with a strategic focus include a climate for safety (e.g., Zohar, 1980), a
22
climate for service (e.g., Schneider, 1990), a climate for updating skills (e.g., Kozlowski & Hults, 1987), transfer of training climate (e.g., Tracey, Tannenbaum, & Kavanagh, 1995) and in the present case, a climate for learning, a climate for performance, and a climate for avoiding failure. In addition, climate has been examined at multiple levels of analysis including the organization (e.g., Pritchard & Karasick, 1973), the sub-unit, such as a branch of a bank or a plant (e.g., Joyce & Slocum, 1984; Schneider, White & Paul, 1998), the work-group (e.g., Hofmann & Stetzer, 1996), and the individual, also referred to as psychological climate (e.g., Jones & James, 1979), levels. Here, the focus is on the work-group level of analysis, and the climate construct at this level of analysis has received considerable empirical support for its validity (Gonzalez-Roma et al, 2002; Howe, 1977; Powell & Butterfield, 1978; Rentsch, 1990; Zohar, 2000). Work-Group Climate: Dimensionality. In my review of approximately twentyfive empirical studies on climate, close to fifty distinct dimensions have been assessed and measured. Despite the wide variance, however, several dimensions emerge as being the most commonly used in research: 1. Structure of the work environment—formality of the work environment, value placed on protocol; used in 29% of the studies reviewed. 2. Rewards—use of various organizational tangible and intangible rewards; used in 29% of the studies reviewed. 3. Support—availability of necessary resources and social support and encouragement; used in 50% of the studies reviewed.
23
4.
Teamwork and cooperation—mutual support among team members; used in 33% of the studies reviewed.
5.
Participation and involvement in decision making—level of influence in work activities and decisions; used in 29% of the studies reviewed.
6.
Goal emphasis—types of outcomes and standards expected; used in 29% of the studies reviewed.
7.
Communication flow—the frequency of communication; used in 29% of the studies reviewed.
8.
Supervisory/managerial values and practices—the actions, values, expectations, and level of commitment demonstrated by the supervisor and/or higher-level manager; used in 33% of the studies reviewed. While these dimensions surfaced as being the most commonly used, some are
more appropriate for inclusion in this study than others. For example, the dimension of teamwork and cooperation is defined as "the extent to which employees perceive that a friendly cooperative work environment exists" (Kozlowski & Doherty, 1989, p. 549). This dimension describes the general environment and is too generic for use in this research. More specifically, this dimension does not help to define distinguishing
characteristics across the three climates of interest in this dissertation. For example, would a climate that had a high level of teamwork and cooperation be more indicative of a climate for learning, a climate for performance, or a climate for avoiding failure? In attempting to tailor this dimension more specifically to highlight distinguishing features across these three settings, the "teamwork and cooperation" dimension becomes redundant with the "support dimension."
24
In addition, studying supervisory practices as a part of climate blurs the distinction between leadership and climate. Perhaps, the frequency with which
researchers measure managerial practices as a part of climate is in part due to the relative rarity of explicitly testing leadership as an antecedent to climate (Barling, Loughlin & Kelloway, 2002; Kozlowski & Doherty, 1989; Schneider et al, 2002 are notable exceptions). However, most recently, researchers have distinguished between these two constructs by specifying leadership as the guiding force or the influence that sets a direction for group members and the work-group climate as the perceptual medium that interprets and implements this direction through a series of ongoing, complex, social interactions among group members, other organizational stakeholders, customers, and the leader (personal communication: B. Schneider, November 20 2002). Furthermore, the dimensions of work-group climate serve to characterize the patterns of these interactions perceived by group members. Seen in this light, it becomes imprecise, and therefore inappropriate, to consider supervisory practices as a part of work-group climate. The other six dimensions align well with the construct definition of climate and are not too generic, nor redundant. They have been adapted so as to properly describe the three climates of interest in this study (i.e., learning, performance, and avoiding failure). Adjustments to the dimensions described above primarily center on more directly assessing the content and focus of rewards and routines, rather than their frequency or level. Incorporation of these changes yield the following climate dimensions and
definitions: 1. Structure of the work environment—extent to which group members perceive that protocols, standardized practices, and means of accomplishing learning,
25
performance of tasks, and/or avoiding failure are clearly specified (adapted from Abbey & Dickson, 1983; Burke, Borucki & Hurley, 1992; Kopelman, Brief & Guzzo, 1990; Kozlowski & Doherty, 1989; Ostroff, 1993; Solomon, 1986; Tagiuri & Litwin, 1968). 2. Rewards—the extent to which group members perceive that various rewards are used to encourage and acknowledge learning, performance, and/or avoiding failure (adapted from Abbey & Dickson, 1983; Burke et al, 1992; Joyce & Slocum, 1984; Kopelman et al, 1990; Ostroff, 1993; Solomon, 1986; Tagiuri & Litwin, 1968). 3. Support—the extent to which group members perceive that the necessary resources and social support, including feedback from colleagues, is available to promote learning, performance, and/or avoiding failure (adapted from Abbey & Dickson, 1983; Burke et al, 1992; Gonzalez-Roma et al, 2002; Jones & James, 1979; Kopelman et al, 1990; Kozlowski & Doherty, 1989; Kozlowski & Farr, 1988; Kozlowski & Hults, 1988; Rouiller & Goldstein, 1993; Solomon, 1986; Tagiuri & Litwin, 1968; Tracey et al, 1995). To reduce redundancy with the leadership construct, this dimension only focuses on support provided by the organization and colleagues, not work-group leaders. 4. Participation and involvement in decision making—the extent to which the group members perceive their involvement in decision making regarding learning, performance, and/or avoiding failure issues (adapted from Abbey & Dickson, 1983; Drexler, 1977; Kozlowski & Doherty, 1989; Kozlowski & Farr, 1988; Ostroff, 1993; Solomon, 1986).
26
5. Goal emphasis—the extent to which group members perceive learning, performance, and/or avoiding failure as valued and expected outcomes (adapted from Burke et al, 1992; Gonzalez-Roma et al, 2002; Kopelman et al, 1990; Kozlowski & Doherty, 1989; Lindell & Brandt, 2000; Rouiller & Goldstein, 1993; Tracey et al, 1995). 6. Communication—the extent to which group members perceive that work-group communication centers on learning, performance, and/or avoiding failure (adapted from Drexler, 1977; Griffin & Neal, 2000; Jones & James, 1979; Kozlowski & Doherty, 1989; Kozlowski & Hults, 1987; Neal, Griffin, & Hart, 2000; Rouiller & Goldstein, 1993; Tracey et al, 1995). Adequate measurement of work-group climate will hinge on the applicability of these dimensions to the context of the sampled organization. For this reason, these dimensions will be further validated for their appropriateness, as well as the identification of additional dimensions, during the qualitative stage of data collection (please see Chapter 3 for more details). Work-group Climate: Description of Three Prototypical Climates. These six
dimensions will be used in describing three prototypical work-group climates (i.e., climate for learning, a climate for performance, a climate for avoiding failure). These climates are not mutually exclusive; instead, each of these climates may be present within the same work group to varying degrees. Because limited organizational resources
demand focused attention on a limited set of valued behaviors and outcomes, it is likely that one, perhaps two, particular climate(s) is/are more dominant than the others as a way to help work-group members focus on the more important priorities.
27
Climate for learning. The concept of a climate for learning builds on previous work in the employee developmental literature and has been referred to as a "continuous learning culture" (Tracey et al, 1995) and "a climate for technical updating" (Kozlowski & Farr, 1988; Kozlowski & Hults, 1987). Training researchers have explored a similar concept, namely a "transfer of training climate" (Burke & Baldwin, 1999; Rouiller & Goldstein, 1993). Consistent with this prior research, a climate for learning emphasizes continuous development of knowledge and skills. Employees perceive that continuous learning and ongoing engagement in developmental activities is valued, supported, and the expected means for achieving individual and work-group goals (Rosow & Zager, 1988; Tracey et al, 1995). In this type of climate, informal practices and formal systems provide opportunities for development and reinforce learning (Dubin, 1990). Intrinsic rewards such as satisfaction gained from continuous improvement and learning are stressed as important rewards; extrinsic rewards, such as public acknowledgement for learning and promotion based on demonstrated dedication to continuous improvement of one's skills, are also emphasized. Co-workers in climates for learning provide the necessary social support, challenge, encouragement to learn, and feedback to assist in developing and maintaining motivation in the face of challenges (Dubin, 1990; Kram, 1985; Noe & Wilk, 1993; VanVelsor et al, 1998); it is perceived that resources are available, such as training, time to engage in development, and access to other developmental activities (Dubin, 1990; Maurer & Tarulli, 1994; Morrison & Brantner, 1992; Noe & Wilk, 1993). Moreover, in this type of climate, incumbents perceive themselves as being an active participant in
28
deciding when, how, and which skills they need to develop. Lastly, given this description of the structure, rewards, support structure, and decision-making, it is probably not surprising that issues of employee development, learning, feedback, and developmental activities are perceived as being commonly discussed within the group. Climate for performance. The construct of a climate for performance is similar to what safety climate researchers have referred to as "high reliability organizations." High reliability organizations are defined as "organizations in which complex technology must be controlled and complex processes carried out in an error free manner" (Hofmann, Jacobs & Landy, 1996, p. 139). Moreover, within the oil industry, Wright (1986)
observed an organizational pressure to complete the work as quickly as possible. The central theme here is on high levels of productivity, without error. A climate for
performance is a bit broader than observations from the safety climate literature in that performance is not limited to productivity levels and/or efficiency. Performance, as in many knowledge-based organizations, also refers to effectiveness in completion of complex tasks or projects. In a climate for performance, employees perceive achieving the performance standard as the most valued outcome. It is expected that these performance standards will be attained efficiently, effectively, and accurately through applying team members' current skills, knowledge, and abilities. Informal practices and policies support the
accomplishment of work tasks, specifying clear methods or task approaches. Extrinsic rewards, such as promotion and salary, are consistently provided for individuals who are viewed as having high ability in meeting or exceeding performance standards. Group
29
members perceive satisfaction (i.e., intrinsic rewards) when particularly difficult, challenging performance standards are met (Locke & Latham, 1990). Perceived support focuses on removing possible barriers to achieving performance goals, such as a lack of materials, information, staff, or time. In addition, group members support one another to accomplish the task by providing advice on difficult issue and encouragement to meet performance standards. In this type of climate, group members perceive themselves as involved in decisions regarding performance issues, such as goal-setting and performance strategies and approaches, and work-group communication primarily centers around goals, goal accomplishment, and performance strategies. Many of these points complement Locke and Latham's (1990) high performance cycle in which they detail how performance goals and standards motivate high levels of performance. Consistent with their view, rewards in this climate are seen as contingent upon performance, and situational constraints are seen as barriers to performance, and therefore, need to be removed. While the concept of a climate for performance
emphasizes more of the environmental conditions, Locke and Latham (1990) work parallels the central emphasis of this particular climate and supports some of its major distinctions. Climate for avoiding failure. In this context, avoiding failure (i.e., making errors) is extremely valued because it is perceived to be instrumental in maintaining the appearance of competence. Being labeled as "incompetent" risks future opportunities (e.g., promotion), limits access to resources (e.g., information, salary increases), and respect and acceptance from colleagues and organizational leaders. The teams and
30
climate literatures documents how some work groups foster a climate that is not supportive of admitting errors, discussing mistakes, or offering dissenting opinions (e.g., Edmondson, 1996; 1999; Hofmann & Stetzer, 1998). In a climate for avoiding failure, the common practice is to engage in "face saving" behaviors (e.g., avoid discussions of errors and mistakes, asking for help, shifting blame). Typically, group members perceive that they are encouraged to avoid blame, resist taking on assignments that risk failure, and/or avoid taking action so as to evade revealing low ability of a particular group member. Group members perceive being rewarded for "not screwing up" with verbal acknowledgements, promotion, and salary increases. It is perceived that committing errors carry significant punishment such as constant reminders of one's mistake and brutal, demeaning confrontations by others (Edmondson, 1996). Because this environment is unforgiving of even the slightest error, group members do not perceive the same type of social support from their colleagues as in a climate for learning or performance. Here, support comes in the form of covering up for one another. This phenomenon of attributing mistakes to external events to shift blame away from group members has been documented in Hofmann and Stetzer's (1998) work on safety climate. Moreover, because errors are not discussed, group members in this climate are not involved in decisions regarding how to improve group functioning. Work-group discussions center around anything but errors and mistakes, potentially at the expense of addressing key issues preventing members from effectively performing their jobs.
31
In summary, work-group climate has been defined as work-group members' "perceptions of events, practices, and procedures and the kinds of behaviors that get rewarded, supported, and expected in a setting" (Schneider, 1990: 384). In this
dissertation, three prototypical climates will be examined: a climate for learning, a climate for performance, and a climate for impression management. These climates have been described using standard dimensions derived from review of existing literature. A summary of these descriptions can be found in Table 2. Table 2. Descriptions of Three Prototypical Work-group Climates
Climate Dimensions Goal emphasis Climate for Learning Continuously developing and learning new skills, knowledge, and capabilities Formal systems and practices provide opportunities for development and reinforce learning For learning Climate for Performance Achieving performance goals and standards Climate for Avoiding Failure Avoid committing and admitting to mistakes
Structure
Formal systems and practices specify how best to achieve performance goals For high performance
Informal practices endorse avoiding discussion of errors
Rewards
Support
Availability of developmental activities; encouragement and feedback on how to develop from peers Involved in decisions pertaining to one's development Focuses on learning and development
Involvement in decision making Communication
Availability of resources necessary to meet performance goals; encouragement and coaching from peers on how to accomplish specific tasks Involved in decisions pertaining to goal accomplishment Focuses on meeting performance goals
For "not screwing up;" severely punished for making mistakes Group members covering up for one another
Not involved in decisions regarding improvement issues Focuses on anything but mistakes and errors
Goal orientation: construct definition. As previously discussed, goal orientation is defined as one's goal preference in achievement settings (Dweck, 1986) and refers to one's desire to develop, realize, or exhibit the capability to perform a specific activity
32
(Dweck & Leggett, 1988). Initial research identified two broad classes of underlying goals that individuals pursue: mastery and performance (Dweck, 1986; Dweck & Leggett, 1988). Individuals displaying a mastery or learning orientation focus on Those favoring a
building one's competence and/or improving their abilities.
performance orientation seek to demonstrate their competence by seeking favorable judgments and/or avoiding negative judgments regarding their capabilities. Goal orientation provides a mental framework that affects how individuals view challenging situations and shapes their responses in these situations (Dweck, 1986; Dweck, 2000; Dweck & Leggett, 1988). The causal mechanism underlying these two broad classes of goals is one's implicit beliefs regarding intellectual ability and effort. Individuals with a learning orientation hold an incremental view of ability and believe their competencies may be developed through effort and experience. In this view, effort and improved task strategies lead to success. In contrast, individuals with a performance orientation ascribe to an entity view of ability, believing that ability is a fixed personal attribute that cannot be changed. According to this mental framework, higher effort is an indication of lack of ability because individuals who are naturally talented should not have to try so hard to be successful. Goal orientation: dimensionality. Our understanding of the dimensionality of goal orientation has evolved. Initially, Dweck and her colleagues left the question of dimensionality unclear, although some scholars have suggested early work posited goal orientation as uni-dimensional (Button et al, 1996; VandeWalle, 1997). The first
measure developed specifically for adult populations was developed and validated by Button and his colleagues (1996). In this research, they provided support for two, distinct
33
dimensions of goal orientation: learning and performance. Their empirical approach was consistent with the goal orientation theory and research of the time and rigorously compared various models consisting of different combinations of learning and performance orientations. No conceptual work at that time had suggested the potential for a three-factor model, and therefore, they did not test this possibility. More recent empirical evidence supports further partitioning performance orientation into two components, i.e., prove and avoid, arguing that "seeking favorable judgments" and "avoiding negative judgments" are two distinct goals (Elliot & Harackiewicz, 1996; Middleton & Midgley, 1997; VandeWalle, 1997). Accordingly, individuals with a proving orientation prefer tasks that allow them the opportunity to demonstrate their capabilities and gain favorable feedback. Their avoiding-oriented
counterparts avoid responsibilities that may reveal their shortcomings to others. It is important to note that both individuals with a proving orientation and individuals with an avoiding orientation subscribe to the entity theory of intellectual ability. Goal orientation: malleability. Researchers consider goal orientation to be a stable, dispositional construct that is susceptible to situational influences. Researchers treating goal orientation as a dispositional trait, measure goal orientation, rather than induce it (Bell & Kozlowski, 2002; Brett & VandeWalle, 1999; Brown, 2001; Chen et al, 2000; Colquitt & Simmering, 1998; Elliot & Church, 1997; Elliot & McGregor, 1999; Fisher & Ford, 1998; Ford et al, 1998; Middleton & Midgley, 1997; Potosky & Ramakrishna, 2002; Phillips & Gully, 1997; Towler & Dipboye, 2001; VandeWalle et al, 1999; VandeWalle, Cron, & Slocum, 2001; VandeWalle & Cummings, 1997). This body
34
of research predicts specific behavioral and motivational outcomes based on individual differences in goal orientations. Researchers treating goal orientation as more malleable induce specific forms of goal orientation through situational influences and often refer to this construct as state goal orientation. A state orientation is a temporary preference for specific achievement goals. Dweck and Leggett (1988) suggest that when the situation offers few cues as to which orientation to adopt, individuals will rely on their dispositional goal orientation. If, on the other hand, the circumstances increase the value or saliency of a specific orientation, "predispositions should be overridden and greater homogeneity among individuals will result" (1988: 269-270). These arguments are consistent with Mischel (1977) who asserts that situational strength impacts the extent to which dispositional variables become observable through behavior. According to Mischel (1977), strong situations encourage individuals to
construe the situation similarly, induce conformity in expectations regarding appropriate behaviors, reward expected behaviors, and require skills individuals possess. On the other hand, weak situations allow for variation in perceptions and behavioral responses. Personality researchers suggest that situations such as experimental settings create strong situations through clearly articulated expectations and requirements and structured tasks (Schneider, 1983; Weiss & Adler, 1984). Consequently, in strong situations, personality influences do not emerge as powerful predictor of behavior. Empirical evidence documents that trait goal orientation has a state counterpart. Through confirmatory factor analysis, Button et al (1996) showed that dispositional and state goal orientations are distinct dimensions, and situational influences may cause
35
individuals to adopt an orientation different than their dispositional one. Additional research has documented that the strength of specific situational cues, such as framing of the task, competition, extrinsic rewards, and evaluations standards, cause individuals to adopt a state goal orientation (Ames & Archer, 1988; Butler, 1993; Elliot & Harackiewicz, 1996; Elliott & Dweck, 1988; Harackiewicz & Elliot, 1993; Nicholls, 1984; Stevens & Gist, 1997). In studies by Butler (1993), Elliot and Harackiewicz (1996), Harackiewicz and Elliot (1993), and Stevens and Gist (1997), researchers induced state goal orientations through experimental manipulations. Adult participants were
randomly assigned to an experimental condition in which a different goal (i.e., performance or mastery) for the experimental task was emphasized. For other studies, state goal orientation was induced through task instructions that emphasized the fixed or incremental nature of ability (Mangos & Steele-Johnson, 2001; Steele-Johnson et al, 2000). Even though the framing of the task was subtle, it proved effective in inducing state goal orientations among participants. Studies conducted with samples of children demonstrate a similar pattern of results. With a sample of fifth graders, Elliott and Dweck (1988) randomly provided each student with a learning or performance goal. Students with the performance goal consistently demonstrated behavior consistent with a performance orientation. Likewise, students with the learning goal exhibited behavior consistent with a learning orientation. Similarly, Ames and Archer (1988) examined how the teacher's emphasis on performance or learning goals in the classroom affected junior high and high school students. Perceptions of learning or performance goals led to similar behavioral patterns
36
found by Elliott and Dweck (1988). These studies highlight the potency of goals and the situational influence of an authority figure on state goal orientation. Given the most recent empirical evidence on the dimensionality and malleability of goal orientation, three dimensions of state goal orientation will be investigated at the individual level of analysis in this dissertation. The current study extends this previous work by examining state goal orientation in an organizational setting with a working adult population, rather than in an experimental setting. This methodology will allow for greater understanding of how naturally occurring contextual factors, such as leadership and work-group climate, affect individual state goal orientation and associated outcomes. Individual outcomes. To better capture the multi-faceted construct of
performance, three primary categories of outcome variables are investigated in this dissertation: task performance, learning strategies, and defensive behaviors (e.g., avoiding challenge, blaming, excuse making). The more specific nature of the outcome variables examined in this study, in part, will be driven by the context of the sampled organization, which has yet to be determined. Therefore, I provide below specific
definitions of these constructs as well as examples from previous research as a way to clarify the meaning of the outcome variables of interest. Task performance refers to the quality, accuracy, and quantity of performance associated with execution of the task and will be investigated at the individual-level of analysis. Examples of task performance from previous studies include exam performance (e.g., Chen et al, 2000; Colquitt & Simmering, 1998; Elliot & McGregor, 1999; Phillips & Gully, 1997), quality of presentation (e.g., Brett & VandeWalle, 1999), performance of skills learned in training and transferred to another context (Ford et al, 1998; Kozlowski
37
et al, 2001), quality of performance on problem-solving tasks (e.g., Butler, 1993), sales volume (e.g., VandeWalle et al, 1999), and accuracy on problem solving tasks (e.g., Steele-Johnson et al, 2001; Mangos & Steele-Johnson, 2001). As argued by other scholars (e.g., Hall, 2002), individual learning and adapting are necessary for employee effectiveness given changing business, organizational and career demands. Learning refers to "a relatively permanent change in knowledge or skill produced by experience" (Weiss, 1990, p. 172). This definition suggests the need to document a relatively stable change through a comparison of pre- and post-assessments of knowledge and/or skill and is difficult to study in cross-sectional designs, such as the one proposed here. For this reason, this dissertation adopts the same approach taken by others and examines precursors of learning, that is, learning strategies (e.g., Fisher & Ford, 1998; Ford et al, 1998; Kozlowski et al, 2001). Learning strategies refer to "an internal process by which learners select and modify their ways of attending, learning, remembering, and thinking" (Gagne, Briggs, & Wager, 1992, p. 66). These strategies become apparent through exhibited behavior. Examples of these types of strategies studied by previous researchers include feedback seeking (e.g., Butler, 1993; Middleton & Midgley, 1997; VandeWalle & Cummings, 1997), metacognition (e.g., Ford et al, 1998), and other strategies, such as practice, self-maintenance activities, organizing knowledge into coherent structures, and elaboration (e.g., Brown, 2001; Fisher & Ford, 1998; Kozlowski et al, 2001; Stevens & Gist, 1997). Consistent with the research cited above, these types of behaviors are viewed as more proximal indicators of learning and will be investigated at the individual-level of analysis.
38
According to some scholars, one barrier to achieve the learning and adaptation necessary in today's business environment is defensive behaviors (e.g., Edmondson, 1996). Defensive behaviors are defined as "reactive and protective actions intended to reduce a perceived threat to or avoid an unwanted demand of an individual or group" (Ashforth & Lee, 1990, p. 622) and may include avoiding delivering "bad news," avoiding blame, and avoiding taking action that might implicate oneself. behaviors have been studied as a part of political behavior (Valle, 1997), means to managing mentor-protégé relationships (Tepper, 1995), and as a barrier to team learning and effectiveness (Ancona & Caldwell, 1992; Edmondson, 1996). Conventional These
wisdom and the extant literature, albeit it limited, suggests that these behaviors are largely dysfunctional and negatively impact group performance and functioning. Here, they are particularly relevant to understanding how different dimensions of goal orientation may affect the frequency in which individuals engage in these potentially dysfunctional behaviors, an area of study that has received no research attention to date. Quality of leader-member exchange. The construct of leader-member exchange relationships, also commonly referred to as LMX, is built on two fundamental and interrelated theoretical perspectives: role making and social exchange theory (Blau, 1964; Graen, 1975). Graen (1975) posited that dyads of interdependent organizational
members engage in the role making process to define how each will behave and agree on the general nature of their relationship. In particular, the relationship between the leader and subordinate, also referred to as vertical dyad linkages in early LMX research, becomes especially important in shaping role expectations of the member (Graen, 1975; Graen & Cashman, 1975; Dienesch & Liden, 1986), and therefore, is the central
39
relationship of interest. Moreover, through the role-making process, differentiated role definitions develop between leaders and each of their members, resulting in varied leader-member exchange relationships within a particular group. Drawing on social exchange theory (Blau, 1964), LMX researchers have argued that the basis for exchange relationships lie in the valued resources each party may provide (e.g., Liden, Sparrowe & Wayne, 1997). For example, leaders may provide challenging assignments, mentoring, and/or informational resources. In exchange,
members offer valued resources such as greater levels of effort and/or loyalty. Based on this theoretical background, the most commonly-used working definition of leader-member exchange is the "quality of the exchange relationship between leader and subordinate" (Schriesheim, Castro, & Cogliser, 1999, p. 77). Often times, LMX is operationalized by having leaders, their members, and/or both parties respond to a set of items assessing the quality of the relationship. The analysis is then conducted at the individual level using the perceptions from the leader, member or both. Recently, however, researchers have questioned this practice because of the logical inconsistency of theoretically defining the construct's level of analysis at the dyadic or relationship level and testing the hypothesized relationships at the individual level (Schriesheim et al, 1999; Schriesheim, Castro, & Yammarino, 2000). While this is a valid criticism, for the purposes of this dissertation, the compelling force believed to drive whether a group member adopts a particular state goal orientation is the member's perception of the quality of the relationship, not necessarily the dyad's assessment of the quality of this relationship. Focusing on member perceptions is logically consistent with examining the role of work-group climate, another perceptual medium. For these
40
reasons, this dissertation will focus on the member's perception of LMX, and it will be investigated at the individual level of analysis. Summary of Construct Definitions and their Level of Analysis. In sum, this dissertation will focus on five key constructs: leader priority, work-group climate, group member state goal orientation, individual outcomes, and member's perception of the quality of leader-member exchange relationship. Each of these constructs has been defined in the preceding discussion, and these definitions are summarized in Table 3 below. Table 3. Construct Definitions
Construct Definition Proposed Investigated of Analysis Group Group Level
Leader priority Work-group climate
State goal orientation Individual Outcomes -Task performance Learning strategies -Defensive behaviors
Member perception of the quality of LMX relationship
An aspect of the group functioning given special attention by the leader (adapted from Random House Dictionary). Group members' "perceptions of events, practices, and procedures and the kinds of behaviors that get rewarded, supported, and expected in a setting" (Schneider, 1990, 384). A temporary preference for specific achievement goals. Task performance: the quality, accuracy, and quantity of performance associated with execution of the task Learning strategy: "an internal process by which learners select and modify their ways of attending, learning, remembering, and thinking" (Gagne, Briggs, & Wager, 1992, p. 66). Defensive behaviors: "reactive and protective actions intended to reduce a perceived threat to or avoid an unwanted demand of an individual or group" (Ashforth & Lee, 1990, p. 622) "quality of the exchange relationship between leader and subordinate" (Schriesheim, Castro, & Cogliser, 1999, p. 77).
Individual Individual
Individual
Hypothesized Relationships As highlighted previously, the model proposed herein posits group leaders' priorities as the key drivers of three unique climates described in the preceding discussion. It is predicted that group members adopt a state goal orientation consistent
41
with the types of behavior that are rewarded, supported and expected, as communicated by the work-group climate. In turn, group members' state goal orientation is posited as the key motivational mechanism impacting individual outcomes. Lastly, it is theorized that the effects of climate and the role of the exchange relationship interact to impact group member state goal orientation. In the following section, the rationale for each of these hypotheses is detailed in the order highlighted above. Relation between leader priority and work-group climate. There is a strong
historical tradition documenting the key role of leadership in establishing work-group climate. In their classic works, Likert (1967) and McGregor (1960) asserted that leaders transmit their beliefs and expectations through the climate they create. McGregor (1960) suggests that managers convey their attitudes about workers, consistent with Theory X or Theory Y, through their behavior. In turn, this behavior establishes the work-group climate. Lewin et al (1939) empirically tested the role of leadership in establishing workgroup climate. Across three leadership conditions (i.e., democratic, authoritarian, and laissez-faire), Lewin et al (1939) found notable differences in followers' attitudes and behaviors between the groups. More recently, Eden and his colleagues' research on the Pygmalion Leadership Effect (2000) shows how leaders transmit their expectations and beliefs regarding their group member ability to create a supportive climate. These
findings suggest a group-level phenomenon that is shaped by the leader and, in turn, affects the behaviors and attitudes of group members. Schneider and Reichers (1983) provide three theoretical explanations for why climates emerge, two of which are relevant to this discussion. First, citing Payne and Pugh (1976), they suggest one way in which climate is influenced is through
42
organizational members' perceptions of objective characteristics of the work setting, such as the extent of centralization of decision making and the degree to which rules and policies constrain behavior. According to Payne and Pugh (1976), this structure shape how organizational members perceive organizational events. This structural argument complements early and contemporary leadership theory. According to researchers involved in the Ohio State and University of Michigan studies (as cited in Bass, 1990; Porter, Steers, & Bigley, 1996), leaders structure the work of their employees through rules, policies and goal setting. This basic idea can be seen in more recent theorizing of leadership in climate. For example, Kopelman et al (1990) and Ostroff and Bowen (2000) argue that human resource practices give rise to climate. In work-group settings, human resource practices are executed in large part by group leaders (Zohar, 2000) and are used to organize and guide the work of group members. In turn, this imposed structure influences group member perceptions of group events and rewards. As an example, Smith-Jentsch, Salas, and Brannick (2001) document that team leadership as exercised through specific job assignment practices, such as delegation of opportunities to perform trained tasks (e.g., Ford, Quinones & Sego, & Sorra, 1992), lead to a transfer of training climate. An additional argument presented by Schneider and Reichers (1983) posits an interactionist perspective, stating that "climates emerge out of the interactions that members of the work group have with each other" (p. 30). Through these interactions, shared meaning and perceptions of objective aspects of the work environment emerge. Drawing on social comparison and social conformity theory, Ashforth (1986) argues
43
greater conformity of perceptions and meaning converge on a particular "fulcrum" when a "compelling referent" exists in the work group. A leader acts as a "compelling referent" in the work group. Their formal status in the organization provides them a heightened degree of saliency and value within the group. Drawing from the safety climate literature, Hofmann et al (1995) echo Ashforth's (1986) assertions by arguing that workers motivation and attitudes mirror the leaders' priorities on safety, thereby impacting the efficacy of safety climate. Leader-group communications have been shown to impact how members perceive the openness of the climate, and in turn, shape their attributions of work events and willingness to admit mistakes (Edmondson, 1996; Hofmann & Stetzer, 1998). In addition, Scott and Bruce (1994) find that leader expectations affect a climate for innovation, and in turn, innovative behavior. Lastly, Barling et al (2002) show that transformational leaders influence climate through expressing idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration. These aspects of leadership provide opportunities for interaction between leaders and their group members. These interactions convey the leader priorities, thereby shaping climate perceptions. In sum, building from Payne and Pugh (1976) and Schneider and Reichers' (1983) structural arguments, it has been argued that, through communication of their priorities, leaders provide the necessary structure to shape group members perceptions of group routines and rewards. A second complementary argument adopted from Schneider and Reichers (1983) and Ashforth (1986) asserts that specific forms of climate emerge through interactions among group members and their leaders. During these interactions, leaders represent a "compelling referent," driving conformity of perceptions to converge
44
on their central priorities. Moreover, more recent empirical evidence substantiates the link between leadership and climate. For these reasons, I offer the following general proposition, accompanied by more specific hypotheses: Proposition 1: The leader's priority directly impacts work-group climate. Hypothesis 1a: The leader's priority on employee development is positively related to a climate for learning Hypothesis 1b: The leader's priority on goal accomplishment is positively related to a climate for performance Hypothesis 1c: The leader's priority on avoiding failure is positively related to a climate for avoiding failure. Relation between work-group climate and group-member state goal orientation. In part, the relationship between work-group climate and group-member state goal orientation hinges on the question of whether one's perceptions of the environment can affect his/her motivation. The broader motivational literature demonstrates a relationship between individual perceptions and one's motivation. For example, equity theory
(Adams, 1965) posits that motivation is a function of how an individual perceives his/her ratio of efforts to rewards as compared to that of another party. These perceptions of oneself in comparison to another drive individual motivation. Similarly, expectancy theory predicts effort will be a function of an individual's perceptions regarding the utility of effort, value of the reward, and probability an achieved outcome will result in receipt of reward (Vroom, 1964). Lastly, Bandura's (1997) triadic reciprocal causation model stipulates that perceptions of factors relating to the environment, individual and his/her behavior continuously interact to affect one another. Moreover, individual
45
factors, such as motivational constructs (e.g., self-efficacy), are products and drivers of the perceived environment and one's behavior. Consistent with this logic, the reasoning underlying the model proposed herein suggests that individual perceptions of their work environments, i.e., work-group climate and their relationship with the leader, affect one's motivation, and in turn, their behavior. Consistent with the broader literature on motivation, research support the logic that climate impacts individuals motivationally, and in turn, their performance outcomes (Griffin & Neal, 2000; Kopelman et al, 1990; Neal et al, 2000). Kopelman et al (1990) posit that climate affect motivation by helping to clarify outcome expectancies through rewards, influencing individuals' self-efficacy through task support, and affecting the valences of particular behaviors and outcomes through rewards and group structures. Empirically, Griffin and Neal (2000) and Neal et al (2000) demonstrate that climate influences individual motivation levels, and in turn, performance. However, as Bandura (1997) points out, person-factors, such as personality, affect causal relationships between environmental factors, motivational states, and behavioral outcomes. Consequently, the relationship of work-group climate to state goal orientation also depends on the extent to which this contextual factor is compelling enough to override personal preferences. Mischel (1977) theorized that situational strength impacts the extent to which personality factors are dominated by contextual influences. He describes strong situations as ones that lead situation participants to hold uniform perceptions of events and expectations of appropriate behaviors and receive rewards for behaving consistent with expectations. On the other hand, weak situations are not
sufficiently compelling to cause situation incumbents to share perceptions regarding
46
events and expected behaviors. Mischel's theory (1977) predicts that in strong situations, individuals are more likely to behave consistent with situational demands. Conversely, in weak situations, individual behavior will be more reflective of individual difference factors. Like other strong situations, a work-group climate provides clear situational cues as to what behaviors are expected, valued, and rewarded. However, why might
individuals conform to the demands of a strong situation, such as work-group climate? Schneider (1975) provides two theoretical arguments that explain why individuals align their attitudes, motivations, and behavior with the existing climate. First, Gestalt
Psychologists posit that the duty of every individual is to understand the order that objectively exists, as represented by situational cues, and to behave consistently with this apprehended order. In this way, Gestalt Psychologists assert that individuals are driven to make sense of their patterned context and to act in accordance with the demands of this environment. Consequently, acting consistent with the demands of the climate, fulfill
individuals' need for order and understanding of their environment. Secondly, Schneider (1975) describes the Functionalist school of thought as an additional rationale for why individuals conform to the existing climate. Similar to Gestalt Psychology, Functionalists assert that individuals perceive individual environmental cues as representative of a larger, patterned order of their environment; however, Functionalists believe that individuals apprehend this order in order to effectively adapt to their environment, not because they are driven to without choice. Moreover, this school of thought asserts that people behave consistently with this perceived order to achieve homeostatis and harmony with their environments. This logic
47
implies that individuals would be greatly challenged to resist conforming to the climate they perceive (Schneider, 1975). Taken together, Gestalt Psychology and the
Functionalist tradition extends the suggestion that work-group climate provides a strong situation to provide rationales for why individuals are driven to adapt their perceptions, motivations, and behavior in accordance with the expectations conveyed by the climate. Recent research and theory supports these theoretical assertions that workgroup/sub-unit climate affect individual outcomes (Joyce & Slocum, 1984; Hofmann, Morgeson, & Gerras, in press; Hofmann & Stetzer, 1996; Hofmann & Stetzer, 1998; Zohar, 2000). In their investigation of three plants within a duty truck manufacturer, Joyce and Slocum (1984) found that unique plant climates were differentially related to individual performance and job satisfaction. Hofmann and Stetzer (1996, 1998) found that a climate for safety shaped attributions for the causes of workplace accidents and impacted the frequency of unsafe behaviors and accidents. Similarly, Zohar (2000) found that a work-group climate that encouraged and valued safe behaviors led to a decrease in injuries on the job. Lastly, Hofmann et al (in press) found that together with leadership, work-group safety climate impacted individual perceptions of their organizational role, and in turn, affected safe behaviors. Taken together, these results suggest that workgroup climate is a strong situation that influences individual-level phenomena. While research has not explored the specific question of whether work-group climate affects individual state goal orientation, theory and empirical evidence regarding goal orientation suggests the existence of such a relationship. Dweck and Leggett (1986) assert that when strong situational cues are present, as in the case of work-group climate, individual dispositional goal orientations may be dominated by a state goal orientation.
48
Nicholls (1984) provides additional clarification regarding the specific conditions that impact state goal orientation. According to Nicholls (1984), individuals are more
inclined to hold a mastery or learning orientation if there is a moderate amount of task challenge and task-related extrinsic rewards are not salient. Because in a climate for learning, growth and development are emphasized, completion of easy tasks is not valued. Consequently, in this climate, on average, employees tend to be faced with moderate to high task challenge. In addition, intrinsic satisfaction, gained through
professional growth, is emphasized. These points suggest that individuals will adopt a state learning orientation when working in a climate for learning. Nicholls (1984) asserts that certain conditions encourage individuals to adopt a performance orientation: (1) when individuals are cognizant that their performance on a task is being evaluated, (2) when competition among group members is fostered, and (3) when task performance is highly visible. The saliency of task evaluation and level of competition are particularly relevant here. In both a climate for performance and avoiding failure, employees receive subtle and explicit messages that their performance is being evaluated. In the climate for performance, employee performance is consistently gauged against set standards. Performance and achievement are emphasized consistently, heightening the saliency of performance evaluation. In addition, task and socio-emotional support are around
attaining performance goals, reinforcing the notion that performance is being monitored and evaluated continuously. In a climate for avoiding failure, the consistent evaluation of performance is subtler. Employees are expected to avoid committing and admitting mistakes in order to
49
appear competent. Consequently, group members are consistently monitoring potential errors that may need to be covered up, thereby heightening the saliency of sub-par performance. Moreover, in these types of climates, the focus is on avoiding negative competence assessments. Consequently, attention is focused almost exclusively on the evaluation of performance rather than on possible task strategies, such as experimentation and learning from mistakes. Typically, organizations have a fixed amount of resources, including rewards. For this reason, emphasis on extrinsic rewards fosters competition among team members. However, in a climate for learning, this "fixed pie" perception is offset by the emphasis on intrinsic rewards such as intrinsic satisfaction for improving and preserving in the face of challenge. In climates for performance and avoiding failure, intrinsic rewards are not salient; therefore, employees more actively compete for extrinsic rewards. Consequently, the perception of a fixed quantity of rewards breeds competition within these climates, encouraging the emergence of state performance orientations. Lastly, researchers have consistently shown that individuals adopt state goal orientations consistent with situational demands. As reviewed earlier, Butler (1993), Elliot and Harackiewicz (1996), Harackiewicz and Elliot (1993), Mangos & SteeleJohnson (2000), Steele-Johnson et al (2000), and Stevens and Gist (1997) induced state goal orientations among respondents that were consistent with the experimental task instructions. Studies conducted with samples of children demonstrate a similar pattern— participants conform their state goal orientation to situational demands (Ames & Archer, 1988; Elliott & Dweck, 1988).
50
Based on motivational theory, climate research and theory, arguments for climate as a strong situation, goal orientation theory and related empirical evidence, I propose the following proposition and hypotheses: Proposition 2: Work-group climate is directly related to team members' state goal orientation. Hypothesis 2a: A climate for learning is positively related to team members' state learning orientation. Hypothesis 2b: A climate for performance is positively related to team members' state prove performance orientation. Hypothesis 2c: A climate for avoiding failure is positively related to team members' state avoid performance orientation. Relation between state goal orientation and outcomes. Consistent with social cognitive theory (Bandura, 1986), which stipulates that learning and behavior is motivated and regulated by one's cognitions, the theory underlying goal orientation centers on a socio-cognitive approach to motivation. Early goal orientation scholars theorized that one's goal orientation, derived from their implicit beliefs of ability, created cognitive frameworks that shape their interpretations of achievement settings (Dweck, 1986; Dweck & Leggett, 1988). More specifically, two sets of underlying cognitions drive behavioral differences: one set center on interpretations of the task while the other set deal with cognitive responses during task performance. Research shows that when state goal orientation is induced, these cognitions emerge in parallel form to their trait goal orientation counterpart (Ames & Archer, 1988; Elliott & Dweck, 1988; Mangos & Steele-Johnson, 2001). For example, Elliott and
51
Dweck (1988) found that children given specific achievement goals (i.e., learning, performance) mirrored the same cognitive, affective, and performance patterns as children relying on their own trait goal orientation. Ames and Archer (1984) found similar results: perceptions of task challenge and causes of success and failure shaped by the type of achievement goal the children's classroom promoted paralleled their respective trait goal orientation cognitions. Lastly, consistent with trait explanations, Mangos and Steele-Johnson (2001) found that a state goal orientation was related to perceptions of task complexity. Because similar sets of cognitions accompany state goal orientations as their trait counterparts and these cognitions are reasoned to drive behavioral differences, I rely on trait and state goal orientation evidence in theorizing about the relationships between state goal orientation and outcomes. For the sake of clarity, however, I note those studies that employed a state approach. Goal orientation theory suggests that learning-oriented individuals view achievement settings as opportunities to develop their competence (Dweck, 1986; Dweck & Leggett, 1988). Toward this end, individuals holding a learning orientation will devote the necessary effort to persevere in the face of failure to achieve higher levels of performance. Empirical evidence supports this theoretical claim and has shown that learning orientation predicts higher levels of task performance (Brett & VandeWalle, 1999; Butler, 1993; Chen et al 2000; Ford et al, 1998; Phillips & Gully, 1997). Brett and VandeWalle (1999) demonstrated that MBA students with a mastery orientation attained higher ratings for the quality of their in-class presentations. In adopting a state approach, Butler (1993) found that individuals subjected to a mastery orientation experimental condition asked more task-related questions and, in turn, received higher scores on a
52
problem-solving task. Ford et al (1998) demonstrated that individuals with a learning orientation made more correct decisions after receiving training on a PC-based, dynamic decision-making simulation program. In addition, Chen et al (2000), Phillips and Gully (1997), and VandeWalle et al (1999) found consistent results indicating a positive, indirect relationship between learning orientation and graded exam performance. Dweck and her colleagues (1986; 1988; 2000) further suggest that learningoriented individuals will employ effective learning strategies to self-regulate their responses to challenging settings. Organizational research provides convincing evidence substantiating this link. VandeWalle and Cummings (1997) document that learningoriented individuals perceive higher value in feedback, and as a result, engage in more feedback seeking behavior. Brett and VandeWalle (1999) found that individuals holding a learning orientation were more likely to set learning goals. Ford et al (1998) found that trainees with a learning orientation monitored their own thought processes so as to pinpoint ways to improve and maintain active mental engagement in the training. Fisher and Ford (1998) found that the effect of mastery orientation is positively related to effort and negatively related to off-task attention. Research on state goal orientation offers similar findings, further substantiating a link between state learning goal orientation and use of learning strategies. Stevens and Gist (1997) found that individuals in a mastery condition engaged in more skillmaintenance activities after receiving negotiation training. Kozlowski et al (2001) and Schmidt, Chambers, Kozlowski & DeShon (2001) found that individuals with a learning state goal orientation engaged in sophisticated cognitive strategies, such as organizing their knowledge into consistent patterns and meta-cognitive activities. Lastly, Ames &
53
Archer (1988) found that students engaged in higher levels of learning strategies, such as self-planning and monitoring, when operating in a state mastery condition. Together, these studies provide a sound empirical justification to predict that state learning orientation is positively related to use of learning strategies. Because a state and trait learning orientation serves to regulate cognitive functioning by focusing individual attention on the task, rather than on preserving one's ego (e.g., Dweck, 1986; Nicholls, 1984), it is unlikely that learning-oriented individuals would engage in defensive behaviors. By their very nature, defensive behaviors require individuals to shift their attention from the task to preserving their appearance of being competent. In this way, these behaviors would require increased off-task attention, a condition that contradicts Fisher and Ford's (1998) finding that mastery orientation is negatively related to off-task attention. While limited empirical evidence exists, theory and logic make it reasonable to expect a negative relationship between state learning goal orientation and defensive behaviors. Theory suggests that individuals with a performance orientation interpret challenging situations as potential threats to their confidence (Dweck, 1986; Dweck & Leggett, 1988. As a result, these individual withdraw from the task and experience significant declines in performance in the face of obstacles (Dweck, 1986, 2000; Dweck & Leggett, 1988). Consistent with this assertion, empirical research has shown a
negative relationship between individuals holding an avoid orientation and task performance. Elliot and Church (1997) and Elliot and McGregor (1999) found that individuals who held avoid performance orientations demonstrated lower levels of performance on exams. VandeWalle et al (1999) found similar results: students high on
54
avoid performance orientation exhibited lower levels of effort and self-efficacy, which in turn, produced lower levels of performance on exams. Elliot and Harackiewicz (1996) showed that a state approach-avoidance, i.e., state avoid orientation, was negatively related to task involvement. While they did not measure task performance, it seems likely that if individuals are disengaged in a task, their task performance would suffer as a result. Theory and evidence suggests that an avoid orientation is negatively related to task performance. Because performance-oriented individuals believe that ability cannot be augmented or developed, theory predicts that they would not engage in strategies to learn new competencies (Dweck, 1986, 2000; Dweck & Leggett, 1988). Holding true to this assertion, research documents that performance orientation was not related to metacognition (Ford et al, 1998), rehearsal of new skills, organization of new material, using examples to better learn (Fisher & Ford, 1998), nor practice level (Brown, 2001). Adopting a state approach, researchers have found similar results: state performance orientation is unrelated to learning strategies (Ames & Archer, 1988) and organizing coherent knowledge structures (Kozlowski et al, 2001). It is important to note that these studies did not tease apart the effects of avoid and prove goal orientation. A few studies provide more fine-grained insights into how avoid and prove orientation might be related to use of learning strategies. VandeWalle and Cummings (1997) found that avoidant individuals perceive higher cost associated with feedback and engaged in less feedback seeking behavior. However, prove orientation was unrelated to perceive cost of feedback and feedback seeking behavior. Elliot and Harackiewicz's (1996) finding that state avoid orientation is negatively related to task involvement
55
suggests the potential for decreased use of learning strategies, particularly given that other researchers have shown a significant relationship between effort and involvement and learning strategies (Fisher & Ford, 1998). These findings do not provide convincing evidence of a relationship between prove orientation and learning strategies; therefore, no formal hypothesis is offered. However, this research does suggest a negative relationship between avoid orientation and learning strategies. Additional research and logic further justifies a hypothesized, negative relationship between avoid orientation and learning strategies. Learning from errors and experimentation have been shown to be effective learning strategies for organizations as well as individuals (e.g., McCall, Lombardo & Morrison, 1988; Sitkin, 1992). However, engaging in these types of behaviors is perceived as particularly threatening to those with an avoid goal orientation. To them, making errors, asking for assistance, and/or
experimenting only invites others to perceive them as incompetent, an outcome they are actively motivated to avoid. Moreover, it seems reasonable to expect that these
individuals will engage in more defensive behaviors when errors do occur (e.g., excuse making, blaming others) in order to preserve the façade of being competent. Taken together, this research and logic suggests that avoid orientation and learning strategies are negatively related, and avoid orientation and defensive behaviors are positively related. Dweck and Leggett (1988) and Heyman and Dweck (1992) note that an emphasis on performance goals may be adaptive in certain circumstances and have a positive effect on task performance. Dweck and Leggett assert: "it is often important for individuals to evaluate their abilities or to gain positive judgments of their competence" (1988, 260). The emphasis on gaining positive competence judgments is consistent with a proving
56
orientation, and research has shown that a prove performance orientation is positively related to task performance. Brett and VandeWalle's (1999) study of MBA students demonstrated that individuals with a proving orientation adopted goals to refine their skills and to perform better than others to positively impact the quality of their presentations. Elliot and McGregor (1999) and Elliot and Church (1997) found a direct, positive relationship between prove performance orientation and exam performance, while VandeWalle et al (1999) found an indirect, positive relationship between prove orientation and exam performance. In sum, guided by extant theory and research, I propose the following proposition and hypotheses: Proposition 3: individual outcomes. Hypothesis 3a: Group members' state learning goal orientation will be positively related to task performance and use of learning strategies and negatively related to defensive behaviors. Hypothesis 3b: Group members' state prove performance goal orientation will be positively related to task performance. Hypothesis 3c: Group members' state avoid goal performance orientation will be negatively related to task performance and use of learning strategies and positively related to defensive behaviors. Role of the quality of leader-member exchange relationship. Leader-member exchange (LMX) describes a relational leadership process. Because leaders develop different types of relationships with each of their members (Graen & Cashman, 1975), Group members' state goal orientation is directly related to
57
work group members experience different levels of access to the leader and varying degrees of a sense of obligation to meet leader expectations (Dienesch & Liden, 1986; Graen & Cashman, 1975; Graen & Uhl-Bien, 1995; Liden et al, 1997). These two mechanisms, level of access and sense of obligation, characterize the quality of the leader-member exchange relationship. It is argued that leader-member relationships
provide a conduit for transmission, receipt, and internalization of the messages being conveyed through a particular climate. Together, work-group climate and the quality of leader-member exchange relationship are expected to affect state goal orientation. The nature of this interaction is detailed below. Because leaders create unique relationships with each of his/her employees, some members enjoy greater levels of access, communication, and interaction with their leader, while others interact with the leader strictly on the basis of their formal employment agreement (Dienesch & Liden, 1986; Graen & Cashman, 1975; Graen & Uhl-Bien, 1995; Liden et al, 1997). Members who perceive themselves as having a better quality
exchange relationship with their leader have an advantage in deciphering expectations of their role and performance as communicated by the work-group climate over members who do not have this same quality of relationship. This advantage has been documented as positively influencing performance (cf., Liden et al, 1997), reducing role conflict, and improving role clarity (Gerstner & Day, 1997). Research demonstrates that leaders and members with high quality LMX relationships are more likely to be in perceptual agreement regarding work issues and messages conveyed by the climate (Graen & Schiemann, 1978; Kozlowski & Doherty, 1989). Holding uniform expectations and perceptions is one crucial precondition to
58
situational strength (Mischel, 1977). Consequently, sharing similar perceptions with one's leader augments the existing situational strength provided by the work-group climate and supplies an additional vehicle through which one can further clarify expectations and heighten one's sense of obligation to conform to these expectations. In this case, group members with high quality LMX relationships are more likely to perceive an even stronger situation than those with low quality LMX relationships. In turn, the perceived augmented strength of the situation for members enjoying highquality LMX is expected produce greater clarity regarding which form of state goal orientation is favored and greater pressure to conform to the ascribed motivational state. This sense of obligation arises out of the convention of reciprocity established by the leader-member exchange relationship (Dienesch & Liden, 1986; Graen & Cashman, 1975; Liden et al, 1997; Sparrowe & Liden, 1997). Built on the notion from social exchange theory (Blau, 1964) that social exchanges between two parties "engender feelings of obligation, gratitude, and trust" (p. 94), LMX researchers have argued that members with higher quality relationships with their leaders are given greater amounts of resources, and in return, offer higher levels of loyalty, behavioral consistency with expectations, effort, performance, and positive work-related attitudes (Gerstner & Day, 1997; Graen & Cashman, 1975; Graen, Novak & Sommerkamp, 1982; Liden et al, 1997; Scandura & Schriesheim, 1994; Scott & Bruce, 1994; Tierney et al, 1999). Employing this logic, Hofmann et al (in press) argue that climate cues members as to which behaviors are emphasized and valued, and higher levels of LMX create a heightened sense of obligation to redefine their roles consistent with these cues. Their empirical findings support this theoretical contention that high-quality LMX relationships produce
59
greater levels of conformance with the expectations conveyed through the work-group climate as compared to lower-quality LMX. The work by Hofmann et al (in press) supports the logic endorsed here. It has been argued that leader-member exchange relationships provide a channel for further transmission, clarification, and internalization of cues conveyed by the work-group climate. In this way, the work-group climate signals the preferred motivational
orientation while the leader-member exchange relationship provide a means for reinforcing climate cues, clarifying expectations conveyed through the climate, and engendering a sense of obligation to adopt the state goal orientation ascribed by the climate. This logic suggests the following proposition and hypotheses: Proposition 4: Work-group climate and group member perceptions of their
leader-member exchange relationship interact to affect member state goal orientation. Hypothesis 4a: A climate for learning is more strongly related to state learning orientation when members perceive a higher-quality LMX relationship than when members perceive LMX to be of low quality. Hypothesis 4b: A climate for performance is more strongly related to state proving orientation when members perceive a higher-quality LMX relationship than when members perceive LMX to be of low quality. Hypothesis 4c: A climate for avoiding failure is more strongly related to state avoid orientation when members perceive a higher-quality LMX relationship than when members perceive LMX to be of low quality. To summarize, it is argued that leader priorities shape three prototypical workgroup climates: a climate for learning, a climate for performance, and a climate for
60
avoiding failure (Proposition 1). It is predicted that group members adopt a state goal orientation consistent with the types of behavior that are perceived to be rewarded, supported and expected, as communicated by the work-group climate (Proposition 2). Further, the type of state goal orientation will be differentially related to individual task performance, use of learning strategies, and defensive behaviors (Proposition 3). Finally, the effects of leader-member exchange are predicted to interact with those of work-group climate to further explain group members' adoption of the state goal orientation (Proposition 4).
61
CHAPTER 3 Research Methods To test the proposed relationships, a multi-phase, multi-method cross-sectional empirical study was conducted. The first phase utilized a qualitative approach to gather data to refine construct measurement and develop a measure of state goal orientation. The second phase involved development and measurement testing of the newly-created state goal orientation measure and is referred to here as the measurement study. The final stage involved surveying respondents via questionnaire in a field setting to test the hypothesized relationships. Each of these phases is discussed in more detail below. Because the findings from the first two phases informed the final stage of research, i.e., the field study, findings from the qualitative and measurement studies are reported here. Research Design: Phase I—Qualitative Stage The purpose of this stage was to gather data to (1) tailor existing measures to be used in the field study to the particular research setting (2) augment and refine the leadership and climate measures, and (3) develop a context-appropriate, valid state goal orientation measure. Sample and design. Six one-on-one interviews were conducted with managers from the regional retail bank that opted to participate in the final phase of this dissertation. This particular bank has a unique reporting structure —financial center managers (FCM) are responsible for overseeing the performance of the branch and directly supervise the personal banking staff (e.g., relationship bankers, loan officers). The area operations managers (AOM) manage the operational aspects of 3-5 different branches, have their office in only one of these branches, and directly supervise the tellers
62
and other administrative staff within those branches. Consequently, I conducted three interviews with FCMs and the other three with AOMs. I worked with senior managers to identify one FCM and one AOM that best represented each of the leadership styles of interest (i.e., places a priority on employee development, goal accomplishment, avoiding failure). Each interview lasted approximately 60-90 minutes and were used primarily to identify the most relevant performance measures for retail bank employees and a range of leader behaviors that might be appropriate for inclusion in the leader priority measures (see Appendix A for the interviewing protocol). In addition, six three-hour focus groups with 4-6 retail bank employees were conducted. Focus group participants were selected by senior managers to represent a diversity of job functions and company tenure. A completely crossed design was used with two primary experimental factors: (1) type of leader that focus group participants report to (e.g., learning oriented, performance-oriented, sensitive to failure) and (2) whether the AOM was co-located with the participant. This approach enabled me to determine the extent to which the physical proximity of the AOM influenced the core processes investigated here (it was found to have a minimal effect). The agenda for the focus groups included gathering data on climate dimensions and sample task situations that required unique motivational orientations, which ultimately provided the basis for the measure of state goal orientation (see Appendix B for the focus group protocol). Key findings. Several important findings emerged in the qualitative portion of this dissertation that guided subsequent data collection and analysis. First, there was high agreement among managers and employees that the Financial Center Manager/Leader is
63
responsible for setting the tone and priorities within the branch.2 The Area Operation Manager was often described as having a supporting role in this regard. This finding is important because it identified the "leader" in this particular research context. Second, focus group data on branch climate revealed that a strong, consistent infrastructure exists in this particular organization that standardizes training and rewards. Focus group
participants did still note that there was variation in the extent to which branch leaders make opportunities and rewards available to their employees. With respect to measure development, data collected during the interviews with the managers indicated that tellers and administrative personnel are assessed on different performance measures than members of the personal banking staff. As a result, using a broad based measure of individual-level performance seemed appropriate. Focus group data on the behaviors that leaders use to communicate their priorities validated Schein's (1992) work and suggested that what leaders pay attention to and measure and the type and focus of their coaching are particularly important in conveying implicit priorities in this particular research context. Lastly, output from the focus group contained a variety of task situations that employees in this organization routinely face. Three task situations were identified as highly relevant to all participants who would be involved in the final study of this dissertation, regardless of job class and branch—generating new business, managing difficult, irate clients, and participating in meetings. In addition, focus group respondents indicated that bank employees adopt a wide range of motivational orientations in these three task situations.
2
Financial Center Leaders differ from Financial Center Managers in that they supervise the operations of
more than one branch.
64
Research Design: Phase II—Measurement Study Sample. A total of 273 part-time MBA students participated in this study
(average response rate: 95.39%). Sixty-four percent of participants are White, 16% are Asian/Pacific Islander, 5% are Latino or Hispanic, 4% are African American, and 4% are of other ethnic backgrounds. Sixty-three percent of this sample is male, and on average, participants are 28 years old (SD=3.51 years). Participants have a wide variety of work experience (overall work experience: mean=5.49 years; SD=3.38) in various industries. Approximately, half of participants do not currently supervise others, 23% of respondents are supervisors, 14% are middle managers, and 3% are executives. Those who manage others have, on average, 1.94 years of experience (SD=2.61). Design. The design of the measurement study was a 3X3 between- and withinsubjects design. The between-subject factor was the achievement orientation of the work-group context (i.e., leader and climate) and was communicated to participants through a description of a hypothetical leader and work group (see Appendix C for experimental conditions). Repeated measures of the state goal orientation measure were captured by asking respondents to assess their motivational orientation under three different task situations—generating new business, handling difficult clients, and participating in meetings (i.e., within-subject factor). A situation-based approach to measure state goal orientation was employed to (1) allow me to test the potential effects of objective task demands on state goal orientation, (2) minimize measurement error by providing respondents the same set of task conditions, and (3) increase reliability by capturing repeated assessments.
65
Data Collection Procedures. Data were collected at two points in time. First, during their core HR course, 260 MBA students were asked to sign a consent form and complete a survey that collected various background (i.e., gender, ethnic background, age, managerial experience, total work experience) and personality measures (i.e., trait goal orientation, implicit theory of ability, core self evaluations, self monitoring). Of these 260 students, 253 provided data for a response rate of 97.24%. Approximately two weeks later, these same students were randomly assigned to one of three different experimental conditions and were asked about their state goal orientation in three task situations and the likelihood they would engage in a variety of behaviors in these situations (e.g., feedback seeking, self promotion, and avoiding blame). In addition, it
should be noted that the ordering of the task situations were randomized so as to remove any possible priming effects. A total of 262 students were asked to participate in the Time 2 data collection, and 244 provided data (response rate: 93.53%). Matching
responses from Time 1 and Time 2 yielded a total matched sample of 227. Surveys administered at Time 1 and 2 appear as Appendices D and E, respectively. To ensure that participants understood the experimental condition in which they participated, responses to the following three items were captured on a 5-point scale ("I responded to questions asked on this survey as if I worked for a boss who emphasizes?(a) the importance of continuous learning, (b) the importance of proving one's ability to others inside and outside the branch, and (c) the avoidance of committing and admitting mistakes"). One-way analysis of variance conducted on these
manipulation checks revealed significant differences across the three experimental conditions (learn check F=136.42, p