Description
The purpose of this study is to learn about data structures in relation to decisions about
family travel being influenced by sellers (i.e. agents), spouses/partners, spouse, kids and
friends/relatives (i.e. influencing units). Concern is with units affecting decisions about duration,
budget, destinations, restaurants, daily tours and accommodation facilities.
International Journal of Culture, Tourism and Hospitality Research
Who influences aspects of family decision making?
Metin Kozak Levent Karadag
Article information:
To cite this document:
Metin Kozak Levent Karadag, (2012),"Who influences aspects of family decision making?", International J ournal of Culture, Tourism and
Hospitality Research, Vol. 6 Iss 1 pp. 8 - 20
Permanent link to this document:
http://dx.doi.org/10.1108/17506181211206216
Downloaded on: 24 January 2016, At: 22:18 (PT)
References: this document contains references to 32 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1576 times since 2012*
Users who downloaded this article also downloaded:
Christina K.C. Lee, Sharon E. Beatty, (2002),"Family structure and influence in family decision making", J ournal of Consumer Marketing,
Vol. 19 Iss 1 pp. 24-41 http://dx.doi.org/10.1108/07363760210414934
Tendai Chikweche, J ohn Stanton, Richard Fletcher, (2012),"Family purchase decision making at the bottom of the pyramid", J ournal of
Consumer Marketing, Vol. 29 Iss 3 pp. 202-213 http://dx.doi.org/10.1108/07363761211221738
Hiral Chavda, Martin Haley, Chris Dunn, (2005),"Adolescents’ influence on family decision-making", Young Consumers, Vol. 6 Iss 3 pp.
68-78 http://dx.doi.org/10.1108/17473610510701223
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service information about
how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/
authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than
290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional
customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and
also works with Portico and the LOCKSS initiative for digital archive preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Who in?uences aspects of family decision
making?
Metin Kozak and Levent Karadag
Abstract
Purpose – The purpose of this study is to learn about data structures in relation to decisions about
family travel being in?uenced by sellers (i.e. agents), spouses/partners, spouse, kids and
friends/relatives (i.e. in?uencing units). Concern is with units affecting decisions about duration,
budget, destinations, restaurants, daily tours and accommodation facilities.
Design/methodology/approach – Research is on survey data collected from British family groups
visiting Turkey in the summer of 2007. Data allowing study of a range of matters were collected by
haphazard but not necessarily random selection of respondents in an airport departure lounge.
Findings – The study identi?es independent effects of in?uencing units on aspects of decision but also
shows that demographic factors affect aspects of decision making such as the choice of destination.
Methodologically, the study identi?es questionnaire structure and analysis methodology to use in
examining further research.
Practical implications – Results suggest how in?uencing units (e.g. self, spouse) affect aspects of
decision making aiding practitioners in understanding in?uences on decisions.
Originality/value – The research opens a new area of investigation by implementing a questionnaire
structure, developing analysis avenues and establishing results for further examination (i.e. to establish
their generality).
Keywords Family travel, Decision making, Units affecting choices, Turkey, Tourism management,
Family life, United Kingdom, Demography, In?uence
Paper type Research paper
Introduction
At each stage of the life cycle, people feel the need for traveling and realize this need in
some way. However, because of developments, innovations and globalization, such as
unrestricted (free) movement of knowledge, rising level of community welfare, increased
leisure time, etc., consumer behavior and decision-making in tourism are changing. From a
marketing perspective, the behavior of consumers has more meaning than just obtaining or
consuming a product (e.g. car, pencil or a meal). Consumer behavior in relation to such
matters as travel involves considering services and activities. Consumer behavior can
include dealing with how to use time effectively and why, where, how, how much and how
often to consume (Hoyer and MacInnis, 2007). In addition, Loudon and Della Bitta (1993)
address decision-making individually or by a group, followed by activity (e.g. going to a
destination or on a particular tour).
For tourism and other service businesses, consumer behavior becomes complicated to
study because of dynamics. For example, tourists will generally have expectations that are
evaluated in selecting a destination. Expectations may be modi?ed by marketing activities.
New information at any stage of destination choice can affect selecting a destination by
casting a destination in a new light (Kozak, 2006). A particular area of dynamic arises with
multiple people being involved in decisions (Decrop, 2006).
PAGE 8
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012, pp. 8-20, Q Emerald Group Publishing Limited, ISSN 1750-6182 DOI 10.1108/17506181211206216
Metin Kozak is a Professor
in the School of Tourismand
Hospitality Management,
Mugla University, Mugla,
Turkey. Levent Karadag is a
Lecturer in the Datca
Vocational School of
Tourism, Mugla University,
Mugla, Turkey
Received: September 2009
Revised: April 2010
Accepted: July 2010
The authors are indebted to Jay
Beaman for his constructive
comments on an earlier version
of this paper.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A group often involved in travel is the family. For family group travel many people can
in?uence travel decisions. Family members (e.g. spouses, children, in-laws) and non-family
(sellers, friends and relatives) are units potentially in?uencing a family’s purchasing
decisions. In this research unit is used in the context just introduced. Namely, units such as
spouses, sellers, etc. are considered.
Because of the importance of families for the viability of some destinations, the purposes of
this study are:
B to determine the in?uence of internal (e.g. family members) and external (e.g. sellers)
units on a family’s decisions of duration, budget, destinations, restaurants, daily tours and
accommodation facilities; and
B to investigate the association between decision making unit and overall satisfaction,
repeat visiting (e.g. loyalty) and word-of-mouth recommendation.
Decision-making in tourism
The decision-making process generally starts from the moment when some consumer
perceives a desire and continues until a ?nal evaluation stage when an assessment of travel
meeting needs occurs. For example, a family member thinking of going for a vacation, just
as an individual thinking of buying a newdress, forms the starting point of a decision-making
process (Hoyer and MacInnis, 2007). However, purchasing of tourism products or services
is the result of a complicated process (e.g. see Huan and Beaman, 2004, and its
references). This process involves a series of sub-decisions, for example where to go, how
to accommodate, how long to stay, and what operator to choose (Swarbrooke and Horner,
2001). In other words, a decision-making process for going on a vacation covers a process
incorporating different decisions (Zalatan, 1998; Blichfeldt, 2008).
Decision making does not just involve making a decision to go somewhere at a speci?c time.
Decisions are taken before and during traveling (Hyde, 2004). The decisions taken before a
vacation often involve the selection of accommodation facilities and travel agencies/tour
operators. Decisions taken during the vacation experience are composed of several
sub-decisions such as visiting special places, choosing restaurants, etc. (Barles et al.,
2007).
In?uence of family members
Over the years, the in?uence of spouses and children on decision making (e.g. purchases or
activities) has been examined (Shoham and Dalakas, 2005). Myers and Moncrief (1978)
appear to be initiators of studies of families’ decision-making processes. Early work is also
by others (Smith, 1979; Bartos, 1982; Nichols and Snepenger, 1988). One approach is
considering roles of family members. The family decision-making process involves the
interaction of family members. The interactions between parents and children can be critical
to a choice (Ndubisi, 2007). A study conducted by Wang et al. (2004) identi?es three types
of decision-making (decisions dominated by the male, by the female and joint decisions). In
other studies that have four different forms of decision-making the in?uence of children is
noted (Kaur and Singh, 2004, p. 26; Xia et al., 2006).
Until the 1950s, man-based decision-making was prominent. However, since women’s
status and rights have changed, often a focus is on joint decision-making (Thornton
et al., 1997). Such changes enhance the contribution of women to families’ purchasing
behavior (Lee and Beatty, 2002). In a study conducted relating to the purchasing
process of Israeli families, the common decision unit is the spouses. Among spouses,
the female spouses play the greatest role (Shoham and Dalakas, 2003). According to
the results of a study conducted in Canada and the US by Zalatan (1998), women seem
to have less effect on decision-making regarding ?nancial matters. However, they
apparently have the greatest effect on the selection of vacation destinations. On all
sub-decisions such as the selection of accommodation facilities and joining local tours,
joint decisions are common. In order to fully understand the complexity in the families’
decision-making process regarding purchasing, the respective roles of family members,
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 9
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
spouses and children need to be analyzed (Henthorne et al., 1997). Recognizing the
effect of children has drawn the attention of marketing experts and researchers
(Labreque and Ricard, 2001; Rose et al., 2002; Caruana and Vassallo, 2003; Chavda
et al., 2005; Flurry, 2007; Wang et al., 2007).
Hypotheses and research strategy
The literature makes clear that when families travel choices are made that are in?uenced by
different units (e.g. spouse, kids). Since statements about the way this research is done are
not explicit, H1 expresses general expectations based on the literature:
H1. For family trips, units (e.g. self, spouse/partner) contribute differently to different
aspects of travel decisions (e.g. choice of destination, accommodation and
budget).
The literature on decision making often considers demographics. H2 expresses the
reasonable expectation that demographics affect contributions to decisions. In other words,
for example, a male spouse/partner may tend to rate some contributions to some aspects of
travel decisions of self, spouse, etc. differently than the female spouse/partner:
H2. A respondent’s attributes affect how she/he rates contributions of Units to Aspects
of decision making.
Finally, in asking a person about contributions to travel decisions, a questionnaire’s structure
can lead to confusing results. A respondent can tend to mark high or low (ipsativity) leading
to a false pattern of correlations. Also, ratings around a mean near the center of a scale can
be more variable than near the top or bottom of a scale. Having the potential to have
relatively high variability without responses being highly skewed can give a false impression
of the importance of a variable. H3 suggests that the research has resulted in a
questionnaire and analysis approaches such that results are not biased by ipsativity or
differing variability on items:
H3. Unit-aspect questions as structured in the questionnaire for this research yield
ratings in which ipsative in?uences are not apparent and for which differing
variability between in?uence rating variables is appropriately controlled by
standardizing responses.
Analysis proceeds in several steps. First, evidence is given that allows seeing if patterns
exist in mean ratings of units for aspects of decision making. Because means do not
necessarily give information about how individuals structure responses, principal
components analysis is employed to determine structure in rating responses. Analysis
continues by examining variables affecting structure.
Methodology
The questionnaire for this research has three parts. The ?rst part is for collection of
information on satisfaction, on willingness to revisit and on word-of mouth recommendations.
These questions are not used except to provide information on numbers of past visits to the
destination area and to develop correlations showing potential for further research. The
research focuses on questions about family members, friends/relatives and sellers (travel
agents or tour sellers) participating in making vacation related decisions. Questions are
given in Figure 1. In the ?gure note that to answer consistently for a unit without just reacting
honestly, the question structure requires that a person remember, look back at a rating or
skip between sections for rating different aspects of travel decisions. The third part of the
questionnaire is for collection of respondents’ demographic information. Though analysis
only considers gender, the discussion pursues the need for and value of considering other
variables.
The population for this study is British families (at least one child) visiting the south-west part
of Turkey for the purpose of pleasure/summer vacation. With the cooperation of the airport
PAGE 10
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
authority, questionnaires were administered in the airport departure lounge of the
international airport operating in the province of Mugla, Turkey. Questionnaires were
administered to persons after tourists had completed their check-in procedures. No
systematic scheme was implemented to have random sampling. However, after an interview
of a member of a family ended, a new person was selected based on who was nearby.
Sampling occurred on different days of the week and at different times of the day in order to
obtain a sample that re?ects the diversity of visitors. Interviewing occurred on a total of eight
days. During those days, 1,400 persons were approached who were actually in a British
family that was traveling. A total of 379 questionnaires were collected in July and August of
2007 resulting in a response rate of 27 percent.
Figure 1 Introductory material and part 2 of the survey questionnaire
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 11
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Results and analysis
Data do not exist that give the percent of existing travelers that should have certain
attributes. However, in the data one has 211 interviews with males and 166 with females so
one is not too far off the 50/50 balance that might be expected. Information for other
demographics is given in Table I. One sees that respondents are spread over the wide range
of attributes that one sees in the area. As for the frequency of taking vacations, 18 percent of
the participants have not individually visited Turkey previously. Another conclusion that can
be mentioned related with this table is that around 80 percent of the participants have visited
Turkey before at least once as a family and decided to take a vacation in Turkey again.
One way of establishing H1 is to examine the pattern of rating regarding how units in?uence
aspects of travel decisions. Figure 2 provides an overview. The patterns for self,
spouse/partner, etc. show that for averages self (?rst bar in each group) is always rated
higher than spouse/partner, but not by much. Kids come in third. Their in?uence is quite high
for selecting the destination, accommodation, local tours and restaurants. However, as is to
be expected, kids in?uence drops substantially for duration, budget and agent/tour
provider. In?uence of kids drops down to the general level that occurs for friends/relatives
and for sellers (agents/tour providers).
Certain in?uences existing may appear obvious. If children are young, their in?uence pattern
should be different than for older children. If several visits to Turkey have occurred, the
in?uence of friends/relatives and sellers should decrease. Such in?uences are not being
investigated in this paper. Because a new approach to in?uence is being considered, the
focus is on general matters. Nevertheless, Table II provides evidence that demographic
variables affect ratings. Given males and females did not differ, male means would exceed
female means about 50 percent of the time. For 6 of 35 variables, the male mean is larger.
Based on the binomial distribution having males higher than 6 or fewer times has a
Table I Demographic pro?le of respondents
Variable No. %
Age
,24 5 1.1
25-34 30 7.6
35-44 185 49.5
45-54 128 35.3
55 . 25 6.4
Gender
Male 210 56.0
Female 163 44.0
Occupation
High school 113 30.3
Technical college 49 13.1
University 105 28.2
Postgraduate 106 28.4
As family
First time 45 11.9
Once 180 47.5
Twice 49 12.9
Three times 30 7.9
Four times 40 10.6
Five times or more 35 9.2
Individually
First time 66 18
Once 170 46.3
Twice 47 12.8
Three times 31 8.4
Four times 25 6.8
Five times or more 28 7.6
PAGE 12
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
probability less than 0.00001 (i.e. probability for 2-tailed test and n times male mean
exceeds female is Prob(j35*0.5 2 nj) $ (17.5 2 6) with (s < 3). Regardless, addressing
general matters refers to steps such as seeing if the 35 variables are related. Addressing
general matters also refers to dealing with values other than averages. Figure 1 and
regressions can be deceptive because a mean (e.g. 2.5) can re?ect half of respondents at
one extreme and half at the other (e.g. half at 0 and half at 5).
To consider individuals and relations between ratings made by an individual, a principal
components analysis discloses relations between in?uence ratings by an individual. An
ipsative relation (e.g. an individual tending to rate high) exists if some people rate high
across all items and some rated low (e.g. see Vaske and Beaman, 1995). If some people
tend to rate in?uences high and some low, in?uence rating variables will be correlated. A
second in?uence on correlations or covariances arises if the magnitude of ratings tends to
correlate with demographics. Another in?uence arises when some ratings tend to be near
the top or bottom of a scale and have relatively low variability compared to ratings
distributed around a central value. From Figure 1, one sees self and spouse ratings tend to
be high while friend/relatives and seller ratings are low to central. Principal components
Table II Mean aspect ratings for units by gender
Aspect
Unit Gender Go to Turkey Accommodation Local tours Restaurants Duration Budget Agent/operator
Self Male 5.7 5.6 5.5 5.6 6.0 6.0 5.5
Female 6.1 6.1 5.5 5.7 6.1 6.3 5.5
Spouse/partner Male 5.6 5.6 5.4 5.6 5.7 5.8 5.2
Female 6.0 5.8 5.3 5.4 5.9 6.0 5.1
Kids Male 4.9 4.8 4.8 4.9 4.6 3.7 3.5
Female 5.5 5.4 5.3 5.3 5.1 3.9 3.5
Friends/relatives Male 3.1 3.0 3.1 3.0 2.9 2.7 2.8
Female 3.4 3.1 2.8 3.0 3.0 2.8 2.8
Agent/tour provider Male 2.3 2.3 2.3 2.3 2.2 2.2 2.2
Female 3.0 3.0 2.7 2.9 2.7 2.7 2.6
Figure 2 Average mean scores of variables
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 13
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
analysis of the covariance matrix will re?ect differing variance while analysis of the
correlation matrix between in?uence variables is adjusted for differing variability. The point is
that principal components analysis of the covariance matrix is likely to yield different results
than analysis of the correlation matrix. However, both can be relevant to understanding
behavior.
Table III provides summary results for principal components analyses. Analysis is justi?ed by
the value of Kaiser-Meyer-Olkin (KMO) of 0.88 and a Bartlett spherosity chi-square of 10,486
for 595 degrees for freedom ( p , 0.0001) showing that principal components analysis is
acceptable. The table shows analysis of both correlation and covariance matrices for the 35
in?uence variables, yields ?ve strongest components that correspond exactly with unit
(e.g. self, spouse/partner). The alpha reliability of each of these exceeds 0.87 leading one to
expect the groups of variables (components) identify scales. For correlation, component
information is given on eight components while for the covariance matrix 7 components were
extracted. Interpretation of components 6 to 8 and 6 to 7 are not pursued here though
variance associated with these is nearly as large as for component 5. Not interpreting these
components is pursued in the discussion.
Several insights come from considering Table III. The covariance has greater total variance
to be explained and the variance is not equally distributed to variables. Therefore, unrotated
factor solutions seem quite different based on the unrotated sums of squared loadings.
However, as soon as variance explained is given as a percent the ?rst component has 32.8
percent of variance for the ?rst component for the correlation matrix and 35 percent for the
covariance matrix. Similarity in percents persists. Labels of Seller, etc. apply to rotated
factors. The reason for labeling is clear from Table IV. In Table IV the asterisks show high
loadings for various components in the analysis of the correlation matrix. For the ?rst
component, the one with the heaviest loading has a asterisks next to loadings for seller that
range from 0.75 to 0.82 values.
Similarly, after rotation, respective percents are 15.7 and 16.7; 13.1 and 15.4; 11.8 and 11.0;
etc. Highest other loadings are in the 0.3 to 0.45 range for friends/relatives. To some degree
people who rate sellers as important rate friends/relatives as important. The same description
given for sellers applies to friends/relatives with the variables interchanged. Also, variance
explained for seller and friends/relatives is, as expected, relatively high because means for
variables in these components are central (around 3 – see Figure 2). However, while based on
mean values one might expect kids to be the third component, spouse/partner is the third
component and close behind is self. This order is partly explained by females typically rating
higher than males (Table II). Nevertheless, to have substantial variance to explain, ratings are
Table III Variance explained for principal components of correlation and covariance matrices
Unrotated sums of
squared loadings
Rotation sum of squared
loadings
Matrix for analysis Component Total % Variance Total % Variance
Correlation (74.5% explained, 8 factors) Seller 11.5 32.8 5.5 15.7
Friends/relatives 5.6 16.1 4.6 13.1
Spouse/partner 2.3 6.7 4.1 11.8
Self 1.6 4.5 3.8 10.9
Kids 1.4 4.0 2.9 8.2
6.00 1.4 3.9 2.0 5.6
7.00 1.2 3.4 1.8 5.2
8.00 1.1 3.3 1.4 4.0
Covariance (72% explained, 7 factors) Seller 51.1 35.0 24.3 16.7
Friend/relative 22.1 15.1 22.5 15.4
Spouse/partner 7.7 5.3 16.1 11.0
Self 7.0 4.8 13.7 9.4
Kids 6.8 4.6 12.4 8.5
6 6.1 4.1 9.6 6.6
7 4.7 3.2 6.9 4.7
PAGE 14
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
skewed with most high but some quite low. Averaging across distributions, percents of
responses from1 (no in?uence) to 7 (highly in?uential) are 11 percent, 0.5, 0.5, 3.6, 9.5, 32 and
43 percent. High variability results from the 11 percent of responses that are 1 when the mean
is about 5.7. Anyway, for both respondent and spouse/partner one sees that the highest
loadings without an asterisk are for the kids and range from 0.11 to 0.32. Lowest loadings
being for setting budget and selecting the seller are reasonable. Analysis considering age of
children could show a positive relation to the age of children (not collected in this research).
Actually, for the kids component (5), the respondent indicates a relatively heavy in?uence of
kids on budget decisions. The value of 0.43 may because asking if the kids were important in
the budget decision resulted in some respondents indicating high importance for the wrong
reason. The response could re?ect what they wanted for the kids not the kids actually
in?uencing the decision by participating in decision making. The question about selecting an
agent is not subject to dual interpretations and has a loading of 0.18.
The principal components analysis results indicate that an issue of primary importance for
further research is ?nding components that correspond with units. Even though in?uence
rating responses for units on different aspects of decisions are in different blocks of the
questionnaire and respondents do not know about the mean for all respondents for each
variable, a person who is high or low compared to the mean for one aspect for a unit (e.g. for
selecting Turkey) tends to be high for all aspects for self. Still, since components are
orthogonal, being high for one unit (e.g. self) has no strong implication about being high for
Table IV Factor structure for Varimax rotation of the correlation matrix principal components
Component
Aspect 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
Turkey as destination Friends/relatives 0.35 0.75
a
0.04 0.06 0.09 20.01 20.03 20.04
Accommodation Friends/relatives 0.36 0.81
a
0.07 0.08 0.14 20.07 20.01 0.00
Participate in local tours Friends/relatives 0.41 0.69
a
0.08 0.01 0.05 0.10 0.31 0.04
Selection of restaurants Friends/relatives 0.36 0.78
a
0.14 0.11 0.06 0.03 0.05 0.15
Length of vacation Friends/relatives 0.31 0.79
a
0.04 0.05 0.13 0.19 20.04 0.05
Budget Friends/relatives 0.36 0.69
a
0.00 20.06 0.12 0.31 0.01 20.03
Select agency/tour operator Friends/relatives 0.44 0.60
a
0.11 0.01 0.02 0.41 0.07 20.01
Turkey as destination Kids 0.12 0.16 0.26 0.21 0.77
a
0.06 0.12 20.12
Accommodation Kids 0.20 0.07 0.25 0.31 0.74
a
0.07 0.15 0.08
Participate in local tours Kids 0.08 0.13 0.20 0.15 0.56
a
0.12 0.63 0.13
Selection of restaurants Kids 0.12 0.11 0.32 0.15 0.52
a
0.09 0.09 0.55
Length of vacation Kids 0.09 0.22 0.19 0.20 0.61
a
0.26 20.02 0.33
Budget Kids 0.19 0.24 0.11 0.08 0.43
a
0.61 20.05 0.11
Select agency/tour operator Kids 0.29 0.22 0.08 0.12 0.18
a
0.76 0.10 0.03
Turkey as destination Self 20.02 0.13 0.16 0.66
a
0.20 0.05 0.15 20.29
Accommodation Self 0.11 20.01 0.15 0.76
a
0.24 0.02 0.08 0.01
Participate in local tours Self 0.03 0.02 0.06 0.49
a
0.07 0.07 0.76 0.11
Selection of restaurants Self 0.06 0.06 0.09 0.61
a
0.13 0.03 0.22 0.58
Length of vacation Self 20.03 0.06 0.16 0.76
a
0.11 20.07 0.09 0.26
Budget Self 0.00 0.06 0.11 0.82
a
0.15 0.02 0.02 0.10
Select agency/tour operator Self 0.11 20.05 0.14 0.70
a
20.09 0.42 0.11 20.04
Turkey as destination Seller 0.75
a
0.31 0.10 0.07 0.06 20.01 20.01 20.03
Accommodation Seller 0.77
a
0.31 0.14 0.08 0.09 20.02 0.03 20.01
Participate in local tours Seller 0.78
a
0.28 0.03 0.03 0.10 0.04 0.19 0.07
Selection of restaurants Seller 0.82
a
0.29 0.05 0.06 0.09 0.03 0.02 0.11
Length of vacation Seller 0.81
a
0.26 0.04 0.03 0.08 0.13 0.01 0.08
Budget Seller 0.76
a
0.26 0.01 20.04 0.11 0.20 20.03 20.03
Select agency/tour operator Seller 0.76
a
0.20 0.04 0.01 0.01 0.33 0.00 20.03
Turkey as destination Spouse/partner 0.07 0.08 0.71
a
0.10 0.38 20.06 0.11 20.27
Accommodation Spouse/partner 0.13 20.01 0.72
a
0.19 0.29 20.03 0.07 0.00
Participate in local tours Spouse/partner 0.08 0.05 0.60
a
0.14 0.16 0.02 0.67 0.00
Selection of restaurants Spouse/partner 0.10 0.04 0.72
a
0.09 0.08 0.05 0.20 0.51
Length of vacation Spouse/partner 0.04 0.09 0.80
a
0.22 0.13 0.06 20.04 0.20
Budget Spouse/partner 0.00 0.11 0.80
a
0.08 0.11 0.09 0.01 0.05
Select agency/tour operator Spouse/partner 0.08 0.03 0.64
a
0.14 20.10 0.48 0.19 20.02
Note:
a
High loadings for various components in the analysis of the correlation matrix
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 15
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
other units. The word strong is used since, as noted, correlations up to 0.45 occur for
in?uences outside a block (e.g. in Table III see component 1 for friends/relatives). While a
relation exists within units, an ipsative in?uence of rating high across all units for all decision
aspects does not occur. Still, orthogonality at the aggregate level cannot be taken to show
that for subsegments orthogonality applies. As with randomization of plots in agricultural
experiments to control for fertility contours, having data across segments can hide structure.
Larger samples are needed to investigate the structure associated with segments.
Given results of principal components analysis, examination of effects such as gender
in?uencing rating can occur for components. However, Figure 3 shows the distributions of
component scores. Component scores have skewed distributions and have tails with
signi?cant mass. They certainly are not approximately normal. Distributions for sellers and
friends/relatives may be bimodal. Strangely enough, the spouse/partner distribution is quite
different from the self distribution which closely resembles the distribution for kids. Analysis
must take into account that distributions have irregular shapes.
Correlations can show principal component variables (e.g. component 1 labeled self) relate
to other variables collected in the survey. Tables V and VI present signi?cant correlations
between principal component scores and visit history, satisfaction, intentions of
recommending and intentions of revisiting. For self (component 1), no signi?cant relations
were found, though two correlations come close to being signi?cant at the 5 percent level
based on a 1-tail test (i.e. p ¼ 0.055). For spouse/partner signi?cant correlations are at the 5
percent level for past family visits and for satisfaction with the current visit. The only
signi?cant correlation for kids is for past family visits. Signi?cance is at the 1 percent level.
Friends/relatives has the most signi?cant relations and has two at the 1 percent level. Finally,
seller has two signi?cant correlations. Signi?cant correlations imply that regression could be
used to establish relations. In this research, the results are just given to show that structure
exists, which can be examined in future research.
Hypotheses and discussion
The ?ndings con?rmH1 in two different ways. Discussion of Figure 1 points out that on average
the units of respondent (self), spouse/partner, kids, friends/relatives and seller have different
Figure 3 Distributions of components from correlation matrix
PAGE 16
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
average rating patterns. The problemwith averages is that two quite different patterns can yield
an average pattern that is not appropriate for considering how to market, plan or manage a
destination. Still, Figure 1 shows that accepting H1 is reasonable given how this hypothesis is
stated. However, for understanding behavior of individuals rather than re?ecting on averages,
results of principal components show that variance in responses tends to be related to unit. A
person who responds above the mean in rating themselves, their spouse/partner, kids, etc.
tends to rate above the mean for all seven decision aspects for the given unit. However, a
respondent rating above the mean for self has no implication for rating for other units. A person
may rate high for self, low on in?uences on decisions for spouse/partner, high for kids and low
for friend/relatives and seller. Furthermore, as noted for Table II, females tend to rate high. The
point is that though a pattern of means exist, without knowing the pattern of means exists,
individuals have their own patterns of high and low responses associated with units. In other
words, H1 is con?rmed because, (1) means show patterns in responses that differ by unit and
(2) individuals show a pattern that relates to the pattern in means for units.
Analysis has eliminated the concern that the results of principal components analysis largely
re?ect the potential for variability in variables. By showing very similar structures for both
varimax solutions for the correlation and covariance matrices, the results show that
respondents tend to be most consistent in rating how sellers deviate from population means
for units and, least consistent for unit ratings for kids (i.e. for component 5). A possibility is
that variability in ratings for kids re?ect the age of children. A baby will not participate in
making a decision. While older children may be involved in budgeting decisions, future
research should clarify that a response regarding budget should distinguish between adults
making decisions because of the children without any participation by kids in choices. Some
thought may result in questions re?ecting how kids affect the budget.
Table V Correlations showing potential for further analysis
Previous times to
Turkey
Family satisfaction
with visits
Unit Measure Self Family Prior Current
Likelihood again
visit Turkey
Likelihood children
want to return
Likelihood spouse
recommends Turkey
Self Corr*** 0.06 20.04 20.08 20.08 20.04 20.03 20.08
Sig*** 0.129 0.203 0.146 0.055 0.242 0.269 0.055
Spouse/partner Corr*** 0.05 0.10* 0.05 20.09* 0.03 0.03 20.05
Sig*** 0.190 0.026 0.223 0.042 0.306 0.300 0.174
Kids Corr*** 0.08 0.13** 0.02 20.05 20.03 0.04 20.01
Sig*** 0.059 0.005 0.383 0.193 0.312 0.245 0.441
Friends/relatives Corr*** 0.07 0.08 0.04 0.05 0.16** 0.15** 0.09*
Sig*** 0.081 0.063 0.314 0.173 0.001 0.002 0.036
Seller Corr*** 0.04 0.08 0.13* 0.04 0.09* 0.04 0.08
Sig*** 0.243 0.053 0.033 0.232 0.046 0.197 0.062
Notes: *Signi?cant at 5 percent level for 1-tail test; **Signi?cant at 1 percent level for 1-tail test; ***Corr is correlation and Sig probability of
the correlation or a larger (+/2) value in its tail
Table VI Five correlations showing potential for further analysis
Previous times
to Turkey
Family satisfaction
with visits
Unit Self Family Prior Current
Likelihood again
visit Turkey
Likelihood children
want to return
Likelihood Spouse
re-commends Turkey
Self 0.06 20.04 20.08 20.08 20.04 20.03 20.08
Spouse/partner 0.05 0.10* 0.05 20.09* 0.03 0.03 20.05
Kids 0.08 0.13** 0.02 20.05 20.03 0.04 20.01
Friends/relatives 0.07 0.08 0.04 0.05 0.16** 0.15** 0.09*
Seller 0.04 0.08 0.13* 0.04 0.09* 0.04 0.08
Notes: *p , 0.05; **p , 0.001
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 17
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The results do not have a simple practical interpretation. However, research avenues are
evident. The analysis has not examined the data by segmentation to determine if, for
example, a segment exists of respondent (self) or spouse/partner members who dominate in
decision making. For spouses/partners in such a segment, the respondent may report
dissatisfaction by other travelers. If a seller or relative/friend is important in selecting a
destination (i.e. Turkey), is that unit rated as important on other decision aspects?
Presumably, the person presents information about costs, accommodation, etc. to in?uence
selection of a destination. However, what is the key element (e.g. cost) that motivates
selection of Turkey? When people or families have been to Turkey previously do Sellers and
Friends become less important with more trips (e.g. see Tables V and VI)? This research
suggests that developing a better understanding of decision making will require information
that may best be obtained by focused questioning based on some preliminary questions
about travel and decision making.
H2 relates to research avenues raised in the last paragraph. However, logically (by showing
one example), H2 can be accepted based on Table II showing that females rate high.
However, discussion in the previous signi?cant correlations in Tables Vand VI also show that
regression analysis of ratings relating to past travel, satisfaction and likelihood of
recommending will lead to signi?cant relations. Therefore, H2 can be accepted. Accepting
the hypothesis is implicitly an acknowledgement that this research has just opened the door
to considering the relations between Units and Aspects of decisions.
A person examining this research might, in fact, be concerned that results are artifacts of
data collection. H3 addresses this matter. If ipsativity, contrary to the hypothesis, was being
manifested by a person rating high or low, principal components would have detected a
correlation across all variables. The fact is that ?ve components correspond with units and
even other components detected do not cut across all variables and are around means
values respondents do not know exist. The patterns shows respondents are not merely
giving ratings without re?ecting on appropriate values. While principal components did not
detect a component showing a tendency to rate high or low, means for females being
greater than for males for all but six of 35 variables, shows an effect. Namely, women are
rating somewhat higher than men on average. However, are females rating higher because
some tend to see in?uence from a subordinate perspective, as suggested by the literature?
Rather than considering women as rating higher because they do, Table II is taken as
indicating that further research will show that reasons exist for somewhat higher average
ratings by females. Regarding ipsativity, H3 can be accepted.
H3 implying that differing variability in in?uence variables is not in?uencing results is
con?rmed. The obvious con?rmation is that analysis of the correlation matrix of in?uence
variables and of the covariance matrix yield virtually identical results in terms of components
and percents of variance explained by them. Appropriately standardized refers to analysis
resulting in components that rank appropriately in terms of variability to explain. The
occurrence of self and spouse/partner ahead of kids actually lead to examining distributions
and ?nding that self and spouse/partner distributions are heavily skewed with a high mean
but with enough lowratings to cause the distributions to have high variance. As discussion of
Figure 2 makes clear, future research cannot proceed based on the assumption of
approximate normality of ratings. A less obvious con?rmation of appropriate consideration of
variability is that components correspond with units. Given that respondents do not knowthe
means about with correlations will be computed and given that means have quite distinct
and variable patterns (Figure 1), the components corresponding with units shows an
unexpected regularity in behavior.
Conclusion and implications
This study has been carried out to investigate using survey research to learn about internal
and external units playing a role on sub-decisions related to vacations. The study reveals the
power of parents supporting the results of earlier studies (Carr, 2006; Wang et al., 2004;
Shoham and Dalakas, 2003). As well, the noticeable in?uence of children on various
sub-decisions supports the evidence of earlier studies showing that the children affect
PAGE 18
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
family decisions (Flurry, 2007; Gram, 2007; Chavda et al., 2005). However, given a
non-random sample of visitors to one country from another, results with implications for
further research are the important contribution or the research.
Research addresses mean scores and respondents having patterns related to those mean.
Note that respondents conform to a pattern without knowing what the means score pattern
was. From a practical perspective, a marketer who looks at means without knowing how
decisions related to deviations from those means is operating on a ?awed picture of
in?uencing decisions.
Yet, detailed research on data on vacation visits by British families to Turkey is not a good
basis for drawing conclusions that apply generally. Even speci?cs such as a weak
relationship between the in?uence ratings of aspects of decisions between kids and self,
and kids and spouse/partner should be subject to a broad examination. One may ?nd
con?rmation in other research (Chavda et al., 2005; Flurry, 2007; Gram, 2007). However, until
further research con?rms data structures found and relations between variables are pursued
with data for more origins and destinations, using results in practical work is based on the
assumption that future research will con?rm the ?ndings. As for researchers, this research
suggests both a questionnaire structure to use in research and suggests hypotheses. If
further research con?rms scale structures based on principal components, proof will exist
that ratings of in?uences deviate around averages in ways that were not known prior to this
research.
To summarize, this study contributes to the literature of consumer behavior by examining the
internal and external units in?uencing family decision making. However, conclusions based
on the research should be tempered by the study’s limitations. First, the questionnaire form
was applied only to the British families vacationing in Turkey. Also, one has a relatively small
non-random sample. Hence, one can suspect but not imply that similar results are
applicable to extended stay travelers frommany countries to Turkey, much less, to extended
stay vacationers to various countries. However, a reasonable consequence of this research
is pursuing the use of the type of questioning and analysis introduced in collecting data that
can con?rm or deny some of the ?ndings apply to a range of situations.
References
Barles, M.J., Bravo, R. and Fraj, E. (2007), ‘‘In?uence of women’s lifestyles on holiday decision: an
emprical study’’, paper presented at Advances in Tourism Marketing Conference, Valencia, September
10-12.
Bartos, R. (1982), ‘‘Women and travel’’, Journal of Travel Research, Vol. 20 No. 4, pp. 3-9.
Blichfeldt, B.S. (2008), ‘‘What to do on our holiday: the case of in situ decision making’’, Anatolia:
An International Journal of Tourism and Hospitality Research, Vol. 19 No. 2, pp. 287-305.
Carr, N. (2006), ‘‘A comparison of adolescents’ and parents’ holiday motivations and desires’’,
Tourism and Hospitality Research, Vol. 6 No. 2, pp. 129-42.
Caruana, A. and Vassallo, R. (2003), ‘‘Children’s perception of their in?uence over purchases: the role of
parental communication pattern’’, Journal of Consumer Marketing, Vol. 20 No. 1, pp. 55-66.
Chavda, H., Haley, M. and Dunn, C. (2005), ‘‘Adolescents’ in?uence on family decision-making’’,
Young Consumers, Vol. 2, pp. 68-78.
Decrop, A. (2006), Vacation Decision Making, CABI, Wallingford.
Flurry, A.L. (2007), ‘‘Children’s in?uence in family decision-making: examining the impact of the
changing American family’’, Journal of Business Research, Vol. 60, pp. 322-30.
Gram, M. (2007), ‘‘Children as co-decision makers in the family? The case of family holidays’’,
Young Consumers, Vol. 8 No. 1, pp. 198-228.
Henthorne, L.T., LaTour, S.M. and Hudson, W.T. (1997), ‘‘Japanese couples’ marital roles in stages of
product purchase decision-making’’, International Marketing Review, Vol. 14 No. 1, pp. 39-58.
Hoyer, D.W. and MacInnis, D.J. (2007), Consumer Behavior, Houghton Mif?in Company, New York, NY.
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 19
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Huan, T.C. and Beaman, J.G. (2004), ‘‘Contexts and dynamics of social interaction and information
search in decision-making for discretionary travel’’, Tourism Analysis, Vol. 8, pp. 177-82.
Hyde, K.F. (2004), ‘‘A duality in vacation decision making’’, Tourism Analysis, Vol. 8, pp. 183-6.
Kaur, P. and Singh, R. (2004), ‘‘Dynamics of purchase decision-making in family’’, South Asian Journal of
Management, Vol. 11 No. 4, pp. 26-41.
Kozak, M. (2006), ‘‘A content analysis of repeaters’ perceptions of tourist destinations’’, The Tourist
Review, Vol. 61 No. 1, pp. 22-6.
Labreque, J.A. and Ricard, L. (2001), ‘‘Children’s in?uence on family decision-making: a restaurant
study’’, Journal of Business Research, Vol. 54, pp. 173-6.
Lee, K-C.C. and Beatty, E.S. (2002), ‘‘Family structure and in?uence in family decision making’’,
Journal of Consumer Marketing, Vol. 19 No. 1, pp. 24-41.
Loudon, L.D. and Della Bitta, J.A. (1993), Consumer Behavior, McGraw Hill, New York, NY.
Myers, B.P. and Moncrief, W.L. (1978), ‘‘Differential leisure travel decision-making between spouses’’,
Annals of Tourism Research, Vol. 5 No. 1, pp. 157-65.
Ndubisi, O.N. (2007), ‘‘Impact of joint product usage and family structure on joint decision to purchase a
vacation by Malaysian spouses’’, Journal of Vacation Marketing, Vol. 13 No. 2, pp. 135-47.
Nichols, M.C. and Snepenger, J.D. (1988), ‘‘Family decision making and tourism behavior and
attitudes’’, Journal of Travel Research, Vol. 26 No. 4, pp. 2-6.
Rose, G., Boush, D. and Shoham, A. (2002), ‘‘Family communication and children’s purchasing
in?uence: a cross-national examination’’, Journal of Business Research, Vol. 55, pp. 867-73.
Shoham, A. and Dalakas, V. (2003), ‘‘Family consumer decision making in Israel: the role of teens and
parents’’, Journal of Consumer Marketing, Vol. 20 No. 3, pp. 238-51.
Shoham, A. and Dalakas, V. (2005), ‘‘He said, she said. . .they said: parents’ and children’s assessment
of children’s in?uence on family consumption decisions’’, The Journal of Consumer Marketing, Vol. 22
Nos 2/3, pp. 152-60.
Smith, V.L. (1979), ‘‘Women the taste-makers in tourism’’, Annals of Tourism Research, Vol. 6 No. 1,
pp. 49-60.
Swarbrooke, J. and Horner, S. (2001), Consumer Behaviour in Tourism, Heinemann, Oxford.
Thornton, R.P., Shaw, G. and Williams, M.A. (1997), ‘‘Tourist group holiday decision making and
behaviour: the in?uence of children’’, Tourism Management, Vol. 18 No. 5, pp. 287-97.
Vaske, J. and Beaman, J.G. (1995), ‘‘An ipsative clustering model for analyzing attitudinal data’’,
Journal of Leisure Research, Vol. 27 No. 2, pp. 168-91.
Wang, K.C., Chen, S.J. and Chou, S.H. (2007), ‘‘Senior tourists’ purchasing decisions in group package
tour’’, Anatolia: An International Journal of Tourism Research, Vol. 18 No. 1, pp. 23-42.
Wang, K.C., Hsieh, A.T., Yeh, Y.C. and Tsai, C.W. (2004), ‘‘Who is the decision-maker: the parents or the
child in group package tours?’’, Tourism Management, Vol. 25, pp. 183-94.
Xia, Y., Ahmed, Z.U., Ghingold, M., Hwa, N.K., Li, T.W. and Ying, W.T.C. (2006), ‘‘Spousal in?uence in
Singaporean family purchase decision-making process: a cross-cultural comparison’’, Asia Paci?c
Journal of Marketing and Logistics, Vol. 18 No. 3, pp. 201-23.
Zalatan, A. (1998), ‘‘Wives’ involvement in tourism decision process’’, Annals of Tourism Research,
Vol. 25 No. 4, pp. 890-903.
Corresponding author
Metin Kozak can be contacted at: [email protected]
PAGE 20
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Tzung-Cheng T.C. Huan. 2016. Seeing further: Honoring John Urry's contributions to tourism and hospitality research. Journal
of Business Research 69, 1228-1233. [CrossRef]
2. Xiaoxiao Fu, Xinran Lehto, Ounjoung Park. 2014. What Does Vacation do to our Family? Contrasting the Perspectives of Parents
and Children. Journal of Travel & Tourism Marketing 31, 461-475. [CrossRef]
3. Mathilda van Niekerk, Melville Saayman. 2013. The influences of tourism awareness on the travel patterns and career choices
of high school students in South Africa. Tourism Review 68:4, 19-33. [Abstract] [Full Text] [PDF]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_816447258.pdf
The purpose of this study is to learn about data structures in relation to decisions about
family travel being influenced by sellers (i.e. agents), spouses/partners, spouse, kids and
friends/relatives (i.e. influencing units). Concern is with units affecting decisions about duration,
budget, destinations, restaurants, daily tours and accommodation facilities.
International Journal of Culture, Tourism and Hospitality Research
Who influences aspects of family decision making?
Metin Kozak Levent Karadag
Article information:
To cite this document:
Metin Kozak Levent Karadag, (2012),"Who influences aspects of family decision making?", International J ournal of Culture, Tourism and
Hospitality Research, Vol. 6 Iss 1 pp. 8 - 20
Permanent link to this document:
http://dx.doi.org/10.1108/17506181211206216
Downloaded on: 24 January 2016, At: 22:18 (PT)
References: this document contains references to 32 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 1576 times since 2012*
Users who downloaded this article also downloaded:
Christina K.C. Lee, Sharon E. Beatty, (2002),"Family structure and influence in family decision making", J ournal of Consumer Marketing,
Vol. 19 Iss 1 pp. 24-41 http://dx.doi.org/10.1108/07363760210414934
Tendai Chikweche, J ohn Stanton, Richard Fletcher, (2012),"Family purchase decision making at the bottom of the pyramid", J ournal of
Consumer Marketing, Vol. 29 Iss 3 pp. 202-213 http://dx.doi.org/10.1108/07363761211221738
Hiral Chavda, Martin Haley, Chris Dunn, (2005),"Adolescents’ influence on family decision-making", Young Consumers, Vol. 6 Iss 3 pp.
68-78 http://dx.doi.org/10.1108/17473610510701223
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service information about
how to choose which publication to write for and submission guidelines are available for all. Please visit www.emeraldinsight.com/
authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of more than
290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online products and additional
customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and
also works with Portico and the LOCKSS initiative for digital archive preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Who in?uences aspects of family decision
making?
Metin Kozak and Levent Karadag
Abstract
Purpose – The purpose of this study is to learn about data structures in relation to decisions about
family travel being in?uenced by sellers (i.e. agents), spouses/partners, spouse, kids and
friends/relatives (i.e. in?uencing units). Concern is with units affecting decisions about duration,
budget, destinations, restaurants, daily tours and accommodation facilities.
Design/methodology/approach – Research is on survey data collected from British family groups
visiting Turkey in the summer of 2007. Data allowing study of a range of matters were collected by
haphazard but not necessarily random selection of respondents in an airport departure lounge.
Findings – The study identi?es independent effects of in?uencing units on aspects of decision but also
shows that demographic factors affect aspects of decision making such as the choice of destination.
Methodologically, the study identi?es questionnaire structure and analysis methodology to use in
examining further research.
Practical implications – Results suggest how in?uencing units (e.g. self, spouse) affect aspects of
decision making aiding practitioners in understanding in?uences on decisions.
Originality/value – The research opens a new area of investigation by implementing a questionnaire
structure, developing analysis avenues and establishing results for further examination (i.e. to establish
their generality).
Keywords Family travel, Decision making, Units affecting choices, Turkey, Tourism management,
Family life, United Kingdom, Demography, In?uence
Paper type Research paper
Introduction
At each stage of the life cycle, people feel the need for traveling and realize this need in
some way. However, because of developments, innovations and globalization, such as
unrestricted (free) movement of knowledge, rising level of community welfare, increased
leisure time, etc., consumer behavior and decision-making in tourism are changing. From a
marketing perspective, the behavior of consumers has more meaning than just obtaining or
consuming a product (e.g. car, pencil or a meal). Consumer behavior in relation to such
matters as travel involves considering services and activities. Consumer behavior can
include dealing with how to use time effectively and why, where, how, how much and how
often to consume (Hoyer and MacInnis, 2007). In addition, Loudon and Della Bitta (1993)
address decision-making individually or by a group, followed by activity (e.g. going to a
destination or on a particular tour).
For tourism and other service businesses, consumer behavior becomes complicated to
study because of dynamics. For example, tourists will generally have expectations that are
evaluated in selecting a destination. Expectations may be modi?ed by marketing activities.
New information at any stage of destination choice can affect selecting a destination by
casting a destination in a new light (Kozak, 2006). A particular area of dynamic arises with
multiple people being involved in decisions (Decrop, 2006).
PAGE 8
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012, pp. 8-20, Q Emerald Group Publishing Limited, ISSN 1750-6182 DOI 10.1108/17506181211206216
Metin Kozak is a Professor
in the School of Tourismand
Hospitality Management,
Mugla University, Mugla,
Turkey. Levent Karadag is a
Lecturer in the Datca
Vocational School of
Tourism, Mugla University,
Mugla, Turkey
Received: September 2009
Revised: April 2010
Accepted: July 2010
The authors are indebted to Jay
Beaman for his constructive
comments on an earlier version
of this paper.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A group often involved in travel is the family. For family group travel many people can
in?uence travel decisions. Family members (e.g. spouses, children, in-laws) and non-family
(sellers, friends and relatives) are units potentially in?uencing a family’s purchasing
decisions. In this research unit is used in the context just introduced. Namely, units such as
spouses, sellers, etc. are considered.
Because of the importance of families for the viability of some destinations, the purposes of
this study are:
B to determine the in?uence of internal (e.g. family members) and external (e.g. sellers)
units on a family’s decisions of duration, budget, destinations, restaurants, daily tours and
accommodation facilities; and
B to investigate the association between decision making unit and overall satisfaction,
repeat visiting (e.g. loyalty) and word-of-mouth recommendation.
Decision-making in tourism
The decision-making process generally starts from the moment when some consumer
perceives a desire and continues until a ?nal evaluation stage when an assessment of travel
meeting needs occurs. For example, a family member thinking of going for a vacation, just
as an individual thinking of buying a newdress, forms the starting point of a decision-making
process (Hoyer and MacInnis, 2007). However, purchasing of tourism products or services
is the result of a complicated process (e.g. see Huan and Beaman, 2004, and its
references). This process involves a series of sub-decisions, for example where to go, how
to accommodate, how long to stay, and what operator to choose (Swarbrooke and Horner,
2001). In other words, a decision-making process for going on a vacation covers a process
incorporating different decisions (Zalatan, 1998; Blichfeldt, 2008).
Decision making does not just involve making a decision to go somewhere at a speci?c time.
Decisions are taken before and during traveling (Hyde, 2004). The decisions taken before a
vacation often involve the selection of accommodation facilities and travel agencies/tour
operators. Decisions taken during the vacation experience are composed of several
sub-decisions such as visiting special places, choosing restaurants, etc. (Barles et al.,
2007).
In?uence of family members
Over the years, the in?uence of spouses and children on decision making (e.g. purchases or
activities) has been examined (Shoham and Dalakas, 2005). Myers and Moncrief (1978)
appear to be initiators of studies of families’ decision-making processes. Early work is also
by others (Smith, 1979; Bartos, 1982; Nichols and Snepenger, 1988). One approach is
considering roles of family members. The family decision-making process involves the
interaction of family members. The interactions between parents and children can be critical
to a choice (Ndubisi, 2007). A study conducted by Wang et al. (2004) identi?es three types
of decision-making (decisions dominated by the male, by the female and joint decisions). In
other studies that have four different forms of decision-making the in?uence of children is
noted (Kaur and Singh, 2004, p. 26; Xia et al., 2006).
Until the 1950s, man-based decision-making was prominent. However, since women’s
status and rights have changed, often a focus is on joint decision-making (Thornton
et al., 1997). Such changes enhance the contribution of women to families’ purchasing
behavior (Lee and Beatty, 2002). In a study conducted relating to the purchasing
process of Israeli families, the common decision unit is the spouses. Among spouses,
the female spouses play the greatest role (Shoham and Dalakas, 2003). According to
the results of a study conducted in Canada and the US by Zalatan (1998), women seem
to have less effect on decision-making regarding ?nancial matters. However, they
apparently have the greatest effect on the selection of vacation destinations. On all
sub-decisions such as the selection of accommodation facilities and joining local tours,
joint decisions are common. In order to fully understand the complexity in the families’
decision-making process regarding purchasing, the respective roles of family members,
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 9
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
spouses and children need to be analyzed (Henthorne et al., 1997). Recognizing the
effect of children has drawn the attention of marketing experts and researchers
(Labreque and Ricard, 2001; Rose et al., 2002; Caruana and Vassallo, 2003; Chavda
et al., 2005; Flurry, 2007; Wang et al., 2007).
Hypotheses and research strategy
The literature makes clear that when families travel choices are made that are in?uenced by
different units (e.g. spouse, kids). Since statements about the way this research is done are
not explicit, H1 expresses general expectations based on the literature:
H1. For family trips, units (e.g. self, spouse/partner) contribute differently to different
aspects of travel decisions (e.g. choice of destination, accommodation and
budget).
The literature on decision making often considers demographics. H2 expresses the
reasonable expectation that demographics affect contributions to decisions. In other words,
for example, a male spouse/partner may tend to rate some contributions to some aspects of
travel decisions of self, spouse, etc. differently than the female spouse/partner:
H2. A respondent’s attributes affect how she/he rates contributions of Units to Aspects
of decision making.
Finally, in asking a person about contributions to travel decisions, a questionnaire’s structure
can lead to confusing results. A respondent can tend to mark high or low (ipsativity) leading
to a false pattern of correlations. Also, ratings around a mean near the center of a scale can
be more variable than near the top or bottom of a scale. Having the potential to have
relatively high variability without responses being highly skewed can give a false impression
of the importance of a variable. H3 suggests that the research has resulted in a
questionnaire and analysis approaches such that results are not biased by ipsativity or
differing variability on items:
H3. Unit-aspect questions as structured in the questionnaire for this research yield
ratings in which ipsative in?uences are not apparent and for which differing
variability between in?uence rating variables is appropriately controlled by
standardizing responses.
Analysis proceeds in several steps. First, evidence is given that allows seeing if patterns
exist in mean ratings of units for aspects of decision making. Because means do not
necessarily give information about how individuals structure responses, principal
components analysis is employed to determine structure in rating responses. Analysis
continues by examining variables affecting structure.
Methodology
The questionnaire for this research has three parts. The ?rst part is for collection of
information on satisfaction, on willingness to revisit and on word-of mouth recommendations.
These questions are not used except to provide information on numbers of past visits to the
destination area and to develop correlations showing potential for further research. The
research focuses on questions about family members, friends/relatives and sellers (travel
agents or tour sellers) participating in making vacation related decisions. Questions are
given in Figure 1. In the ?gure note that to answer consistently for a unit without just reacting
honestly, the question structure requires that a person remember, look back at a rating or
skip between sections for rating different aspects of travel decisions. The third part of the
questionnaire is for collection of respondents’ demographic information. Though analysis
only considers gender, the discussion pursues the need for and value of considering other
variables.
The population for this study is British families (at least one child) visiting the south-west part
of Turkey for the purpose of pleasure/summer vacation. With the cooperation of the airport
PAGE 10
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
authority, questionnaires were administered in the airport departure lounge of the
international airport operating in the province of Mugla, Turkey. Questionnaires were
administered to persons after tourists had completed their check-in procedures. No
systematic scheme was implemented to have random sampling. However, after an interview
of a member of a family ended, a new person was selected based on who was nearby.
Sampling occurred on different days of the week and at different times of the day in order to
obtain a sample that re?ects the diversity of visitors. Interviewing occurred on a total of eight
days. During those days, 1,400 persons were approached who were actually in a British
family that was traveling. A total of 379 questionnaires were collected in July and August of
2007 resulting in a response rate of 27 percent.
Figure 1 Introductory material and part 2 of the survey questionnaire
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 11
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Results and analysis
Data do not exist that give the percent of existing travelers that should have certain
attributes. However, in the data one has 211 interviews with males and 166 with females so
one is not too far off the 50/50 balance that might be expected. Information for other
demographics is given in Table I. One sees that respondents are spread over the wide range
of attributes that one sees in the area. As for the frequency of taking vacations, 18 percent of
the participants have not individually visited Turkey previously. Another conclusion that can
be mentioned related with this table is that around 80 percent of the participants have visited
Turkey before at least once as a family and decided to take a vacation in Turkey again.
One way of establishing H1 is to examine the pattern of rating regarding how units in?uence
aspects of travel decisions. Figure 2 provides an overview. The patterns for self,
spouse/partner, etc. show that for averages self (?rst bar in each group) is always rated
higher than spouse/partner, but not by much. Kids come in third. Their in?uence is quite high
for selecting the destination, accommodation, local tours and restaurants. However, as is to
be expected, kids in?uence drops substantially for duration, budget and agent/tour
provider. In?uence of kids drops down to the general level that occurs for friends/relatives
and for sellers (agents/tour providers).
Certain in?uences existing may appear obvious. If children are young, their in?uence pattern
should be different than for older children. If several visits to Turkey have occurred, the
in?uence of friends/relatives and sellers should decrease. Such in?uences are not being
investigated in this paper. Because a new approach to in?uence is being considered, the
focus is on general matters. Nevertheless, Table II provides evidence that demographic
variables affect ratings. Given males and females did not differ, male means would exceed
female means about 50 percent of the time. For 6 of 35 variables, the male mean is larger.
Based on the binomial distribution having males higher than 6 or fewer times has a
Table I Demographic pro?le of respondents
Variable No. %
Age
,24 5 1.1
25-34 30 7.6
35-44 185 49.5
45-54 128 35.3
55 . 25 6.4
Gender
Male 210 56.0
Female 163 44.0
Occupation
High school 113 30.3
Technical college 49 13.1
University 105 28.2
Postgraduate 106 28.4
As family
First time 45 11.9
Once 180 47.5
Twice 49 12.9
Three times 30 7.9
Four times 40 10.6
Five times or more 35 9.2
Individually
First time 66 18
Once 170 46.3
Twice 47 12.8
Three times 31 8.4
Four times 25 6.8
Five times or more 28 7.6
PAGE 12
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
probability less than 0.00001 (i.e. probability for 2-tailed test and n times male mean
exceeds female is Prob(j35*0.5 2 nj) $ (17.5 2 6) with (s < 3). Regardless, addressing
general matters refers to steps such as seeing if the 35 variables are related. Addressing
general matters also refers to dealing with values other than averages. Figure 1 and
regressions can be deceptive because a mean (e.g. 2.5) can re?ect half of respondents at
one extreme and half at the other (e.g. half at 0 and half at 5).
To consider individuals and relations between ratings made by an individual, a principal
components analysis discloses relations between in?uence ratings by an individual. An
ipsative relation (e.g. an individual tending to rate high) exists if some people rate high
across all items and some rated low (e.g. see Vaske and Beaman, 1995). If some people
tend to rate in?uences high and some low, in?uence rating variables will be correlated. A
second in?uence on correlations or covariances arises if the magnitude of ratings tends to
correlate with demographics. Another in?uence arises when some ratings tend to be near
the top or bottom of a scale and have relatively low variability compared to ratings
distributed around a central value. From Figure 1, one sees self and spouse ratings tend to
be high while friend/relatives and seller ratings are low to central. Principal components
Table II Mean aspect ratings for units by gender
Aspect
Unit Gender Go to Turkey Accommodation Local tours Restaurants Duration Budget Agent/operator
Self Male 5.7 5.6 5.5 5.6 6.0 6.0 5.5
Female 6.1 6.1 5.5 5.7 6.1 6.3 5.5
Spouse/partner Male 5.6 5.6 5.4 5.6 5.7 5.8 5.2
Female 6.0 5.8 5.3 5.4 5.9 6.0 5.1
Kids Male 4.9 4.8 4.8 4.9 4.6 3.7 3.5
Female 5.5 5.4 5.3 5.3 5.1 3.9 3.5
Friends/relatives Male 3.1 3.0 3.1 3.0 2.9 2.7 2.8
Female 3.4 3.1 2.8 3.0 3.0 2.8 2.8
Agent/tour provider Male 2.3 2.3 2.3 2.3 2.2 2.2 2.2
Female 3.0 3.0 2.7 2.9 2.7 2.7 2.6
Figure 2 Average mean scores of variables
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 13
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
analysis of the covariance matrix will re?ect differing variance while analysis of the
correlation matrix between in?uence variables is adjusted for differing variability. The point is
that principal components analysis of the covariance matrix is likely to yield different results
than analysis of the correlation matrix. However, both can be relevant to understanding
behavior.
Table III provides summary results for principal components analyses. Analysis is justi?ed by
the value of Kaiser-Meyer-Olkin (KMO) of 0.88 and a Bartlett spherosity chi-square of 10,486
for 595 degrees for freedom ( p , 0.0001) showing that principal components analysis is
acceptable. The table shows analysis of both correlation and covariance matrices for the 35
in?uence variables, yields ?ve strongest components that correspond exactly with unit
(e.g. self, spouse/partner). The alpha reliability of each of these exceeds 0.87 leading one to
expect the groups of variables (components) identify scales. For correlation, component
information is given on eight components while for the covariance matrix 7 components were
extracted. Interpretation of components 6 to 8 and 6 to 7 are not pursued here though
variance associated with these is nearly as large as for component 5. Not interpreting these
components is pursued in the discussion.
Several insights come from considering Table III. The covariance has greater total variance
to be explained and the variance is not equally distributed to variables. Therefore, unrotated
factor solutions seem quite different based on the unrotated sums of squared loadings.
However, as soon as variance explained is given as a percent the ?rst component has 32.8
percent of variance for the ?rst component for the correlation matrix and 35 percent for the
covariance matrix. Similarity in percents persists. Labels of Seller, etc. apply to rotated
factors. The reason for labeling is clear from Table IV. In Table IV the asterisks show high
loadings for various components in the analysis of the correlation matrix. For the ?rst
component, the one with the heaviest loading has a asterisks next to loadings for seller that
range from 0.75 to 0.82 values.
Similarly, after rotation, respective percents are 15.7 and 16.7; 13.1 and 15.4; 11.8 and 11.0;
etc. Highest other loadings are in the 0.3 to 0.45 range for friends/relatives. To some degree
people who rate sellers as important rate friends/relatives as important. The same description
given for sellers applies to friends/relatives with the variables interchanged. Also, variance
explained for seller and friends/relatives is, as expected, relatively high because means for
variables in these components are central (around 3 – see Figure 2). However, while based on
mean values one might expect kids to be the third component, spouse/partner is the third
component and close behind is self. This order is partly explained by females typically rating
higher than males (Table II). Nevertheless, to have substantial variance to explain, ratings are
Table III Variance explained for principal components of correlation and covariance matrices
Unrotated sums of
squared loadings
Rotation sum of squared
loadings
Matrix for analysis Component Total % Variance Total % Variance
Correlation (74.5% explained, 8 factors) Seller 11.5 32.8 5.5 15.7
Friends/relatives 5.6 16.1 4.6 13.1
Spouse/partner 2.3 6.7 4.1 11.8
Self 1.6 4.5 3.8 10.9
Kids 1.4 4.0 2.9 8.2
6.00 1.4 3.9 2.0 5.6
7.00 1.2 3.4 1.8 5.2
8.00 1.1 3.3 1.4 4.0
Covariance (72% explained, 7 factors) Seller 51.1 35.0 24.3 16.7
Friend/relative 22.1 15.1 22.5 15.4
Spouse/partner 7.7 5.3 16.1 11.0
Self 7.0 4.8 13.7 9.4
Kids 6.8 4.6 12.4 8.5
6 6.1 4.1 9.6 6.6
7 4.7 3.2 6.9 4.7
PAGE 14
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
skewed with most high but some quite low. Averaging across distributions, percents of
responses from1 (no in?uence) to 7 (highly in?uential) are 11 percent, 0.5, 0.5, 3.6, 9.5, 32 and
43 percent. High variability results from the 11 percent of responses that are 1 when the mean
is about 5.7. Anyway, for both respondent and spouse/partner one sees that the highest
loadings without an asterisk are for the kids and range from 0.11 to 0.32. Lowest loadings
being for setting budget and selecting the seller are reasonable. Analysis considering age of
children could show a positive relation to the age of children (not collected in this research).
Actually, for the kids component (5), the respondent indicates a relatively heavy in?uence of
kids on budget decisions. The value of 0.43 may because asking if the kids were important in
the budget decision resulted in some respondents indicating high importance for the wrong
reason. The response could re?ect what they wanted for the kids not the kids actually
in?uencing the decision by participating in decision making. The question about selecting an
agent is not subject to dual interpretations and has a loading of 0.18.
The principal components analysis results indicate that an issue of primary importance for
further research is ?nding components that correspond with units. Even though in?uence
rating responses for units on different aspects of decisions are in different blocks of the
questionnaire and respondents do not know about the mean for all respondents for each
variable, a person who is high or low compared to the mean for one aspect for a unit (e.g. for
selecting Turkey) tends to be high for all aspects for self. Still, since components are
orthogonal, being high for one unit (e.g. self) has no strong implication about being high for
Table IV Factor structure for Varimax rotation of the correlation matrix principal components
Component
Aspect 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
Turkey as destination Friends/relatives 0.35 0.75
a
0.04 0.06 0.09 20.01 20.03 20.04
Accommodation Friends/relatives 0.36 0.81
a
0.07 0.08 0.14 20.07 20.01 0.00
Participate in local tours Friends/relatives 0.41 0.69
a
0.08 0.01 0.05 0.10 0.31 0.04
Selection of restaurants Friends/relatives 0.36 0.78
a
0.14 0.11 0.06 0.03 0.05 0.15
Length of vacation Friends/relatives 0.31 0.79
a
0.04 0.05 0.13 0.19 20.04 0.05
Budget Friends/relatives 0.36 0.69
a
0.00 20.06 0.12 0.31 0.01 20.03
Select agency/tour operator Friends/relatives 0.44 0.60
a
0.11 0.01 0.02 0.41 0.07 20.01
Turkey as destination Kids 0.12 0.16 0.26 0.21 0.77
a
0.06 0.12 20.12
Accommodation Kids 0.20 0.07 0.25 0.31 0.74
a
0.07 0.15 0.08
Participate in local tours Kids 0.08 0.13 0.20 0.15 0.56
a
0.12 0.63 0.13
Selection of restaurants Kids 0.12 0.11 0.32 0.15 0.52
a
0.09 0.09 0.55
Length of vacation Kids 0.09 0.22 0.19 0.20 0.61
a
0.26 20.02 0.33
Budget Kids 0.19 0.24 0.11 0.08 0.43
a
0.61 20.05 0.11
Select agency/tour operator Kids 0.29 0.22 0.08 0.12 0.18
a
0.76 0.10 0.03
Turkey as destination Self 20.02 0.13 0.16 0.66
a
0.20 0.05 0.15 20.29
Accommodation Self 0.11 20.01 0.15 0.76
a
0.24 0.02 0.08 0.01
Participate in local tours Self 0.03 0.02 0.06 0.49
a
0.07 0.07 0.76 0.11
Selection of restaurants Self 0.06 0.06 0.09 0.61
a
0.13 0.03 0.22 0.58
Length of vacation Self 20.03 0.06 0.16 0.76
a
0.11 20.07 0.09 0.26
Budget Self 0.00 0.06 0.11 0.82
a
0.15 0.02 0.02 0.10
Select agency/tour operator Self 0.11 20.05 0.14 0.70
a
20.09 0.42 0.11 20.04
Turkey as destination Seller 0.75
a
0.31 0.10 0.07 0.06 20.01 20.01 20.03
Accommodation Seller 0.77
a
0.31 0.14 0.08 0.09 20.02 0.03 20.01
Participate in local tours Seller 0.78
a
0.28 0.03 0.03 0.10 0.04 0.19 0.07
Selection of restaurants Seller 0.82
a
0.29 0.05 0.06 0.09 0.03 0.02 0.11
Length of vacation Seller 0.81
a
0.26 0.04 0.03 0.08 0.13 0.01 0.08
Budget Seller 0.76
a
0.26 0.01 20.04 0.11 0.20 20.03 20.03
Select agency/tour operator Seller 0.76
a
0.20 0.04 0.01 0.01 0.33 0.00 20.03
Turkey as destination Spouse/partner 0.07 0.08 0.71
a
0.10 0.38 20.06 0.11 20.27
Accommodation Spouse/partner 0.13 20.01 0.72
a
0.19 0.29 20.03 0.07 0.00
Participate in local tours Spouse/partner 0.08 0.05 0.60
a
0.14 0.16 0.02 0.67 0.00
Selection of restaurants Spouse/partner 0.10 0.04 0.72
a
0.09 0.08 0.05 0.20 0.51
Length of vacation Spouse/partner 0.04 0.09 0.80
a
0.22 0.13 0.06 20.04 0.20
Budget Spouse/partner 0.00 0.11 0.80
a
0.08 0.11 0.09 0.01 0.05
Select agency/tour operator Spouse/partner 0.08 0.03 0.64
a
0.14 20.10 0.48 0.19 20.02
Note:
a
High loadings for various components in the analysis of the correlation matrix
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 15
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
other units. The word strong is used since, as noted, correlations up to 0.45 occur for
in?uences outside a block (e.g. in Table III see component 1 for friends/relatives). While a
relation exists within units, an ipsative in?uence of rating high across all units for all decision
aspects does not occur. Still, orthogonality at the aggregate level cannot be taken to show
that for subsegments orthogonality applies. As with randomization of plots in agricultural
experiments to control for fertility contours, having data across segments can hide structure.
Larger samples are needed to investigate the structure associated with segments.
Given results of principal components analysis, examination of effects such as gender
in?uencing rating can occur for components. However, Figure 3 shows the distributions of
component scores. Component scores have skewed distributions and have tails with
signi?cant mass. They certainly are not approximately normal. Distributions for sellers and
friends/relatives may be bimodal. Strangely enough, the spouse/partner distribution is quite
different from the self distribution which closely resembles the distribution for kids. Analysis
must take into account that distributions have irregular shapes.
Correlations can show principal component variables (e.g. component 1 labeled self) relate
to other variables collected in the survey. Tables V and VI present signi?cant correlations
between principal component scores and visit history, satisfaction, intentions of
recommending and intentions of revisiting. For self (component 1), no signi?cant relations
were found, though two correlations come close to being signi?cant at the 5 percent level
based on a 1-tail test (i.e. p ¼ 0.055). For spouse/partner signi?cant correlations are at the 5
percent level for past family visits and for satisfaction with the current visit. The only
signi?cant correlation for kids is for past family visits. Signi?cance is at the 1 percent level.
Friends/relatives has the most signi?cant relations and has two at the 1 percent level. Finally,
seller has two signi?cant correlations. Signi?cant correlations imply that regression could be
used to establish relations. In this research, the results are just given to show that structure
exists, which can be examined in future research.
Hypotheses and discussion
The ?ndings con?rmH1 in two different ways. Discussion of Figure 1 points out that on average
the units of respondent (self), spouse/partner, kids, friends/relatives and seller have different
Figure 3 Distributions of components from correlation matrix
PAGE 16
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
average rating patterns. The problemwith averages is that two quite different patterns can yield
an average pattern that is not appropriate for considering how to market, plan or manage a
destination. Still, Figure 1 shows that accepting H1 is reasonable given how this hypothesis is
stated. However, for understanding behavior of individuals rather than re?ecting on averages,
results of principal components show that variance in responses tends to be related to unit. A
person who responds above the mean in rating themselves, their spouse/partner, kids, etc.
tends to rate above the mean for all seven decision aspects for the given unit. However, a
respondent rating above the mean for self has no implication for rating for other units. A person
may rate high for self, low on in?uences on decisions for spouse/partner, high for kids and low
for friend/relatives and seller. Furthermore, as noted for Table II, females tend to rate high. The
point is that though a pattern of means exist, without knowing the pattern of means exists,
individuals have their own patterns of high and low responses associated with units. In other
words, H1 is con?rmed because, (1) means show patterns in responses that differ by unit and
(2) individuals show a pattern that relates to the pattern in means for units.
Analysis has eliminated the concern that the results of principal components analysis largely
re?ect the potential for variability in variables. By showing very similar structures for both
varimax solutions for the correlation and covariance matrices, the results show that
respondents tend to be most consistent in rating how sellers deviate from population means
for units and, least consistent for unit ratings for kids (i.e. for component 5). A possibility is
that variability in ratings for kids re?ect the age of children. A baby will not participate in
making a decision. While older children may be involved in budgeting decisions, future
research should clarify that a response regarding budget should distinguish between adults
making decisions because of the children without any participation by kids in choices. Some
thought may result in questions re?ecting how kids affect the budget.
Table V Correlations showing potential for further analysis
Previous times to
Turkey
Family satisfaction
with visits
Unit Measure Self Family Prior Current
Likelihood again
visit Turkey
Likelihood children
want to return
Likelihood spouse
recommends Turkey
Self Corr*** 0.06 20.04 20.08 20.08 20.04 20.03 20.08
Sig*** 0.129 0.203 0.146 0.055 0.242 0.269 0.055
Spouse/partner Corr*** 0.05 0.10* 0.05 20.09* 0.03 0.03 20.05
Sig*** 0.190 0.026 0.223 0.042 0.306 0.300 0.174
Kids Corr*** 0.08 0.13** 0.02 20.05 20.03 0.04 20.01
Sig*** 0.059 0.005 0.383 0.193 0.312 0.245 0.441
Friends/relatives Corr*** 0.07 0.08 0.04 0.05 0.16** 0.15** 0.09*
Sig*** 0.081 0.063 0.314 0.173 0.001 0.002 0.036
Seller Corr*** 0.04 0.08 0.13* 0.04 0.09* 0.04 0.08
Sig*** 0.243 0.053 0.033 0.232 0.046 0.197 0.062
Notes: *Signi?cant at 5 percent level for 1-tail test; **Signi?cant at 1 percent level for 1-tail test; ***Corr is correlation and Sig probability of
the correlation or a larger (+/2) value in its tail
Table VI Five correlations showing potential for further analysis
Previous times
to Turkey
Family satisfaction
with visits
Unit Self Family Prior Current
Likelihood again
visit Turkey
Likelihood children
want to return
Likelihood Spouse
re-commends Turkey
Self 0.06 20.04 20.08 20.08 20.04 20.03 20.08
Spouse/partner 0.05 0.10* 0.05 20.09* 0.03 0.03 20.05
Kids 0.08 0.13** 0.02 20.05 20.03 0.04 20.01
Friends/relatives 0.07 0.08 0.04 0.05 0.16** 0.15** 0.09*
Seller 0.04 0.08 0.13* 0.04 0.09* 0.04 0.08
Notes: *p , 0.05; **p , 0.001
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 17
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The results do not have a simple practical interpretation. However, research avenues are
evident. The analysis has not examined the data by segmentation to determine if, for
example, a segment exists of respondent (self) or spouse/partner members who dominate in
decision making. For spouses/partners in such a segment, the respondent may report
dissatisfaction by other travelers. If a seller or relative/friend is important in selecting a
destination (i.e. Turkey), is that unit rated as important on other decision aspects?
Presumably, the person presents information about costs, accommodation, etc. to in?uence
selection of a destination. However, what is the key element (e.g. cost) that motivates
selection of Turkey? When people or families have been to Turkey previously do Sellers and
Friends become less important with more trips (e.g. see Tables V and VI)? This research
suggests that developing a better understanding of decision making will require information
that may best be obtained by focused questioning based on some preliminary questions
about travel and decision making.
H2 relates to research avenues raised in the last paragraph. However, logically (by showing
one example), H2 can be accepted based on Table II showing that females rate high.
However, discussion in the previous signi?cant correlations in Tables Vand VI also show that
regression analysis of ratings relating to past travel, satisfaction and likelihood of
recommending will lead to signi?cant relations. Therefore, H2 can be accepted. Accepting
the hypothesis is implicitly an acknowledgement that this research has just opened the door
to considering the relations between Units and Aspects of decisions.
A person examining this research might, in fact, be concerned that results are artifacts of
data collection. H3 addresses this matter. If ipsativity, contrary to the hypothesis, was being
manifested by a person rating high or low, principal components would have detected a
correlation across all variables. The fact is that ?ve components correspond with units and
even other components detected do not cut across all variables and are around means
values respondents do not know exist. The patterns shows respondents are not merely
giving ratings without re?ecting on appropriate values. While principal components did not
detect a component showing a tendency to rate high or low, means for females being
greater than for males for all but six of 35 variables, shows an effect. Namely, women are
rating somewhat higher than men on average. However, are females rating higher because
some tend to see in?uence from a subordinate perspective, as suggested by the literature?
Rather than considering women as rating higher because they do, Table II is taken as
indicating that further research will show that reasons exist for somewhat higher average
ratings by females. Regarding ipsativity, H3 can be accepted.
H3 implying that differing variability in in?uence variables is not in?uencing results is
con?rmed. The obvious con?rmation is that analysis of the correlation matrix of in?uence
variables and of the covariance matrix yield virtually identical results in terms of components
and percents of variance explained by them. Appropriately standardized refers to analysis
resulting in components that rank appropriately in terms of variability to explain. The
occurrence of self and spouse/partner ahead of kids actually lead to examining distributions
and ?nding that self and spouse/partner distributions are heavily skewed with a high mean
but with enough lowratings to cause the distributions to have high variance. As discussion of
Figure 2 makes clear, future research cannot proceed based on the assumption of
approximate normality of ratings. A less obvious con?rmation of appropriate consideration of
variability is that components correspond with units. Given that respondents do not knowthe
means about with correlations will be computed and given that means have quite distinct
and variable patterns (Figure 1), the components corresponding with units shows an
unexpected regularity in behavior.
Conclusion and implications
This study has been carried out to investigate using survey research to learn about internal
and external units playing a role on sub-decisions related to vacations. The study reveals the
power of parents supporting the results of earlier studies (Carr, 2006; Wang et al., 2004;
Shoham and Dalakas, 2003). As well, the noticeable in?uence of children on various
sub-decisions supports the evidence of earlier studies showing that the children affect
PAGE 18
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
family decisions (Flurry, 2007; Gram, 2007; Chavda et al., 2005). However, given a
non-random sample of visitors to one country from another, results with implications for
further research are the important contribution or the research.
Research addresses mean scores and respondents having patterns related to those mean.
Note that respondents conform to a pattern without knowing what the means score pattern
was. From a practical perspective, a marketer who looks at means without knowing how
decisions related to deviations from those means is operating on a ?awed picture of
in?uencing decisions.
Yet, detailed research on data on vacation visits by British families to Turkey is not a good
basis for drawing conclusions that apply generally. Even speci?cs such as a weak
relationship between the in?uence ratings of aspects of decisions between kids and self,
and kids and spouse/partner should be subject to a broad examination. One may ?nd
con?rmation in other research (Chavda et al., 2005; Flurry, 2007; Gram, 2007). However, until
further research con?rms data structures found and relations between variables are pursued
with data for more origins and destinations, using results in practical work is based on the
assumption that future research will con?rm the ?ndings. As for researchers, this research
suggests both a questionnaire structure to use in research and suggests hypotheses. If
further research con?rms scale structures based on principal components, proof will exist
that ratings of in?uences deviate around averages in ways that were not known prior to this
research.
To summarize, this study contributes to the literature of consumer behavior by examining the
internal and external units in?uencing family decision making. However, conclusions based
on the research should be tempered by the study’s limitations. First, the questionnaire form
was applied only to the British families vacationing in Turkey. Also, one has a relatively small
non-random sample. Hence, one can suspect but not imply that similar results are
applicable to extended stay travelers frommany countries to Turkey, much less, to extended
stay vacationers to various countries. However, a reasonable consequence of this research
is pursuing the use of the type of questioning and analysis introduced in collecting data that
can con?rm or deny some of the ?ndings apply to a range of situations.
References
Barles, M.J., Bravo, R. and Fraj, E. (2007), ‘‘In?uence of women’s lifestyles on holiday decision: an
emprical study’’, paper presented at Advances in Tourism Marketing Conference, Valencia, September
10-12.
Bartos, R. (1982), ‘‘Women and travel’’, Journal of Travel Research, Vol. 20 No. 4, pp. 3-9.
Blichfeldt, B.S. (2008), ‘‘What to do on our holiday: the case of in situ decision making’’, Anatolia:
An International Journal of Tourism and Hospitality Research, Vol. 19 No. 2, pp. 287-305.
Carr, N. (2006), ‘‘A comparison of adolescents’ and parents’ holiday motivations and desires’’,
Tourism and Hospitality Research, Vol. 6 No. 2, pp. 129-42.
Caruana, A. and Vassallo, R. (2003), ‘‘Children’s perception of their in?uence over purchases: the role of
parental communication pattern’’, Journal of Consumer Marketing, Vol. 20 No. 1, pp. 55-66.
Chavda, H., Haley, M. and Dunn, C. (2005), ‘‘Adolescents’ in?uence on family decision-making’’,
Young Consumers, Vol. 2, pp. 68-78.
Decrop, A. (2006), Vacation Decision Making, CABI, Wallingford.
Flurry, A.L. (2007), ‘‘Children’s in?uence in family decision-making: examining the impact of the
changing American family’’, Journal of Business Research, Vol. 60, pp. 322-30.
Gram, M. (2007), ‘‘Children as co-decision makers in the family? The case of family holidays’’,
Young Consumers, Vol. 8 No. 1, pp. 198-228.
Henthorne, L.T., LaTour, S.M. and Hudson, W.T. (1997), ‘‘Japanese couples’ marital roles in stages of
product purchase decision-making’’, International Marketing Review, Vol. 14 No. 1, pp. 39-58.
Hoyer, D.W. and MacInnis, D.J. (2007), Consumer Behavior, Houghton Mif?in Company, New York, NY.
VOL. 6 NO. 1 2012
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
PAGE 19
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Huan, T.C. and Beaman, J.G. (2004), ‘‘Contexts and dynamics of social interaction and information
search in decision-making for discretionary travel’’, Tourism Analysis, Vol. 8, pp. 177-82.
Hyde, K.F. (2004), ‘‘A duality in vacation decision making’’, Tourism Analysis, Vol. 8, pp. 183-6.
Kaur, P. and Singh, R. (2004), ‘‘Dynamics of purchase decision-making in family’’, South Asian Journal of
Management, Vol. 11 No. 4, pp. 26-41.
Kozak, M. (2006), ‘‘A content analysis of repeaters’ perceptions of tourist destinations’’, The Tourist
Review, Vol. 61 No. 1, pp. 22-6.
Labreque, J.A. and Ricard, L. (2001), ‘‘Children’s in?uence on family decision-making: a restaurant
study’’, Journal of Business Research, Vol. 54, pp. 173-6.
Lee, K-C.C. and Beatty, E.S. (2002), ‘‘Family structure and in?uence in family decision making’’,
Journal of Consumer Marketing, Vol. 19 No. 1, pp. 24-41.
Loudon, L.D. and Della Bitta, J.A. (1993), Consumer Behavior, McGraw Hill, New York, NY.
Myers, B.P. and Moncrief, W.L. (1978), ‘‘Differential leisure travel decision-making between spouses’’,
Annals of Tourism Research, Vol. 5 No. 1, pp. 157-65.
Ndubisi, O.N. (2007), ‘‘Impact of joint product usage and family structure on joint decision to purchase a
vacation by Malaysian spouses’’, Journal of Vacation Marketing, Vol. 13 No. 2, pp. 135-47.
Nichols, M.C. and Snepenger, J.D. (1988), ‘‘Family decision making and tourism behavior and
attitudes’’, Journal of Travel Research, Vol. 26 No. 4, pp. 2-6.
Rose, G., Boush, D. and Shoham, A. (2002), ‘‘Family communication and children’s purchasing
in?uence: a cross-national examination’’, Journal of Business Research, Vol. 55, pp. 867-73.
Shoham, A. and Dalakas, V. (2003), ‘‘Family consumer decision making in Israel: the role of teens and
parents’’, Journal of Consumer Marketing, Vol. 20 No. 3, pp. 238-51.
Shoham, A. and Dalakas, V. (2005), ‘‘He said, she said. . .they said: parents’ and children’s assessment
of children’s in?uence on family consumption decisions’’, The Journal of Consumer Marketing, Vol. 22
Nos 2/3, pp. 152-60.
Smith, V.L. (1979), ‘‘Women the taste-makers in tourism’’, Annals of Tourism Research, Vol. 6 No. 1,
pp. 49-60.
Swarbrooke, J. and Horner, S. (2001), Consumer Behaviour in Tourism, Heinemann, Oxford.
Thornton, R.P., Shaw, G. and Williams, M.A. (1997), ‘‘Tourist group holiday decision making and
behaviour: the in?uence of children’’, Tourism Management, Vol. 18 No. 5, pp. 287-97.
Vaske, J. and Beaman, J.G. (1995), ‘‘An ipsative clustering model for analyzing attitudinal data’’,
Journal of Leisure Research, Vol. 27 No. 2, pp. 168-91.
Wang, K.C., Chen, S.J. and Chou, S.H. (2007), ‘‘Senior tourists’ purchasing decisions in group package
tour’’, Anatolia: An International Journal of Tourism Research, Vol. 18 No. 1, pp. 23-42.
Wang, K.C., Hsieh, A.T., Yeh, Y.C. and Tsai, C.W. (2004), ‘‘Who is the decision-maker: the parents or the
child in group package tours?’’, Tourism Management, Vol. 25, pp. 183-94.
Xia, Y., Ahmed, Z.U., Ghingold, M., Hwa, N.K., Li, T.W. and Ying, W.T.C. (2006), ‘‘Spousal in?uence in
Singaporean family purchase decision-making process: a cross-cultural comparison’’, Asia Paci?c
Journal of Marketing and Logistics, Vol. 18 No. 3, pp. 201-23.
Zalatan, A. (1998), ‘‘Wives’ involvement in tourism decision process’’, Annals of Tourism Research,
Vol. 25 No. 4, pp. 890-903.
Corresponding author
Metin Kozak can be contacted at: [email protected]
PAGE 20
j
INTERNATIONAL JOURNAL OF CULTURE, TOURISM AND HOSPITALITY RESEARCH
j
VOL. 6 NO. 1 2012
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Tzung-Cheng T.C. Huan. 2016. Seeing further: Honoring John Urry's contributions to tourism and hospitality research. Journal
of Business Research 69, 1228-1233. [CrossRef]
2. Xiaoxiao Fu, Xinran Lehto, Ounjoung Park. 2014. What Does Vacation do to our Family? Contrasting the Perspectives of Parents
and Children. Journal of Travel & Tourism Marketing 31, 461-475. [CrossRef]
3. Mathilda van Niekerk, Melville Saayman. 2013. The influences of tourism awareness on the travel patterns and career choices
of high school students in South Africa. Tourism Review 68:4, 19-33. [Abstract] [Full Text] [PDF]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
2
:
1
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_816447258.pdf