Description
The purpose of this study is to consider the effect of an internship experience on tax
accountants’ professional performance
Accounting Research Journal
A survival analysis of tax professionals’ performance and internship experience
Carl R. Borgia Philip H. Siegel Dennis Ortiz
Article information:
To cite this document:
Carl R. Borgia Philip H. Siegel Dennis Ortiz , (2014),"A survival analysis of tax professionals’ performance
and internship experience", Accounting Research J ournal, Vol. 27 Iss 3 pp. 266 - 285
Permanent link to this document:
http://dx.doi.org/10.1108/ARJ -04-2013-0018
Downloaded on: 24 January 2016, At: 21:21 (PT)
References: this document contains references to 59 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 502 times since 2014*
Users who downloaded this article also downloaded:
Kim Watty, Satoshi Sugahara, Nadana Abayadeera, Luckmika Perera, J ade McKay, (2014),"Towards a
Global Model of Accounting Education", Accounting Research J ournal, Vol. 27 Iss 3 pp. 286-300 http://
dx.doi.org/10.1108/ARJ -08-2013-0054
Nicholas Apergis, Christina Christou, Christis Hassapis, (2014),"Accounting standards convergence
dynamics: International evidence from club convergence and clustering", Accounting Research J ournal, Vol.
27 Iss 3 pp. 226-248 http://dx.doi.org/10.1108/ARJ -06-2013-0031
Mirela Malin, (2014),"Enhancing lecture presentation through tablet technology", Accounting Research
J ournal, Vol. 27 Iss 3 pp. 212-225 http://dx.doi.org/10.1108/ARJ -09-2013-0069
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A survival analysis of tax
professionals’ performance and
internship experience
Carl R. Borgia
Florida Atlantic University, Boca Raton, Florida, USA
Philip H. Siegel
McCurry & Company CPAs, Boca Raton, Florida, USA, and
Dennis Ortiz
University of Texas, Pan American, Edinburg, Texas, USA
Abstract
Purpose – The purpose of this study is to consider the effect of an internship experience on tax
accountants’ professional performance.
Design/methodology/approach – It uses survival analysis, a dynamic methodology that allows for
more precise modeling than static traditional methods used to study promotion and turnover rates in
the past. The hypotheses were tested using a longitudinal database obtained from the human resource
departments of regional Certifed Public Accountant frms located in the southeastern and mid-south
areas of the USA.
Findings – Results were mixed. As in previous studies on the effects of internships on subsequent
professional performance, tax accounting professionals with a master’s degree and prior internship
experience had signifcantly faster promotion rates than those professionals with a master’s degree and
no internship experience. However, tax professionals with a master’s degree and prior internship
experience did not demonstrate a signifcant difference in turnover rate when compared to the
no-internship group.
Practical implications – This research provides evidence that students, employers and institutions
of higher education can use to guide them in their decisions regarding the effects of structured
internships on professional performance – in this case, the professional performance of tax accountants.
Originality/value – Previous research on tax professionals’ performance and internship experience
made use of static research methodologies. This study uses the more dynamic methodology of survival
analysis to see if different fndings result.
Keywords Taxation, Internships, Survival analysis
Paper type Research paper
Introduction
Evidence suggests that there is a positive relationship between internship experience as
a student and subsequent professional performance as a Certifed Public Accountant
(CPA), (Siegel and Rigby, 1988; Spiceland et al., 1992; Siegel and Rigby, 1998;
O’Shaughnessy and Naser-Tavakolian, 2000; Siegel et al., 2010). This is consistent with
conventional wisdom that a carefully managed internship program is an effective
component in the education process, and leads to more effective performance by
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1030-9616.htm
ARJ
27,3
266
Accounting Research Journal
Vol. 27 No. 3, 2014
pp. 266-285
© Emerald Group Publishing Limited
1030-9616
DOI 10.1108/ARJ-04-2013-0018
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
accountants in the workforce (Kolb et al., 2000; Cummings and Tataman, 2007; National
Association of Colleges and Educators, 2009, 2011).
Accounting research on the relationship between professional accounting
performance and internship experience uses traditional static methodology (Siegel and
Rigby, 1988; Siegel et al., 1992b) and focuses on auditors as subjects. Other research on
professional performance and internship experience uses traditional “static” statistical
methodology and focuses on tax professionals as subjects (Siegel et al., 2010). However,
Spiceland et al. (1992) are the only researchers to use survival analysis to study
subsequent accounting performance. The above research has been fairly consistent in
supporting the proposition that prior internship experience improves performance of
public accounting professionals.
Performance is measured in the literature in three ways: promotion rate or velocity
(Siegel et al., 1992b; Spiceland et al., 1992; Siegel et al., 2010), turnover rate (Spiceland
et al., 1992; Siegel et al., 2010) and performance evaluations (Siegel et al., 1992b, 2010).
This paper uses the frst two of these performance measures. Performance evaluations
were not made available, and are not considered in this study.
Prior research on tax professionals’ performance and internship experience makes
use of static research methodologies. Generally, these studies found that tax
professionals with master degrees and internship experience had better promotion and
turnover rates (Siegel et al., 1992b, 2010). This study uses survival analysis, a more
dynamic methodology which analyzes subjects as they enter, progress and leave the
frm as opposed to static studies which only use one data point. Multiple data points
allow the measurement of data to be verifed without misclassifcation, and therefore
lessening the chance of data bias.
Tax professionals were selected to see if results coincide with previous studies using
static and dynamic approaches. This study and prior studies in accounting show that
tax and auditing professionals with graduate degrees and internship experience have
better promotion rates. However, unlike in prior studies, no signifcant differences
were found in the present study on tax professional turnover rates. These fndings
indicate that future research into the turnover variable maybe warranted because of
these different fndings.
Thus, the current study further extends the use of survival analysis by applying it to
the study of tax professionals’ performance and its hypothesized association with
internship experience. This analysis will assist us in predicting the turnover and
promotion rate of tax professionals with internship experience.
The remainder of the paper is organized as follows. The frst section consists of a
literature review of the trends and history of internships; the benefts of internships to
students, employers and sponsors; the empirical fndings of the effect of internships on
subsequent performance; the cultural differences between auditing and taxation; and
the beneft of survival analysis. The second section presents the methodology, which
includes the hypotheses, sampling design, procedures and application of the research.
Finally, the third section presents and discusses the results of the survival analysis.
Literature review
To understand the contribution this study makes to the literature, this section provides
a review of fve research areas relevant to the development of this paper:
267
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
(1) the trend and history of internship requirements in public accounting;
(2) internships benefts to students, employers and sponsoring educational
institutions;
(3) empirical evidence of the effect of internships on subsequent performance;
(4) cultural differences between tax professionals and other public accounting
professionals; and
(5) benefts of survival analysis over traditional, static analysis.
Internship requirements in public accounting: trends and history
Over the years, two trends emerge regarding internship requirements in public
accounting. One trend shows increasing participation in internship programs. The other
indicates that internship programs are becoming more effectively managed and
structured.
The increasing participation of internship programs over the years are shown in
research studies from 1965 to the present. Lowe (1965) reports that 23 per cent of
institutions surveyed sponsored internship programs. Over 20 years later, Schmutte
(1986) reports 69 per cent of responding institutions offer student internships. This trend
clearly shows a sharp increase in sponsored programs during this 23-year period.
Beard (1998) reports a continuing increase in structured internship programs. For
example, most structured programs are fairly young, are for credit only, occur during
the junior year, are paid and require a written project report by the student. However,
Beard reports that most internship programs in accounting are not structured. They do
not have full- or part-time coordinators. Most do not require on-site visits, and share the
responsibility for identifying internship sites with the student and others.
Thompson (2011) reports the results of studies he performed during 1998 and 2003
indicating that internship opportunities are always available, regardless of institutional
support. Thompson essentially views institutional support of internships as a matter of
degree, ranging from none to full institutional support.
Other research show that internships programs better managed and structured over
the years. For example Beard (2007) reports that accounting internship program
administrators have established internship assessment plans. She cites Herring and
Izard (1992), DeMong et al. (1994), and Akers et al. (1997), who recommend that plans
include the identifcation of goals and objectives, employer guidelines for intern
supervision, establishment of internship requirements (a weekly diary, a fnal paper and
an oral presentation) and an assessment of the internship experience by the student and
internship supervisor.
Internships benefts to students, employers and sponsoring educational institutions
In 1986, Schmutte recalls the endorsements of internships by important accounting
organizations (AAA, 1952; AAA and AIA, 1955; AICPA, 1978). He summarizes the
advantages of internships to the participating student, the business employer and the
participating academic institution. Students beneft fromthe consolidation of classroom
experience with practice. They have the opportunity to return to school and correct
educational defciencies discovered during the internship, and the opportunity to clarify
career objectives before graduation. The business employer benefts from the qualifed
students serving as temporary workers during peak periods, the opportunity to assess
ARJ
27,3
268
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
company training and supervision functions and an enhanced reputation among
students. The sponsoring educational institutions beneft from better graduates and
interaction with CPA frms.
Student benefts. The Commission on Professional Accounting Education (1983)
indicates that an internship experience for accounting and tax professionals provides
performance benefts such as faster promotion and lower personnel turnover. These
benefts can help lower the high costs associated with recruitment, training and
replacing personnel (Cluskey and Vaux, 1997). Beard (2007), in her study of accounting
internships, cites employers’ anecdotal claims indicating that there are lower turnover
rates for employees who have participated in an internship. Other more formal research
reports that professionals with internships had a faster promotion rate than those
without an internship experience (Siegel and Rigby, 1988; Siegel et al., 1992b; Spiceland
et al., 1992; Siegel et al., 2010). Knechel and Snowball (1987) and English and Koeppen
(1993) show signifcant improvement in post-internship academic performance. Other
research indicates that accounting students who have internships adapt better to the
professional world.
Research fnds positive effects from internship experience on job interviewing and
the likelihood of a job offer (Pasewark et al., 1989; Feldman et al., 1998), on orientation
and career choice (Blank et al., 1991), on the ability to deal with real-world problems and
experiences; on confdence in the accounting profession (McCombs and VanSyckle,
1994) and on accounting knowledge base and motivation level (Beard, 1998). Research
also reports that internships have a positive impact on technical preparation (Johnstone
and Biggs, 1998), on development of categorical tax structures (Maletta et al., 1990), on
integration into the 150-hour requirement (Duncan and Schmutte, 2006) and on students’
job expectations (Broaddus, 2010).
Mentoring and student benefts. One of the benefts to students of a tax internship is
the possibility of being mentored by a more senior professional. Mentorships have been
shown to have a positive effect on subsequent professional performance and
development.
Viator and Scandura (1991) suggest that mentorship reduces turnover. Similarly,
Kleinman et al. (2001) indicate that mentoring can reduce turnover by imparting frm
values and goals. However, Hall and Smith (2009) indicate that mentoring may increase
turnover by increasing market skills and making the employee more valuable to other
employers. Kleinman et al. (2001) indicate that mentoring can ameliorate the high stress
levels in public accounting. Viator (1999) indicated that mentoring can reduce stress, but
success is conditional upon the mentoring design structure.
Accounting has a history of weak supervision of engagements generally not
encountered in other industries, which can often result in role ambiguity. However,
accordingly, Viator (2001) indicates that mentoring can have a role in reducing this
ambiguity.
Tax accountants often encounter job abstraction. Ostroff and Kozlowski (1993) argue
that mentors are able to reduce abstracting by focusing the younger professional on
knowledge that can be applied to other projects. Moreover, Viator and Scandura (1991)
indicate that mentoring provides opportunities to fnd signifcant assignments to
showcase the young professional’s skills.
Employer benefts. McCombs and VanSyckle (1994) report student internships can
reduce CPA frms’ recruiting and training costs. Internships provide the frm with a
269
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
non-obligatory look at prospective employees and to their training before hiring.
Internships also enable the CPA frm to provide for peak or seasonal personnel needs.
They help both current and prospective employers maintain visibility and professional
stature among accounting students. Beard (1998) indicates that internships help smaller
frms to establish a recruiting system and to link with universities and colleges.
Benefts to sponsoring educational institutions. Pasework et al. (1989) report that the
internship experience enhances the placement opportunities of tax accounting
graduates. Another beneft is reported by Beard (1998), who indicates that internship
programs can assist in the development of CPA frm support for accounting programs.
In addition, internships can provide feedback concerning the accounting curriculumand
help identify potential accounting advisory board members (Beard, 1998).
Empirical evidence of the effect of internships on subsequent performance
Internships are designed to enhance students’ subsequent academic and professional
performance. The extant literature provides us with feedback on these expectations.
Internships and other areas of performance. The internship experience provides
students with learning opportunities and frst-hand, experiential professional
knowledge and development. The extant literature shows that internships have positive
effects on subsequent academic performance. Beard (1998) indicates that the internship
experience makes the subsequent academic studies more meaningful, thereby helping
students become more successful in entering the market place. Other researchers fnd
post-internship students perform better in their auditing and accounting courses
(Knechel and Snowball, 1987), and have better academic performance (English and
Koeppen, 1993; O’Shaughnessy and Naser-Tavakolian, 2000).
Pasework et al. (1989) showprevious internship experience increases the likelihood of
getting an interview with a CPA frm. Beck and Halim (2008) fnd that a previous
internship experience helps students with personal and interpersonal skills in their
professional development, their frst job and their subsequent career.
This prior research supports the contention that internship experience can enhance
the students’ knowledge base and academic performance. The limitation of these studies
is that they do not investigate how post-internship students perform in the workplace.
Internships and subsequent professional performance. Internships have positive
effects on the subsequent professional performance of auditors. There has been limited
research on the post-graduate performance of students who have had accounting
internships because of a lack of data due to the diffculty in obtaining performance data.
However, Siegel and Rigby (1988) were able to followstudents after they graduated from
their degree program. They studied the students’ performance in public accounting and
found that internship students performed signifcantly better than those without an
internship experience.
Similar results were found using survival analysis methodology (Spiceland et al.,
1992). Both Siegel and Rigby (1988) and Spiceland et al. (1992) studies fnd that
professionals with internships had a faster promotion rate than those without an
internship experience. However, the survival analysis approach provides more accurate
results because it is not static. It takes into account the varying time element of
individuals entering and exiting the frm.
Siegel et al. (2010) studied the subsequent professional performance of tax
professionals with and without an internship experience. They fnd that internships
ARJ
27,3
270
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
play a signifcant role in the subsequent performance of tax professionals because they
perform better than those without a similar experience.
Culture differences in tax accountants
Siegel et al. (1997) note that, while the audit and tax functions exist within the same
public accounting frm, cultural differences exist between them. For example, while the
threat of litigation is common to both the audit and tax environments, there are three
fundamental differences:
(1) Role differences: The watchdog role of the auditor is quite different from the
advocate role of the tax professional.
(2) Sociological differences: The offce versus the feld nature of the respective jobs
may differentially impact the effectiveness of audit versus tax internships.
(3) Task differences: Audit personnel are frequently called on to exercise
professional judgment as compared with their tax counterparts, whose job tends
to be more technical in nature.
Given these differences between audit and tax work environments, and the attention prior
research has paid to the effect of internship experience on auditors, an important purpose of
the current study is to examine the effect of internship experience on tax professionals.
Earlier research (Benke and Rhodes, 1980) fnds similar sociological differences
between the team approach of auditors and the autonomous approach of the tax staff.
Also, Anderson-Gough et al. (1998) indicate that the sociological differences of these two
professional groups create separate subcultures.
Bertolini et al. (2010) study social interaction preferences of accounting professionals
working in US public accounting frms. They measure social interaction preferences,
fnding signifcant differences between tax and other public accounting professionals.
The results reveal an elevated need for social interaction among successful, experienced
tax professionals versus their non-tax counterparts. Specifcally, tax professionals
demonstrate a desire to be included and liked by others and seek common goals. This
unique set of social preferences leads us to theorize that internship experience – with its
socialization objective – should have a signifcant performance beneft for tax
professionals in public accounting frms.
Methodology
The current study examines tax professionals using a superior, dynamic methodology.
Somers and Birnbaum’s (1999) comparison of traditional turnover research with survival
analysis indicates signifcant divergence between the two methods, and concludes with a
call to consider survival methodology as a focus of future turnover studies.
Survival analysis incorporates a temporal, dynamic component for analyzing turnover
behavior. There are two methodological advantages of survival over traditional analysis.
First, there is less chance of misclassifying data, and therefore, the measurement method is
verifable. Subjects are classifed as they enter and exit the frm – a simple classifcation.
Second, there is less chance of data bias as a result of censored data.
Past survival analysis research in behavioral accounting has been limited to just two
studies. Spiceland et al. (1992) use educational background to address auditor promotion
and turnover. Their application permits more precise modeling of the entry and exit of
271
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the audit professionals in their study. Bröcheler et al. (2004) examine auditor human
capital, education and experience to predict frm survival.
Kolb indicates that internship experiences provide experiential learning to the
participant and provide a valuable work experience. These experiences are particularly
important for professions such as accounting that require technical and
problem-solving skills (Kolb et al., 2000).
The current paper focuses more on the question of when tax professionals are
promoted rather than why they are promoted. This is an important improvement over
traditional, static approaches, which more often address whether internship experience
is effective in improving performance.
Hypotheses
A comparison is developed for tax professionals with internship experience (I) and
without internship experience (NI). For this comparison, three hypotheses are
investigated:
H1A. S
I
(t) ?S
NI
(t), for all t.
H1B. M
I
(t) ?M
NI
(t), for all t.
H2. T
I
(t) ?T
NI
(t), for all t.
Where
S
i
(t) ? Probability that the promotion velocity to senior for subject t in
group i exceeds t.
M
i
(t) ? Probability that the promotion velocity to manger for subject t in
group i exceeds t.
T
i
(t) ?Probability that the turnover rate for subject t in group i exceeds t.
T ?0, and i ? I and NI (for the two internship experience groups, internship and
non-internship).
The relevant alternative hypotheses (H1A, H1Band H2) are that the survival curves for
tax professionals with and without internship experience are not uniformly the same.
Participants
This study incorporates a similar data gathering approach following prior internship
studies (Siegel and Rigby, 1988; Siegel et al., 2010). Ten regional offces of CPA frms were
contactedthat hadformal internships lastingbetween12and15weeks. Feldmanet al. (1998)
indicate that the most critical internship design characteristics include job autonomy, task
identity(theabilitytobeginandcompleteanassignment) andopportunitiestodevelopsocial
andcommunicationskills. Because researchindicates that entry-level workers are socialized
more quickly when they have more structured training (Blank et al., 1991), we included only
CPA frms that offered such training programs. This can insure uniformity of the frms’
input and output expectations for participants.
The frms were located in the southeastern and mid-south areas of the USA. Each
offce provided educational and professional performance data for tax professionals
who had an internship with the frm and subsequently joined the frm during the years
2001-2010. This time period was selected to allowenough time for the study participants
to be promoted to senior and manager levels. Similar information was obtained for tax
ARJ
27,3
272
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
personnel who did not have an internship experience for the same time period. Appendix
A1 provides a summary of participants.
Design and procedure
Interviews were conducted by one of the research authors working with the human
resource personnel fromeach of the participating offces. The human resource personnel
were instructed to select tax professionals who had a master’s degree in taxation. Those
tax professionals who had participated in the frms’ internship program and were
subsequently hired by the frm were identifed to divide the professionals into two
groups, internship and non-internship experience. The internship experience had to be
in the tax area. To control for prior work experience, individuals with accounting
experience prior to their employment with the frm or with signifcant work experience
in another functional area of the frm, such as auditing, were not included in the sample.
The information obtained fromthe public accounting frms’ personnel fles regarding
each of the tax professionals included the following:
• Name of college and graduation dates from both degrees.
• Number of months employed with the frm if terminated before end of study
period.
• CPA certifcation.
• Number of months lapsed before promotion to senior.
• Number of months lapsed before promotion to manager.
• Year started employment with the frm.
• Gender.
All of the frms used the designations senior and manager.
Internship participation enabled the division of the sample into two distinct groups:
those tax professionals who had an internship experience and those who did not
participate in an internship as a component of their educational programs. The sample
included 151 tax professionals who had an internship. The non-internship group had a
sample size of 121. Appendix A1 provides a profle of the study sample.
In this study, two measures of professional performance, promotion velocity and
turnover rate, are used to assess the relationship between performances, and prior tax
and no-prior tax internship experience. These variables have been tested and
successfully used in similar post-education research (Siegel and Rigby, 1988; Spiceland
et al., 1992; Siegel et al., 1992a; Siegel et al., 2010).
Promotion velocity to senior level in the frm is identifed as the number of months
lapsed from the date of initial employment at the frm until promotion to senior.
Similarly, promotion velocity to manager from senior can be defned as the number of
months lapsed from promotion to senior until the individual is promoted to manager.
Turnover velocity is defned as the number of months lapsed from the date
employment commences until employment terminates. No distinction was made
between voluntary and involuntary turnover. Within the groups, any bias associated
with voluntary versus involuntary turnover should be controlled by our selection
process. While it might seem that individuals with master’s degrees in tax would have
greater employment opportunities, in this case, all the sample participants held an
273
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
advanced tax degree. Thus, it is argued that differences in turnover rates, whether
voluntary or not, could be associated with an internship experience.
Survival analysis application
Survival analysis is a statistical method that has its foundation in medical research
wherein a terminal or fnal event occurs (often referred to as a failure) after a measured
length of time from an initial event. In medical research, detection of a disease is usually
the initial event, and death of the patient is the terminal event. In our study, the initial
events are hire or promotion to senior, and the terminal events are promotion to senior or
manager, and departure from the frm.
Survival analysis has been used in medical intervention studies for drug treatments
and effectiveness studies for disease prevention (Motzer et al., 1999; Pencina and
D’Agostino, 2004). Survival analysis represents an improvement over more
conventional statistical analysis when examining the relationship between education,
experience and professional performance (Spiceland et al., 1992). The application of
survival analysis is appropriate for research where subjects enter the study at varying
times and are observed for different lengths of time (longitudinal studies). Thus,
subjects can be followed from different time periods.
Moritaet al. (1989), andlater Somers andBirnbaum(1999) stronglyrecommendthe use of
survival analysis using turnover as a test variable. With longitudinal studies, some
observations are censored during the study because an individual may leave the study
before the event of interest occurs, or the study may conclude before the event of interest
occurs. For example, an individual who seeks employment elsewhere or is terminated prior
to promotion to senior is a censored observation for that event of interest. Furthermore,
right-censored data are data that are censored after the study concludes. It consists of
personnel who either did not leave or were not promoted. These individuals will not be
observable in the study because they are censored far after the end of the study observation
window. Therefore, longer-lived subjects are more likely to be censored over time.
Survival analysis requires explicit knowledge of three items: the initial event, a
measurement unit for the passage of time and a clear meaning of the terminal event (Cox
and Oaks, 1984). For promotion velocity, survival analysis is used in this study to
compare time in 12-month time intervals. There are two initial and two terminal events:
the subject’s initial employment (initial event) to their promotion to senior (terminal
event), and the time period frompromotion to senior status (initial event) to promotion to
manager (terminal event). Similarly, for turnover velocity, survival analysis is used to
compare the time in 12-month intervals from the subject’s initial employment, to their
termination of employment with the frm.
This approach enables the comparison of the survival experience of all subjects
participating during the study period. Survival analysis utilizes partial information
inferred from the fact that subjects enter the study at varying intervals. For some of
these subjects, the terminal event did not occur within the study period, but the potential
remains for the terminal event to occur at some later point in time. Traditional “static”
analysis, such as average time to promotion, fails to consider these observations,
resulting in less reliable fndings (Spiceland et al., 1992).
Survival analysis is also used to test for differences in turnover rates between the two
study groups; internship experience and non-internship experience. In this case, the
terminal event is a subject’s cessation of employment with the frm. Individuals leaving
ARJ
27,3
274
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the frm within the study period will have experienced the terminal event. Those
subjects still employed by the frm at the end of the study period will not have
experienced the terminal event. Individuals entering later in the study period may not
have experienced the terminal event due to the short-time period between
commencement of employment and the end of the study period. While a conventional
means test ignores these individuals, survival analysis utilizes information associated
with subjects entering and leaving at different times within a study period.
There are two steps involved in the survival analysis methodology. First, the
survival experience or survival distribution function (SDF) for the two groups of
subjects is estimated. The SDF describes the lifetimes of the population of interest (SAS
Institute, 2009). Second, survival curves are compared to test whether survival
experience is signifcantly different between the groups. The survival curve is estimated
by the construction of life tables. Alife table divides the sample into intervals (12-month
periods in this study) and identifes the number of individuals who are under study at
the beginning of each time interval.
The life tables show the time of entry for each subject to be zero. The conditional
probability of being promoted for each interval, or terminated as the case may be, can be
developed observing the number of individuals entering each promotion level, the
number who have been promoted, the number censored and the number of individuals
not promoted. The individuals who are terminated, who leave, or are promoted are
classifed as not surviving the interval. The estimated survival curve is the accumulated
product of these conditional probabilities.
Results
Promotion velocity
The Commission on Professional Accounting Educator (1983) indicated the tax
professional’s advancement rate to be a signifcant indicator of professional
performance. Therefore, if internship experience has a positive impact on a tax
professional’s preparation, individuals with internship experience should advance from
one level to another at a faster rate than those without such experience.
Table I shows the promotion velocity to senior. The tax professionals who had an
internship experience are in panel A; and the tax professionals who did not have an
internship experience are in panel B. In the sample, 96 individuals (64 per cent of
professional tax staff hires) were promoted in the internship experience group, while 59
(49 per cent) were promoted to senior in the non-internship group. Signifcant differences
in life tables were tested using the Wilcoxon statistic. The Wilcoxon statistic for
promotion velocity to senior is 8.4966 with one degree of freedom[1], and a probability
level (p value) of 0.0036. The survey analysis indicates that the promotion velocity to
senior was signifcantly faster for the internship group than for the non-internship
group. Thus, the null hypothesis, H1a, was rejected.
Survival curves for promotion velocity to senior are shown in Figure 1. The vertical
axis shows the cumulative probability of promotion to senior. The horizontal axis shows
the time interval in 12-month increments. The fgure clearly shows the promotion
velocity to senior for the internship group exceeds that of the non-internship group for
the frst 36 months. After 36 months, both groups tend to level out until the 60-month
mark when the cumulative probability of promotion to senior of the non-internship
group actually surpasses their internship experienced counterparts
275
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table II shows the promotion velocity from senior to manager. The internship group is
shown in panel A, and the non-internship group is in panel B. In the sample, 42
individuals (44 per cent of those promoted to senior) in the internship group were
promoted to manager. However, only nine (10 per cent of those promoted to senior) in the
non-internship group were promoted to manager. The Wilcoxon statistic for promotion
velocity to manager is 71.4700 with one degree of freedom. The probability level (p
value) is 0.0001. These fndings strongly indicate that there is a signifcant difference
between the two groups in promotion velocity to manager. Thus, the null hypothesis,
H1b, was also rejected. We tested for gender differences and found no signifcant
differences between the genders. Using survival analysis, we also tested for signifcant
Table I.
Life tables – promotion
velocity to senior
Interval start
(months) No. entering
a
No.
censored
b
No.
exposed
to risk
c
No.
promoted
Proportion
promoted
d
Cumulative
probability
e
Standard
error
f
Panel A – Internship Group (N ? 151)
0 151 0 151 9 0.0596 0.0596 0.0000
12 142 18 133 51 0.3835 0.4202 0.0600
24 73 21 62 33 0.5323 0.7288 0.0720
36 19 12 13 3 0.2308 0.7914 0.0718
48 4 4 2 0 0.0000 0.7914 0.0790
55 96
Panel B – Non-internship Group (N ? 121)
0 121 1 120.5 0 0.0000 0.0000 0.0000
12 120 28 106.0 32 0.3019 0.3019 0.0030
24 60 15 52.5 25 0.4762 0.6343 0.0209
36 20 2 19.0 1 0.0526 0.6536 0.0246
48 18 9 13.5 1 0.0740 0.6792 0.0256
60 8 4 6.0 0 0.0000 0.6792 0.0305
72 2 1 1.5 1 0.6667 0.8931 0.0305
60 60
g
Notes:
a
Subjects still employed at the start of a particular interval (year) and not having been
promoted in a prior period – for example, in panel A, the number entering the interval starting after 12
months (142) is the number entering the preceding interval (151) minus those censored (0) or promoted
(9) during the preceding interval;
b
subjects lost to follow-up during the interval because of leaving the
frm prior to experiencing the terminal event or not yet experiencing the terminal event at the data
collection date;
c
the average number of subjects exposed to the possibility of being promoted during a
particular interval, that is, the number entering that interval minus one-half the number censored that
interval;
d
the number promoted divided by the number exposed – an estimate of the conditional
probability of being promoted during the interval (i.e., the probability of being promoted in the interval,
conditional on being with the frm and not being promoted previously;
e
the unconditional probability
of being promoted up to and including the current interval; calculated as one minus the accumulated
product of the conditional probabilities of not being promoted in all previous years–for example, in
panel A, he cumulative probability of being promoted after 36 months (0.7914) is the complement of the
accumulated product of the complements of the proportions promoted in each interval up to and
including the current interval: 0.7914 ? 1 – (1 ? 0.0596)(1 ? 0.3835)(1 ? 0.5323)(1 ? 0.2308);
f
the
standard error of the cumulative probability estimates;
g
panel B number censored plus promoted
should sum to 121, The difference of one is most likely due to rounding
ARJ
27,3
276
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
differences between the two groups with respect to promotion velocity from entry-level
to manger. As expected, the internship group was promoted signifcantly faster than the
non-internship group. The Wilcoxon statistic is 64.6332 with one degree of freedom and
a p value of 0.0001.
The Wilcoxon Mann–Whitney test (Siegel, 1956) is one of the most powerful of the
non-parametric tests for comparing two populations. It is used to test the null hypothesis
that two populations have identical distribution functions against the alternative
hypothesis that the two distribution functions differ only with respect to location
(median).
Survival curves for promotion velocity to manager are shown in Figure 2. The
vertical axis shows the cumulative probability of promotion to manager. The horizontal
axis shows the time interval in 12-month increments. The fgure shows the promotion
velocity to manager for the internship group, once again, clearly exceeds that of the
non-internship group for the frst 36 months. After 36 months, the internship group has
reached a plateau, while their non-internship counterparts continue to climb until the
60-month mark. At 60 months, both groups plateau with the cumulative probability of
promotion to manager for the internship group exceeding that of the non-internship
group.
Turnover rate
Turnover rate is an important measure of an accounting professional’s performance
(Spiceland et al., 1992; Dalton et al., 1997; Hall and Smith, 2009; Siegel et al., 2010). Few
studies, however, use actual accounting frmturnover data (Spiceland et al., 1992; Dalton
et al., 1997; Siegel et al., 2010). Following the survival technique of Spiceland et al. (1992),
turnover event is defned as the cessation of employment.
Table III shows the turnover rate for the two groups. The tax professionals who had an
internship experience are in panel A; and the tax professionals who did not have an
internship experience are in panel B. In the sample, 56 individuals (37 per cent)
terminated employment in the internship experience group, while 43 (36 per cent)
terminated employment in the non-internship group. The Wilcoxon statistic for
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 12 24 36 48 60 72
C
u
m
u
l
a
t
i
v
e
P
r
o
b
a
b
i
l
i
t
y
Months
Internship
Non-Internship
Figure 1.
Survival curve-promotion
velocity to senior
277
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
turnover rate is 2.5517 with one degree of freedom, and a probability level (p value) of
0.1102. Using only the last 60 months (thus, dropping the frst 24 months of data) for the
analysis produces a Wilcoxon value of 2.72, with a marginally signifcant p value of
0.0993. The survey analysis indicates that the turnover rate was not signifcantly faster
for the internship group versus that for the non-internship group. Thus, the null
hypothesis, H2, was not rejected.
Survival curves for turnover rate are shown in Figure 3. The vertical axis shows
the cumulative probability of exiting the frm. The horizontal axis shows the time
Table II.
Life tables – promotion
velocity, senior to
manager
Interval start
(months)
No.
entering
a
No.
censored
b
No.
exposed
to risk
c
No.
promoted
Proportion
promoted
d
Cumulative
probability
e
Standard
error
f
Panel A – Internship Group (N ? 96)
0 96 0 96.0 0 0.0000 0.0000 0.0000
12 96 3 94.5 0 0.0000 0.0000 0.0000
24 93 6 90.0 18 0.2000 0.2000 0.0000
36 69 18 60.0 24 0.4000 0.5200 0.0730
??48 27 15 19.5 0 0.0000 0.5200 0.0098
60 12 9 7.5 0 0.0000 0.5200 0.0098
72 3 3 1.5 0 0.0000 0.5200 0.0098
54 42
Panel B – Non-internship Group (N ? 62)
0 62 0 302.0 0 0.0000 0.0000 0.0000
12 62 1 61.5 0 0.0000 0.0000 0.0000
24 61 15 53.5 0 0.0000 0.0000 0.0000
36 46 21 35.5 3 0.0845 0.0845 0.0038
48 22 10 17.0 3 0.1765 0.2461 0.0206
60 9 4 7.0 2 0.2857 0.4615 0.0449
72 3 2 2.0 0 0.0000 0.4615 0.0718
53 8
g
Notes:
a
Subjects still employed at the start of a particular interval (year) and not having been
promoted in a prior period – for example, in panel A, the number entering the interval starting after 12
months (96) is the number entering the preceding interval (96) minus those censored (0) or promoted (0)
during the preceding interval;
b
subjects lost to follow-up during the interval because of leaving the
frm prior to experiencing the terminal event or not yet experiencing the terminal event at the data
collection date;
c
the average number of subjects exposed to the possibility of being promoted during a
particular interval, that is, the number entering that interval minus one-half the number censored that
interval;
d
the number promoted divided by the number exposed – an estimate of the conditional
probability of being promoted during the interval (i.e., the probability of being promoted in the interval,
conditional on being with the frm and not being promoted previously;
e
the unconditional probability
of being promoted up to and including the current interval; calculated as one minus the accumulated
product of the conditional probabilities of not being promoted in all previous years–for example, in
panel A, he cumulative probability of being promoted after 36 months (0.5200) is the complement of the
accumulated product of the complements of the proportions promoted in each interval up to and
including the current interval: 0.5200 ? 1 – (1 ? 0.0000)(1 ? 0.0000)(1 ? 0.2000)(1 ? 0.4000);
f
the
standard error of the cumulative probability estimates;
g
panel B number censored plus promoted
should sum to 62, The difference of one is most likely due to rounding
ARJ
27,3
278
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
interval in 12-month increments. The fgure shows that after the initial 24-month
period, the turnover rate for the non-internship group exceeds that of the internship
group. As the Wilcoxon test indicates, however, the difference is not statistically
signifcant.
Discussion and conclusion
This study considered the effect of an internship experience on tax accountants’
professional performance. It used survival analysis to study promotion (frst to senior
and then from senior to manager) and turnover rates (accountant’s survival in the frm),
which are important measures of tax accountants’ professional performance. The
hypotheses were tested using a longitudinal database obtained from the human
resource departments of regional CPA frms located in the southeastern and mid-south
areas of the USA.
In this study, the internship group was promoted at a signifcantly faster rate
than the non-internship group, both to senior and manager levels. These results are
generally consistent with prior research in auditing (Spiceland et al., 1992) using
survival analysis. The results also are consistent with traditional statistical
analysis of internships for both tax and auditing professionals (Siegel et al., 2010,
and Siegel and Rigby, 1988), which support the predicted positive correlation
between internship experience and subsequent professional performance.
Understanding this result should have important implications for those planning
careers as tax professionals in public accounting. The advancement rate fndings
are also an important indication to both educators and CPA frms of the value of a
well-planned and well-managed internship program.
Intuitively, it would appear that individuals with the same educational background
plus professional internship experience would be promoted earlier than those
individuals with similar backgrounds but without internship experience. Also, prior
research on tax accountant performance shows this to be true (Siegel et al., 1992a,
2010). However, there are key distinctions between these studies and the present
research. First, this study has increased control over the sample selection, resulting
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0 12 24 36 48 60 72
C
u
m
u
l
a
t
i
v
e
P
r
o
b
a
b
i
l
i
t
y
Months
Internship
Non-Internship
Figure 2.
Survival curve-promotion
velocity to manager
279
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
in a more homogeneous sample. Second, it uses a statistical methodology that
improves the testing process by accounting for participants entering and exiting the
sample at different time periods. Somers and Birnbaum (1999) would have
researchers also statistically compare the results of traditional versus survival
analysis. This could further improve our understanding of survival analysis’
contribution to the role internships play in employee performance.
The present study reports no signifcant differences in turnover rates between
the two tax accountant groups. These results are at variance with prior internship
studies using analysis of variance (Siegel and Rigby, 1988; Siegel et al., 2010) and
Table III.
Life tables – turnover rate
Interval atart
(months)
No.
entering
a
No.
censored
b
No.
exposed
to risk
c
No.
leaving
d
Proportion
leaving
Cumulative
probability
e
Standard
error
f
Panel A – Internship Group (N ? 151)
0 151 0 151.0 0 0.0000 0.0000 0.0000
12 151 0 151.0 11 0.0728 0.0728 0.0000
24 130 0 130.0 14 0.1077 0.1727 0.0491
36 106 1 105.5 22 0.2085 0.3452 0.0652
48 78 3 76.5 9 0.1176 0.4222 0.0716
60 66 7 62.5 0 0.0000 0.4222 0.0721
72 59 42 38.0 0 0.0000 0.4222 0.0721
84 42 42 21.0 0 0.0000 0.4222 0.0721
95 56
Panel B – Non-internship Group (N ? 121)
0 121 0 121.0 1 0.0083 0.0083 0.0000
12 120 23 108.5 5 0.0461 0.0540 0.0031
24 92 19 82.5 14 0.1697 0.2146 0.0102
36 59 14 52.0 16 0.3077 0.4562 0.0203
48 29 9 24.5 5 0.2041 0.5672 0.0274
60 15 7 11.5 2 0.1739 0.6425 0.0303
72 6 6 3.0 0 0.0000 0.6425 0.0340
78 43
Notes:
a
Subjects still employed at the start of a particular interval (year) and not terminating
employment in a prior period – for example, in panel A, the number entering the interval starting after
12 months (151) is the number entering the preceding interval (151) minus those censored (0) or
terminating (9) during the preceding interval;
b
subjects lost to follow-up during the interval because of
not yet terminating at the data collection date;
c
the average number of subjects exposed to the
possibility of terminating during a particular interval, that is, the number entering that interval minus
one-half the number censored that interval;
d
the number terminated divided by the number exposed –
an estimate of the conditional probability of terminating during the interval (i.e., the probability of
terminating in the interval, conditional on not terminating previously;
e
the unconditional probability
of terminating up to and including the current interval; calculated as one minus the accumulated
product of the conditional probabilities of not terminating in all previous years – for example, in panel
A, the cumulative probability of terminating after 36 months (0.3452) is the complement of the
accumulated product of the complements of the proportions terminating in each interval up to and
including the current interval: 0.3452 ? 1 – (1 ? 0.0000)(1 ? 0.0728)(1 ? 0.1077)(1 ? 0.2085);
f
the
standard error of the cumulative probability estimates
ARJ
27,3
280
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
using survival analysis (Spiceland et al., 1992). These studies reported signifcant
differences in turnover rates between internship and non-internship groups. As
discussed earlier, perhaps the study results would be different if the analysis were
extended to a longer time period. Also, those tax accountants with internships are
more likely to be in greater demand due to their experience and perhaps leave the
frm for a more lucrative position. Hagen and Siegel (1995) argue that contrasting
results have also been found in other turnover studies (Wright, 1988; Siegel et al.,
1988), suggesting that other factors, such as socialization, outside employment
opportunities and different stress levels, may infuence the retention rate. Future
research might address these conficting results by controlling or specifcally
considering these other factors.
The association between personality types and professional performance should
also be examined. Research has shown that stress levels and voluntary and
involuntary turnover often relate to personality types and exogenous factors such as
mergers, time constraints and management changes (Satava, 1996; Choo, 1995).
These studies report that stress level is a signifcant variable relating to turnover.
Several limitations and recommendations regarding our results should be
considered. As noted earlier, a positive correlation between internship experience
and promotion velocity does not prove a causal effect. Other factors, such as those
discussed above, may play an integral role in the promotion and retention of tax
professionals. For example, the internship group may be more highly motivated
toward professional work and/or have different personality traits that are conducive
to professional tax careers. These individuals may be more aggressive and mature
than their non-internship counterparts. To the extent that these characteristics are
factors in promotion velocity and frm retention, the performance of tax accountants
with internship experience might be upwardly biased.
Our sample was drawn for a specifed geographic area and from regional frms.
Therefore, the results may not be generalizable to other regions and other types of
frms, such as large international or small local CPA frms. Changing market
conditions within the profession and student time constraints may also infuence the
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 12 24 36 48 60 72 84
C
u
m
u
l
a
t
i
v
e
P
r
o
b
a
b
i
l
i
t
y
Months
Internship
Non-Internship
Figure 3.
Survival curve – turnover
rate
281
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
nature and success of internship experiences. The present study contributes to our
understanding of the effect of an internship on tax professionals’ performance, and
shows the potential of survival analysis as an appropriate method for shedding
further light on this topic.
Note
1. Degrees of freedom (df) are calculated as: df ?2 groups (internship and non-internship) – 1.
References
Akers, M., Giacomino, D. and Trebby, J. (1997), “Designing and implementing an accounting
assessment program”, Issues in Accounting Education, Vol. 12 No. 2, pp. 259-280.
American Accounting Association (1952), “Report of the committee on internship programs”, The
Accounting Review, Vol. 27 No. 3, pp. 361-323.
American Accounting Association and American Institute of Accountants Committee on Faculty
Residency and Internship Programs and Committee on Accounting Personnel (1955),
“Statement of standards and responsibilities under public accounting internship
programs”, The Accounting Review, Vol. 30 No. 2, pp. 206-210.
American Institute of Certifed Public Accountants, Task Force on the Committee on Education
and Experience Requirements for CPAs (1978), Education Requirements for Entry into the
Accounting Profession, AICPA, New York, NY.
Anderson-Gough, F., Grey, C. and Robson, K. (1988), Making up Accountants: the Organizational
and Professional Socialization of Trainee Chartered Accountants, Ashgate, London.
Beard, D.F. (1998), “The status of internships/cooperative education experiences in accounting
education”, Journal of Accounting Education, Vol. 16 No. 3, pp. 507-516.
Beard, D.F. (2007), “Assessment of internship experiences and accounting core competencies”,
Accounting Education, Vol. 16 No. 2, pp. 207-220.
Beck, E.B. and Halim, H. (2008), “Undergraduate internships in accounting: what and how do
Singapore interns learn fromexperience?”, Accounting Education, Vol. 17 No. 2, pp. 151-172.
Benke, R.L. and Rhodes, J.G. (1980), “The job satisfaction of higher level employees in large
certifed public accounting frms”, Accounting Organizations and Society, Vol. 5 No. 2,
pp. 187-201.
Bertolini, M., Borgia, C. and Siegel, P.H. (2010), “The social skill preferences of tax professionals in
CPA frms: a FIRO-B analysis”, The Journal of Applied Business Research, Vol. 26 No. 2,
pp. 105-114.
Blank, M.B., Siegel, P.H. and Rigsby, J.T. (1991), “Determinants of international CPA frm
orientation among accounting students”, The British Accounting Review, Vol. 23 No. 4,
pp. 281-300.
Broaddus, B. (2010), “Graduate accounting internships: counting on a transformation”, Baylor
Business Review, Vol. 28 No. 2, pp. 24-27.
Bröcheler, V., Maijoor, S. andVanWitteloostuijn, A. (2004), “Auditor humancapital andaudit frm
survival: the Dutch audit industry in 1930–1992”, Accounting, Organizations and Society,
Vol. 29 No. 7, pp. 627-646.
Choo, F. (1995), “Auditors’ judgment performance under stress: a test of the predicted relationship
by three theoretical models”, Journal of Accounting, Auditing & Finance, Vol. 10 No. 3,
p. 611.
Cluskey, G.R. and Vaux, A.C. (1997), “Is seasonal stress a career choice of professional
accountants?”, Journal of Employment Counseling, Vol. 34 No. 1, pp. 7-19.
ARJ
27,3
282
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Commission on Professional Accounting Educator (1983), A Post-baccalaureate Education
Requirement for the CPA Profession, National Association of State Boards of Accounting,
New York, NY.
Cox, D.R. and Oaks, D. (1984), Analysis of Survival Data, Chapman and Hall, London.
Cummings, R. and Tataman, W. (2007), “Accounting internships: a student driven approach”,
Journal of Business and Economic Research, Vol. 5 No. 12, pp. 1-12.
Dalton, D.R., Hill, J.W. and Ramsay, R.J. (1997), “Women as managers and partners: context
specifc predictors of turnover in international public accounting frms”, Auditing: A
Journal of Practice & Theory, Vol. 16 No. 1, pp. 29-50.
DeMong, R.F., Lindgren, J.H. and Perry, S.E. (1994), “Designing an assessment program for
accounting”, Issues in Accounting Education, Vol. 9 No. 1, pp. 11-27.
Duncan, J. and Schmutte, J. (2006), “Change in accounting program: the impact of infuences and
constraints”, The Accounting Educators Journal, Vol. 16, pp. 52-81.
English, D.M. and Koeppen, D.R. (1993), “The relationship of accounting internships and
subsequent academic performance”, Issues in Accounting Education, Vol. 8 No. 4,
pp. 292-299.
Feldman, D.C., Folks, W.R. and Turnley, W.H. (1998), “The socialization of expatriate interns”,
Journal of Managerial Issues, Vol. 10 No. 4, pp. 403-418.
Hagen, J. and Siegel, P.H. (1995), “Mentoring for tax professionals: an examination of stress during
a CPA frm merger”, paper presented at the American Accounting Association Annual
Meeting, Orlando, FL.
Hall, M. and Smith, D. (2009), “Mentoring and turnover intentions in public accounting frms: a
research note”, Accounting, Organization and Society, Vol. 34 Nos 6/7, pp. 695-704.
Herring, H.C. III and Izard, C.D. (1992), “Outcomes assessment of accounting majors”, Issues in
Accounting Education, Vol. 7 No. 1, pp. 1-17.
Johnstone, K. and Biggs, S.F. (1998), “Problem-based learning: introduction, analysis, and
accounting curricula implications”, Journal of Accounting Education, Vol. 16 Nos 3/4,
pp. 407-427.
Kleinman, G., Siegel, P.H. and Eckstein, E. (2001), “Mentoring and learning: the case of CPA
frms”, Leadership & Organization Development Journal, Vol. 22 No. 1, pp. 22-34.
Knechel, W.R. and Snowball, D. (1987), “Accounting internships and subsequent
academic performance: an empirical study”, The Accounting Review, Vol. 65 No. 4,
pp. 799-807.
Kolb, D.A., Boyatzis, R.E. and Mainemelis, C. (2000), “Experiential learning theory: previous
research and new directions”, in Sternberg, R.J. and Zhang, L.F. (Eds), Perspectives on
Cognitive, Learning, and Thinking Styles, Lawrence Erlbaum, NJ.
Lowe, R.E. (1965), “Public accounting internships”, The Accounting Review, Vol. 40 No. 4,
pp. 839-846.
McCombs, G. and VanSyckle, L. (1994), “Accounting internships: a win-win arrangement”, The
National Public Accountant, Vol. 39 No. 5, pp. 21-23.
Maletta, M., Anderson, B. and Angelini, J. (1990), “Experience, instruction and knowledge
acquisition: a study in taxation”, Journal of Accounting Education, Vol. 17 No. 4,
pp. 351-366.
Morita, J.G., Lee, T.W. and Mowday, R.T. (1989), “Introducing survival analysis to organizational
researchers: a selected application to turnover research”, Journal of Applied Psychology,
Vol. 74 No. 2, pp. 280-292.
283
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Motzer, R.J., Mazumdar, N., Block, J., Berg, W., Amsterdam, A. andFerrara, J. (1999), “Survival and
prognostic stratifcation of 670 patients with advanced renal cell carcinoma”, Journal of
Clinical Oncology, Vol. 17 No. 8, pp. 2530-2540.
National Association of Colleges and Employers (2009), “Research brief: 2009 experiential
education survey”, Bethlehem, PA, available at: https://naceweb.org/
uploadedFiles/intern.pdf (accessed 17 December 2012).
National Association of Colleges and Employers (2011), “Research brief: 2011 experiential
education survey”, Bethlehem, PA, available at: https://naceweb.org/
uploadedFiles/intern.pdf (accessed 17 December 2012).
O’Shaughnessy, J. and Naser-Tavakolian, N. (2000), “Internal auditing internships and
subsequent academic performance: some empirical evidence”, Internal Auditing, Vol. 15
No. 6, pp. 16-23.
Ostroff, C. and Kozlowski, S.W. (1993), “The role of mentoring in the information gathering
processes of newcomers during early organizational socialization”, Journal of Vocational
Behavior, Vol. 42 No. 2, pp. 170-183.
Pasewark, W.R., Strawser, J. and Wilkerson, J. (1989), “An empirical examination of the effect of
previous internship on interviewing success”, Journal of Accounting Education, Vol. 7 No. 1,
pp. 25-40.
Pencina, M. and D’Agostino, R.B. (2004), “Overall C as a measure of discrimination in survival
analysis: model specifc population value and confdence interval estimation”, Statistics in
Medicine, Vol. 23 No. 13, pp. 2109-2123.
SAS Institute (2009), SAS/Stat 9.2 Users Guide: Survival Analysis, SAS Institute, Cary, NC.
Satava, D. (1996), “Personality types of CPA: national versus local frms)”, Journal of Psychological
Type, Vol. 36 No. 2, pp. 36-41.
Schmutte, J. (1986), “Accounting internships: the state of the art”, Journal of Accounting
Education, Vol. 4 No. 1, pp. 227-236.
Siegel, P.H. and Rigby, J.T. (1988), “The relationship of accounting internships and subsequent
professional performance”, Issues in Accounting Education, Vol. 3 No. 2, pp. 423-432.
Siegel, P.H. and Rigby, J.T. (1998), “Institutionalization and structuring of certifed public
accountants: an analysis of the development of education and experience requirements
for certifed public accountants”, Journal of Management History, Vol. 4 No. 2,
pp. 81-93.
Siegel, P.H., Persellin, M.B. and Rigby, J.T. (1992a), “The relation between graduate tax education
and professional performance”, The Journal of the American Taxation Association, Vol. 14
No. 2, pp. 134-143.
Siegel, P.H., Rigsby, J.T. and Leavins, J. (1992b), “An analysis of the relative contribution of
experience/educations to the professional development of auditors”, Advances in
Accounting, Vol. 10, pp. 143-158.
Siegel, P.H., Rutledge, R.W. and Hagen, J.M. (1997), “The mentor relationship within the public
accounting frm: its impact on tax professionals’performance”, Advances in Taxation,
Vol. 9, pp. 175-199.
Siegel, P.H., Blackwood, B.J. and Landy, S.D. (2010), “Tax professional internships and
subsequent professional performance”, American Journal of Business Education, Vol. 3
No. 5, pp. 51-60.
Siegel, S. (1956), Nonparmetric Statistics for the Behavioral Sciences, McGraw-Hill, NewYork, NY.
ARJ
27,3
284
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Somers, M.J. and Birnbaum, D. (1999), “Survival versus traditional methodologies for studying
employee turnover: differences, divergences and directions for future research”, Journal of
Organizational Behavior, Vol. 20 No. 2, pp. 273-284.
Spiceland, J.D., Siegel, P.H. and George, C.R. (1992), “Educational preparation of auditors,
promotion time, and turnover: a survival analysis”, Advances in Accounting, Vol. 10,
pp. 61-76.
Thompson, J.H. (2011), “A comparative empirical analysis of characteristics associated with
accounting internships”, International Journal of Business, Humanities and Technology,
Vol. 1 No. 1, pp. 54-68.
Viator, R.E. (1999), “An analysis of formal mentoring programs and perceived barriers to
obtaining a mentor at large public accounting frms”, Accounting Horizons, Vol. 13 No. 1,
pp. 37-53.
Viator, R.E. (2001), “The association of formal and informal public accounting mentoring with role
stress and related job outcomes”, Accounting, Organizations and Society, Vol. 26 No. 1,
pp. 73-93.
Viator, R.E. and Scandura, T.A. (1991), “A study of mentor-protégé relationships in large public
accounting frms”, Accounting Horizons, Vol. 5 No. 3, pp. 20-30.
Wright, A. (1988), “The comparative performance of MBAs vs. undergraduate accounting majors
in public accounting”, The Accounting Review, Vol. 63 No. 1, pp. 123-136.
Further reading
Siegel, P.H. and Rigby, J.T. (1989), “An analysis of the development of education and experience
requirements for CPAs”, Research in Accounting Regulation, Vol. 3, pp. 45-68.
Appendix
Corresponding author
Dennis Ortiz can be contacted at: [email protected]
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
Table AI.
Summary of participants
(N ?272)
Internship Group Non-internship Group Totals
N 151 121 272
Gender:
Male 82 58 140
Female 69 63 132
CPA Firm Region:
Southeast 72 59 131
Mid-South 79 62 141
285
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_579662344.pdf
The purpose of this study is to consider the effect of an internship experience on tax
accountants’ professional performance
Accounting Research Journal
A survival analysis of tax professionals’ performance and internship experience
Carl R. Borgia Philip H. Siegel Dennis Ortiz
Article information:
To cite this document:
Carl R. Borgia Philip H. Siegel Dennis Ortiz , (2014),"A survival analysis of tax professionals’ performance
and internship experience", Accounting Research J ournal, Vol. 27 Iss 3 pp. 266 - 285
Permanent link to this document:
http://dx.doi.org/10.1108/ARJ -04-2013-0018
Downloaded on: 24 January 2016, At: 21:21 (PT)
References: this document contains references to 59 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 502 times since 2014*
Users who downloaded this article also downloaded:
Kim Watty, Satoshi Sugahara, Nadana Abayadeera, Luckmika Perera, J ade McKay, (2014),"Towards a
Global Model of Accounting Education", Accounting Research J ournal, Vol. 27 Iss 3 pp. 286-300 http://
dx.doi.org/10.1108/ARJ -08-2013-0054
Nicholas Apergis, Christina Christou, Christis Hassapis, (2014),"Accounting standards convergence
dynamics: International evidence from club convergence and clustering", Accounting Research J ournal, Vol.
27 Iss 3 pp. 226-248 http://dx.doi.org/10.1108/ARJ -06-2013-0031
Mirela Malin, (2014),"Enhancing lecture presentation through tablet technology", Accounting Research
J ournal, Vol. 27 Iss 3 pp. 212-225 http://dx.doi.org/10.1108/ARJ -09-2013-0069
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A survival analysis of tax
professionals’ performance and
internship experience
Carl R. Borgia
Florida Atlantic University, Boca Raton, Florida, USA
Philip H. Siegel
McCurry & Company CPAs, Boca Raton, Florida, USA, and
Dennis Ortiz
University of Texas, Pan American, Edinburg, Texas, USA
Abstract
Purpose – The purpose of this study is to consider the effect of an internship experience on tax
accountants’ professional performance.
Design/methodology/approach – It uses survival analysis, a dynamic methodology that allows for
more precise modeling than static traditional methods used to study promotion and turnover rates in
the past. The hypotheses were tested using a longitudinal database obtained from the human resource
departments of regional Certifed Public Accountant frms located in the southeastern and mid-south
areas of the USA.
Findings – Results were mixed. As in previous studies on the effects of internships on subsequent
professional performance, tax accounting professionals with a master’s degree and prior internship
experience had signifcantly faster promotion rates than those professionals with a master’s degree and
no internship experience. However, tax professionals with a master’s degree and prior internship
experience did not demonstrate a signifcant difference in turnover rate when compared to the
no-internship group.
Practical implications – This research provides evidence that students, employers and institutions
of higher education can use to guide them in their decisions regarding the effects of structured
internships on professional performance – in this case, the professional performance of tax accountants.
Originality/value – Previous research on tax professionals’ performance and internship experience
made use of static research methodologies. This study uses the more dynamic methodology of survival
analysis to see if different fndings result.
Keywords Taxation, Internships, Survival analysis
Paper type Research paper
Introduction
Evidence suggests that there is a positive relationship between internship experience as
a student and subsequent professional performance as a Certifed Public Accountant
(CPA), (Siegel and Rigby, 1988; Spiceland et al., 1992; Siegel and Rigby, 1998;
O’Shaughnessy and Naser-Tavakolian, 2000; Siegel et al., 2010). This is consistent with
conventional wisdom that a carefully managed internship program is an effective
component in the education process, and leads to more effective performance by
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1030-9616.htm
ARJ
27,3
266
Accounting Research Journal
Vol. 27 No. 3, 2014
pp. 266-285
© Emerald Group Publishing Limited
1030-9616
DOI 10.1108/ARJ-04-2013-0018
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
accountants in the workforce (Kolb et al., 2000; Cummings and Tataman, 2007; National
Association of Colleges and Educators, 2009, 2011).
Accounting research on the relationship between professional accounting
performance and internship experience uses traditional static methodology (Siegel and
Rigby, 1988; Siegel et al., 1992b) and focuses on auditors as subjects. Other research on
professional performance and internship experience uses traditional “static” statistical
methodology and focuses on tax professionals as subjects (Siegel et al., 2010). However,
Spiceland et al. (1992) are the only researchers to use survival analysis to study
subsequent accounting performance. The above research has been fairly consistent in
supporting the proposition that prior internship experience improves performance of
public accounting professionals.
Performance is measured in the literature in three ways: promotion rate or velocity
(Siegel et al., 1992b; Spiceland et al., 1992; Siegel et al., 2010), turnover rate (Spiceland
et al., 1992; Siegel et al., 2010) and performance evaluations (Siegel et al., 1992b, 2010).
This paper uses the frst two of these performance measures. Performance evaluations
were not made available, and are not considered in this study.
Prior research on tax professionals’ performance and internship experience makes
use of static research methodologies. Generally, these studies found that tax
professionals with master degrees and internship experience had better promotion and
turnover rates (Siegel et al., 1992b, 2010). This study uses survival analysis, a more
dynamic methodology which analyzes subjects as they enter, progress and leave the
frm as opposed to static studies which only use one data point. Multiple data points
allow the measurement of data to be verifed without misclassifcation, and therefore
lessening the chance of data bias.
Tax professionals were selected to see if results coincide with previous studies using
static and dynamic approaches. This study and prior studies in accounting show that
tax and auditing professionals with graduate degrees and internship experience have
better promotion rates. However, unlike in prior studies, no signifcant differences
were found in the present study on tax professional turnover rates. These fndings
indicate that future research into the turnover variable maybe warranted because of
these different fndings.
Thus, the current study further extends the use of survival analysis by applying it to
the study of tax professionals’ performance and its hypothesized association with
internship experience. This analysis will assist us in predicting the turnover and
promotion rate of tax professionals with internship experience.
The remainder of the paper is organized as follows. The frst section consists of a
literature review of the trends and history of internships; the benefts of internships to
students, employers and sponsors; the empirical fndings of the effect of internships on
subsequent performance; the cultural differences between auditing and taxation; and
the beneft of survival analysis. The second section presents the methodology, which
includes the hypotheses, sampling design, procedures and application of the research.
Finally, the third section presents and discusses the results of the survival analysis.
Literature review
To understand the contribution this study makes to the literature, this section provides
a review of fve research areas relevant to the development of this paper:
267
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
(1) the trend and history of internship requirements in public accounting;
(2) internships benefts to students, employers and sponsoring educational
institutions;
(3) empirical evidence of the effect of internships on subsequent performance;
(4) cultural differences between tax professionals and other public accounting
professionals; and
(5) benefts of survival analysis over traditional, static analysis.
Internship requirements in public accounting: trends and history
Over the years, two trends emerge regarding internship requirements in public
accounting. One trend shows increasing participation in internship programs. The other
indicates that internship programs are becoming more effectively managed and
structured.
The increasing participation of internship programs over the years are shown in
research studies from 1965 to the present. Lowe (1965) reports that 23 per cent of
institutions surveyed sponsored internship programs. Over 20 years later, Schmutte
(1986) reports 69 per cent of responding institutions offer student internships. This trend
clearly shows a sharp increase in sponsored programs during this 23-year period.
Beard (1998) reports a continuing increase in structured internship programs. For
example, most structured programs are fairly young, are for credit only, occur during
the junior year, are paid and require a written project report by the student. However,
Beard reports that most internship programs in accounting are not structured. They do
not have full- or part-time coordinators. Most do not require on-site visits, and share the
responsibility for identifying internship sites with the student and others.
Thompson (2011) reports the results of studies he performed during 1998 and 2003
indicating that internship opportunities are always available, regardless of institutional
support. Thompson essentially views institutional support of internships as a matter of
degree, ranging from none to full institutional support.
Other research show that internships programs better managed and structured over
the years. For example Beard (2007) reports that accounting internship program
administrators have established internship assessment plans. She cites Herring and
Izard (1992), DeMong et al. (1994), and Akers et al. (1997), who recommend that plans
include the identifcation of goals and objectives, employer guidelines for intern
supervision, establishment of internship requirements (a weekly diary, a fnal paper and
an oral presentation) and an assessment of the internship experience by the student and
internship supervisor.
Internships benefts to students, employers and sponsoring educational institutions
In 1986, Schmutte recalls the endorsements of internships by important accounting
organizations (AAA, 1952; AAA and AIA, 1955; AICPA, 1978). He summarizes the
advantages of internships to the participating student, the business employer and the
participating academic institution. Students beneft fromthe consolidation of classroom
experience with practice. They have the opportunity to return to school and correct
educational defciencies discovered during the internship, and the opportunity to clarify
career objectives before graduation. The business employer benefts from the qualifed
students serving as temporary workers during peak periods, the opportunity to assess
ARJ
27,3
268
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
company training and supervision functions and an enhanced reputation among
students. The sponsoring educational institutions beneft from better graduates and
interaction with CPA frms.
Student benefts. The Commission on Professional Accounting Education (1983)
indicates that an internship experience for accounting and tax professionals provides
performance benefts such as faster promotion and lower personnel turnover. These
benefts can help lower the high costs associated with recruitment, training and
replacing personnel (Cluskey and Vaux, 1997). Beard (2007), in her study of accounting
internships, cites employers’ anecdotal claims indicating that there are lower turnover
rates for employees who have participated in an internship. Other more formal research
reports that professionals with internships had a faster promotion rate than those
without an internship experience (Siegel and Rigby, 1988; Siegel et al., 1992b; Spiceland
et al., 1992; Siegel et al., 2010). Knechel and Snowball (1987) and English and Koeppen
(1993) show signifcant improvement in post-internship academic performance. Other
research indicates that accounting students who have internships adapt better to the
professional world.
Research fnds positive effects from internship experience on job interviewing and
the likelihood of a job offer (Pasewark et al., 1989; Feldman et al., 1998), on orientation
and career choice (Blank et al., 1991), on the ability to deal with real-world problems and
experiences; on confdence in the accounting profession (McCombs and VanSyckle,
1994) and on accounting knowledge base and motivation level (Beard, 1998). Research
also reports that internships have a positive impact on technical preparation (Johnstone
and Biggs, 1998), on development of categorical tax structures (Maletta et al., 1990), on
integration into the 150-hour requirement (Duncan and Schmutte, 2006) and on students’
job expectations (Broaddus, 2010).
Mentoring and student benefts. One of the benefts to students of a tax internship is
the possibility of being mentored by a more senior professional. Mentorships have been
shown to have a positive effect on subsequent professional performance and
development.
Viator and Scandura (1991) suggest that mentorship reduces turnover. Similarly,
Kleinman et al. (2001) indicate that mentoring can reduce turnover by imparting frm
values and goals. However, Hall and Smith (2009) indicate that mentoring may increase
turnover by increasing market skills and making the employee more valuable to other
employers. Kleinman et al. (2001) indicate that mentoring can ameliorate the high stress
levels in public accounting. Viator (1999) indicated that mentoring can reduce stress, but
success is conditional upon the mentoring design structure.
Accounting has a history of weak supervision of engagements generally not
encountered in other industries, which can often result in role ambiguity. However,
accordingly, Viator (2001) indicates that mentoring can have a role in reducing this
ambiguity.
Tax accountants often encounter job abstraction. Ostroff and Kozlowski (1993) argue
that mentors are able to reduce abstracting by focusing the younger professional on
knowledge that can be applied to other projects. Moreover, Viator and Scandura (1991)
indicate that mentoring provides opportunities to fnd signifcant assignments to
showcase the young professional’s skills.
Employer benefts. McCombs and VanSyckle (1994) report student internships can
reduce CPA frms’ recruiting and training costs. Internships provide the frm with a
269
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
non-obligatory look at prospective employees and to their training before hiring.
Internships also enable the CPA frm to provide for peak or seasonal personnel needs.
They help both current and prospective employers maintain visibility and professional
stature among accounting students. Beard (1998) indicates that internships help smaller
frms to establish a recruiting system and to link with universities and colleges.
Benefts to sponsoring educational institutions. Pasework et al. (1989) report that the
internship experience enhances the placement opportunities of tax accounting
graduates. Another beneft is reported by Beard (1998), who indicates that internship
programs can assist in the development of CPA frm support for accounting programs.
In addition, internships can provide feedback concerning the accounting curriculumand
help identify potential accounting advisory board members (Beard, 1998).
Empirical evidence of the effect of internships on subsequent performance
Internships are designed to enhance students’ subsequent academic and professional
performance. The extant literature provides us with feedback on these expectations.
Internships and other areas of performance. The internship experience provides
students with learning opportunities and frst-hand, experiential professional
knowledge and development. The extant literature shows that internships have positive
effects on subsequent academic performance. Beard (1998) indicates that the internship
experience makes the subsequent academic studies more meaningful, thereby helping
students become more successful in entering the market place. Other researchers fnd
post-internship students perform better in their auditing and accounting courses
(Knechel and Snowball, 1987), and have better academic performance (English and
Koeppen, 1993; O’Shaughnessy and Naser-Tavakolian, 2000).
Pasework et al. (1989) showprevious internship experience increases the likelihood of
getting an interview with a CPA frm. Beck and Halim (2008) fnd that a previous
internship experience helps students with personal and interpersonal skills in their
professional development, their frst job and their subsequent career.
This prior research supports the contention that internship experience can enhance
the students’ knowledge base and academic performance. The limitation of these studies
is that they do not investigate how post-internship students perform in the workplace.
Internships and subsequent professional performance. Internships have positive
effects on the subsequent professional performance of auditors. There has been limited
research on the post-graduate performance of students who have had accounting
internships because of a lack of data due to the diffculty in obtaining performance data.
However, Siegel and Rigby (1988) were able to followstudents after they graduated from
their degree program. They studied the students’ performance in public accounting and
found that internship students performed signifcantly better than those without an
internship experience.
Similar results were found using survival analysis methodology (Spiceland et al.,
1992). Both Siegel and Rigby (1988) and Spiceland et al. (1992) studies fnd that
professionals with internships had a faster promotion rate than those without an
internship experience. However, the survival analysis approach provides more accurate
results because it is not static. It takes into account the varying time element of
individuals entering and exiting the frm.
Siegel et al. (2010) studied the subsequent professional performance of tax
professionals with and without an internship experience. They fnd that internships
ARJ
27,3
270
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
play a signifcant role in the subsequent performance of tax professionals because they
perform better than those without a similar experience.
Culture differences in tax accountants
Siegel et al. (1997) note that, while the audit and tax functions exist within the same
public accounting frm, cultural differences exist between them. For example, while the
threat of litigation is common to both the audit and tax environments, there are three
fundamental differences:
(1) Role differences: The watchdog role of the auditor is quite different from the
advocate role of the tax professional.
(2) Sociological differences: The offce versus the feld nature of the respective jobs
may differentially impact the effectiveness of audit versus tax internships.
(3) Task differences: Audit personnel are frequently called on to exercise
professional judgment as compared with their tax counterparts, whose job tends
to be more technical in nature.
Given these differences between audit and tax work environments, and the attention prior
research has paid to the effect of internship experience on auditors, an important purpose of
the current study is to examine the effect of internship experience on tax professionals.
Earlier research (Benke and Rhodes, 1980) fnds similar sociological differences
between the team approach of auditors and the autonomous approach of the tax staff.
Also, Anderson-Gough et al. (1998) indicate that the sociological differences of these two
professional groups create separate subcultures.
Bertolini et al. (2010) study social interaction preferences of accounting professionals
working in US public accounting frms. They measure social interaction preferences,
fnding signifcant differences between tax and other public accounting professionals.
The results reveal an elevated need for social interaction among successful, experienced
tax professionals versus their non-tax counterparts. Specifcally, tax professionals
demonstrate a desire to be included and liked by others and seek common goals. This
unique set of social preferences leads us to theorize that internship experience – with its
socialization objective – should have a signifcant performance beneft for tax
professionals in public accounting frms.
Methodology
The current study examines tax professionals using a superior, dynamic methodology.
Somers and Birnbaum’s (1999) comparison of traditional turnover research with survival
analysis indicates signifcant divergence between the two methods, and concludes with a
call to consider survival methodology as a focus of future turnover studies.
Survival analysis incorporates a temporal, dynamic component for analyzing turnover
behavior. There are two methodological advantages of survival over traditional analysis.
First, there is less chance of misclassifying data, and therefore, the measurement method is
verifable. Subjects are classifed as they enter and exit the frm – a simple classifcation.
Second, there is less chance of data bias as a result of censored data.
Past survival analysis research in behavioral accounting has been limited to just two
studies. Spiceland et al. (1992) use educational background to address auditor promotion
and turnover. Their application permits more precise modeling of the entry and exit of
271
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the audit professionals in their study. Bröcheler et al. (2004) examine auditor human
capital, education and experience to predict frm survival.
Kolb indicates that internship experiences provide experiential learning to the
participant and provide a valuable work experience. These experiences are particularly
important for professions such as accounting that require technical and
problem-solving skills (Kolb et al., 2000).
The current paper focuses more on the question of when tax professionals are
promoted rather than why they are promoted. This is an important improvement over
traditional, static approaches, which more often address whether internship experience
is effective in improving performance.
Hypotheses
A comparison is developed for tax professionals with internship experience (I) and
without internship experience (NI). For this comparison, three hypotheses are
investigated:
H1A. S
I
(t) ?S
NI
(t), for all t.
H1B. M
I
(t) ?M
NI
(t), for all t.
H2. T
I
(t) ?T
NI
(t), for all t.
Where
S
i
(t) ? Probability that the promotion velocity to senior for subject t in
group i exceeds t.
M
i
(t) ? Probability that the promotion velocity to manger for subject t in
group i exceeds t.
T
i
(t) ?Probability that the turnover rate for subject t in group i exceeds t.
T ?0, and i ? I and NI (for the two internship experience groups, internship and
non-internship).
The relevant alternative hypotheses (H1A, H1Band H2) are that the survival curves for
tax professionals with and without internship experience are not uniformly the same.
Participants
This study incorporates a similar data gathering approach following prior internship
studies (Siegel and Rigby, 1988; Siegel et al., 2010). Ten regional offces of CPA frms were
contactedthat hadformal internships lastingbetween12and15weeks. Feldmanet al. (1998)
indicate that the most critical internship design characteristics include job autonomy, task
identity(theabilitytobeginandcompleteanassignment) andopportunitiestodevelopsocial
andcommunicationskills. Because researchindicates that entry-level workers are socialized
more quickly when they have more structured training (Blank et al., 1991), we included only
CPA frms that offered such training programs. This can insure uniformity of the frms’
input and output expectations for participants.
The frms were located in the southeastern and mid-south areas of the USA. Each
offce provided educational and professional performance data for tax professionals
who had an internship with the frm and subsequently joined the frm during the years
2001-2010. This time period was selected to allowenough time for the study participants
to be promoted to senior and manager levels. Similar information was obtained for tax
ARJ
27,3
272
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
personnel who did not have an internship experience for the same time period. Appendix
A1 provides a summary of participants.
Design and procedure
Interviews were conducted by one of the research authors working with the human
resource personnel fromeach of the participating offces. The human resource personnel
were instructed to select tax professionals who had a master’s degree in taxation. Those
tax professionals who had participated in the frms’ internship program and were
subsequently hired by the frm were identifed to divide the professionals into two
groups, internship and non-internship experience. The internship experience had to be
in the tax area. To control for prior work experience, individuals with accounting
experience prior to their employment with the frm or with signifcant work experience
in another functional area of the frm, such as auditing, were not included in the sample.
The information obtained fromthe public accounting frms’ personnel fles regarding
each of the tax professionals included the following:
• Name of college and graduation dates from both degrees.
• Number of months employed with the frm if terminated before end of study
period.
• CPA certifcation.
• Number of months lapsed before promotion to senior.
• Number of months lapsed before promotion to manager.
• Year started employment with the frm.
• Gender.
All of the frms used the designations senior and manager.
Internship participation enabled the division of the sample into two distinct groups:
those tax professionals who had an internship experience and those who did not
participate in an internship as a component of their educational programs. The sample
included 151 tax professionals who had an internship. The non-internship group had a
sample size of 121. Appendix A1 provides a profle of the study sample.
In this study, two measures of professional performance, promotion velocity and
turnover rate, are used to assess the relationship between performances, and prior tax
and no-prior tax internship experience. These variables have been tested and
successfully used in similar post-education research (Siegel and Rigby, 1988; Spiceland
et al., 1992; Siegel et al., 1992a; Siegel et al., 2010).
Promotion velocity to senior level in the frm is identifed as the number of months
lapsed from the date of initial employment at the frm until promotion to senior.
Similarly, promotion velocity to manager from senior can be defned as the number of
months lapsed from promotion to senior until the individual is promoted to manager.
Turnover velocity is defned as the number of months lapsed from the date
employment commences until employment terminates. No distinction was made
between voluntary and involuntary turnover. Within the groups, any bias associated
with voluntary versus involuntary turnover should be controlled by our selection
process. While it might seem that individuals with master’s degrees in tax would have
greater employment opportunities, in this case, all the sample participants held an
273
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
advanced tax degree. Thus, it is argued that differences in turnover rates, whether
voluntary or not, could be associated with an internship experience.
Survival analysis application
Survival analysis is a statistical method that has its foundation in medical research
wherein a terminal or fnal event occurs (often referred to as a failure) after a measured
length of time from an initial event. In medical research, detection of a disease is usually
the initial event, and death of the patient is the terminal event. In our study, the initial
events are hire or promotion to senior, and the terminal events are promotion to senior or
manager, and departure from the frm.
Survival analysis has been used in medical intervention studies for drug treatments
and effectiveness studies for disease prevention (Motzer et al., 1999; Pencina and
D’Agostino, 2004). Survival analysis represents an improvement over more
conventional statistical analysis when examining the relationship between education,
experience and professional performance (Spiceland et al., 1992). The application of
survival analysis is appropriate for research where subjects enter the study at varying
times and are observed for different lengths of time (longitudinal studies). Thus,
subjects can be followed from different time periods.
Moritaet al. (1989), andlater Somers andBirnbaum(1999) stronglyrecommendthe use of
survival analysis using turnover as a test variable. With longitudinal studies, some
observations are censored during the study because an individual may leave the study
before the event of interest occurs, or the study may conclude before the event of interest
occurs. For example, an individual who seeks employment elsewhere or is terminated prior
to promotion to senior is a censored observation for that event of interest. Furthermore,
right-censored data are data that are censored after the study concludes. It consists of
personnel who either did not leave or were not promoted. These individuals will not be
observable in the study because they are censored far after the end of the study observation
window. Therefore, longer-lived subjects are more likely to be censored over time.
Survival analysis requires explicit knowledge of three items: the initial event, a
measurement unit for the passage of time and a clear meaning of the terminal event (Cox
and Oaks, 1984). For promotion velocity, survival analysis is used in this study to
compare time in 12-month time intervals. There are two initial and two terminal events:
the subject’s initial employment (initial event) to their promotion to senior (terminal
event), and the time period frompromotion to senior status (initial event) to promotion to
manager (terminal event). Similarly, for turnover velocity, survival analysis is used to
compare the time in 12-month intervals from the subject’s initial employment, to their
termination of employment with the frm.
This approach enables the comparison of the survival experience of all subjects
participating during the study period. Survival analysis utilizes partial information
inferred from the fact that subjects enter the study at varying intervals. For some of
these subjects, the terminal event did not occur within the study period, but the potential
remains for the terminal event to occur at some later point in time. Traditional “static”
analysis, such as average time to promotion, fails to consider these observations,
resulting in less reliable fndings (Spiceland et al., 1992).
Survival analysis is also used to test for differences in turnover rates between the two
study groups; internship experience and non-internship experience. In this case, the
terminal event is a subject’s cessation of employment with the frm. Individuals leaving
ARJ
27,3
274
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the frm within the study period will have experienced the terminal event. Those
subjects still employed by the frm at the end of the study period will not have
experienced the terminal event. Individuals entering later in the study period may not
have experienced the terminal event due to the short-time period between
commencement of employment and the end of the study period. While a conventional
means test ignores these individuals, survival analysis utilizes information associated
with subjects entering and leaving at different times within a study period.
There are two steps involved in the survival analysis methodology. First, the
survival experience or survival distribution function (SDF) for the two groups of
subjects is estimated. The SDF describes the lifetimes of the population of interest (SAS
Institute, 2009). Second, survival curves are compared to test whether survival
experience is signifcantly different between the groups. The survival curve is estimated
by the construction of life tables. Alife table divides the sample into intervals (12-month
periods in this study) and identifes the number of individuals who are under study at
the beginning of each time interval.
The life tables show the time of entry for each subject to be zero. The conditional
probability of being promoted for each interval, or terminated as the case may be, can be
developed observing the number of individuals entering each promotion level, the
number who have been promoted, the number censored and the number of individuals
not promoted. The individuals who are terminated, who leave, or are promoted are
classifed as not surviving the interval. The estimated survival curve is the accumulated
product of these conditional probabilities.
Results
Promotion velocity
The Commission on Professional Accounting Educator (1983) indicated the tax
professional’s advancement rate to be a signifcant indicator of professional
performance. Therefore, if internship experience has a positive impact on a tax
professional’s preparation, individuals with internship experience should advance from
one level to another at a faster rate than those without such experience.
Table I shows the promotion velocity to senior. The tax professionals who had an
internship experience are in panel A; and the tax professionals who did not have an
internship experience are in panel B. In the sample, 96 individuals (64 per cent of
professional tax staff hires) were promoted in the internship experience group, while 59
(49 per cent) were promoted to senior in the non-internship group. Signifcant differences
in life tables were tested using the Wilcoxon statistic. The Wilcoxon statistic for
promotion velocity to senior is 8.4966 with one degree of freedom[1], and a probability
level (p value) of 0.0036. The survey analysis indicates that the promotion velocity to
senior was signifcantly faster for the internship group than for the non-internship
group. Thus, the null hypothesis, H1a, was rejected.
Survival curves for promotion velocity to senior are shown in Figure 1. The vertical
axis shows the cumulative probability of promotion to senior. The horizontal axis shows
the time interval in 12-month increments. The fgure clearly shows the promotion
velocity to senior for the internship group exceeds that of the non-internship group for
the frst 36 months. After 36 months, both groups tend to level out until the 60-month
mark when the cumulative probability of promotion to senior of the non-internship
group actually surpasses their internship experienced counterparts
275
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Table II shows the promotion velocity from senior to manager. The internship group is
shown in panel A, and the non-internship group is in panel B. In the sample, 42
individuals (44 per cent of those promoted to senior) in the internship group were
promoted to manager. However, only nine (10 per cent of those promoted to senior) in the
non-internship group were promoted to manager. The Wilcoxon statistic for promotion
velocity to manager is 71.4700 with one degree of freedom. The probability level (p
value) is 0.0001. These fndings strongly indicate that there is a signifcant difference
between the two groups in promotion velocity to manager. Thus, the null hypothesis,
H1b, was also rejected. We tested for gender differences and found no signifcant
differences between the genders. Using survival analysis, we also tested for signifcant
Table I.
Life tables – promotion
velocity to senior
Interval start
(months) No. entering
a
No.
censored
b
No.
exposed
to risk
c
No.
promoted
Proportion
promoted
d
Cumulative
probability
e
Standard
error
f
Panel A – Internship Group (N ? 151)
0 151 0 151 9 0.0596 0.0596 0.0000
12 142 18 133 51 0.3835 0.4202 0.0600
24 73 21 62 33 0.5323 0.7288 0.0720
36 19 12 13 3 0.2308 0.7914 0.0718
48 4 4 2 0 0.0000 0.7914 0.0790
55 96
Panel B – Non-internship Group (N ? 121)
0 121 1 120.5 0 0.0000 0.0000 0.0000
12 120 28 106.0 32 0.3019 0.3019 0.0030
24 60 15 52.5 25 0.4762 0.6343 0.0209
36 20 2 19.0 1 0.0526 0.6536 0.0246
48 18 9 13.5 1 0.0740 0.6792 0.0256
60 8 4 6.0 0 0.0000 0.6792 0.0305
72 2 1 1.5 1 0.6667 0.8931 0.0305
60 60
g
Notes:
a
Subjects still employed at the start of a particular interval (year) and not having been
promoted in a prior period – for example, in panel A, the number entering the interval starting after 12
months (142) is the number entering the preceding interval (151) minus those censored (0) or promoted
(9) during the preceding interval;
b
subjects lost to follow-up during the interval because of leaving the
frm prior to experiencing the terminal event or not yet experiencing the terminal event at the data
collection date;
c
the average number of subjects exposed to the possibility of being promoted during a
particular interval, that is, the number entering that interval minus one-half the number censored that
interval;
d
the number promoted divided by the number exposed – an estimate of the conditional
probability of being promoted during the interval (i.e., the probability of being promoted in the interval,
conditional on being with the frm and not being promoted previously;
e
the unconditional probability
of being promoted up to and including the current interval; calculated as one minus the accumulated
product of the conditional probabilities of not being promoted in all previous years–for example, in
panel A, he cumulative probability of being promoted after 36 months (0.7914) is the complement of the
accumulated product of the complements of the proportions promoted in each interval up to and
including the current interval: 0.7914 ? 1 – (1 ? 0.0596)(1 ? 0.3835)(1 ? 0.5323)(1 ? 0.2308);
f
the
standard error of the cumulative probability estimates;
g
panel B number censored plus promoted
should sum to 121, The difference of one is most likely due to rounding
ARJ
27,3
276
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
differences between the two groups with respect to promotion velocity from entry-level
to manger. As expected, the internship group was promoted signifcantly faster than the
non-internship group. The Wilcoxon statistic is 64.6332 with one degree of freedom and
a p value of 0.0001.
The Wilcoxon Mann–Whitney test (Siegel, 1956) is one of the most powerful of the
non-parametric tests for comparing two populations. It is used to test the null hypothesis
that two populations have identical distribution functions against the alternative
hypothesis that the two distribution functions differ only with respect to location
(median).
Survival curves for promotion velocity to manager are shown in Figure 2. The
vertical axis shows the cumulative probability of promotion to manager. The horizontal
axis shows the time interval in 12-month increments. The fgure shows the promotion
velocity to manager for the internship group, once again, clearly exceeds that of the
non-internship group for the frst 36 months. After 36 months, the internship group has
reached a plateau, while their non-internship counterparts continue to climb until the
60-month mark. At 60 months, both groups plateau with the cumulative probability of
promotion to manager for the internship group exceeding that of the non-internship
group.
Turnover rate
Turnover rate is an important measure of an accounting professional’s performance
(Spiceland et al., 1992; Dalton et al., 1997; Hall and Smith, 2009; Siegel et al., 2010). Few
studies, however, use actual accounting frmturnover data (Spiceland et al., 1992; Dalton
et al., 1997; Siegel et al., 2010). Following the survival technique of Spiceland et al. (1992),
turnover event is defned as the cessation of employment.
Table III shows the turnover rate for the two groups. The tax professionals who had an
internship experience are in panel A; and the tax professionals who did not have an
internship experience are in panel B. In the sample, 56 individuals (37 per cent)
terminated employment in the internship experience group, while 43 (36 per cent)
terminated employment in the non-internship group. The Wilcoxon statistic for
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 12 24 36 48 60 72
C
u
m
u
l
a
t
i
v
e
P
r
o
b
a
b
i
l
i
t
y
Months
Internship
Non-Internship
Figure 1.
Survival curve-promotion
velocity to senior
277
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
turnover rate is 2.5517 with one degree of freedom, and a probability level (p value) of
0.1102. Using only the last 60 months (thus, dropping the frst 24 months of data) for the
analysis produces a Wilcoxon value of 2.72, with a marginally signifcant p value of
0.0993. The survey analysis indicates that the turnover rate was not signifcantly faster
for the internship group versus that for the non-internship group. Thus, the null
hypothesis, H2, was not rejected.
Survival curves for turnover rate are shown in Figure 3. The vertical axis shows
the cumulative probability of exiting the frm. The horizontal axis shows the time
Table II.
Life tables – promotion
velocity, senior to
manager
Interval start
(months)
No.
entering
a
No.
censored
b
No.
exposed
to risk
c
No.
promoted
Proportion
promoted
d
Cumulative
probability
e
Standard
error
f
Panel A – Internship Group (N ? 96)
0 96 0 96.0 0 0.0000 0.0000 0.0000
12 96 3 94.5 0 0.0000 0.0000 0.0000
24 93 6 90.0 18 0.2000 0.2000 0.0000
36 69 18 60.0 24 0.4000 0.5200 0.0730
??48 27 15 19.5 0 0.0000 0.5200 0.0098
60 12 9 7.5 0 0.0000 0.5200 0.0098
72 3 3 1.5 0 0.0000 0.5200 0.0098
54 42
Panel B – Non-internship Group (N ? 62)
0 62 0 302.0 0 0.0000 0.0000 0.0000
12 62 1 61.5 0 0.0000 0.0000 0.0000
24 61 15 53.5 0 0.0000 0.0000 0.0000
36 46 21 35.5 3 0.0845 0.0845 0.0038
48 22 10 17.0 3 0.1765 0.2461 0.0206
60 9 4 7.0 2 0.2857 0.4615 0.0449
72 3 2 2.0 0 0.0000 0.4615 0.0718
53 8
g
Notes:
a
Subjects still employed at the start of a particular interval (year) and not having been
promoted in a prior period – for example, in panel A, the number entering the interval starting after 12
months (96) is the number entering the preceding interval (96) minus those censored (0) or promoted (0)
during the preceding interval;
b
subjects lost to follow-up during the interval because of leaving the
frm prior to experiencing the terminal event or not yet experiencing the terminal event at the data
collection date;
c
the average number of subjects exposed to the possibility of being promoted during a
particular interval, that is, the number entering that interval minus one-half the number censored that
interval;
d
the number promoted divided by the number exposed – an estimate of the conditional
probability of being promoted during the interval (i.e., the probability of being promoted in the interval,
conditional on being with the frm and not being promoted previously;
e
the unconditional probability
of being promoted up to and including the current interval; calculated as one minus the accumulated
product of the conditional probabilities of not being promoted in all previous years–for example, in
panel A, he cumulative probability of being promoted after 36 months (0.5200) is the complement of the
accumulated product of the complements of the proportions promoted in each interval up to and
including the current interval: 0.5200 ? 1 – (1 ? 0.0000)(1 ? 0.0000)(1 ? 0.2000)(1 ? 0.4000);
f
the
standard error of the cumulative probability estimates;
g
panel B number censored plus promoted
should sum to 62, The difference of one is most likely due to rounding
ARJ
27,3
278
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
interval in 12-month increments. The fgure shows that after the initial 24-month
period, the turnover rate for the non-internship group exceeds that of the internship
group. As the Wilcoxon test indicates, however, the difference is not statistically
signifcant.
Discussion and conclusion
This study considered the effect of an internship experience on tax accountants’
professional performance. It used survival analysis to study promotion (frst to senior
and then from senior to manager) and turnover rates (accountant’s survival in the frm),
which are important measures of tax accountants’ professional performance. The
hypotheses were tested using a longitudinal database obtained from the human
resource departments of regional CPA frms located in the southeastern and mid-south
areas of the USA.
In this study, the internship group was promoted at a signifcantly faster rate
than the non-internship group, both to senior and manager levels. These results are
generally consistent with prior research in auditing (Spiceland et al., 1992) using
survival analysis. The results also are consistent with traditional statistical
analysis of internships for both tax and auditing professionals (Siegel et al., 2010,
and Siegel and Rigby, 1988), which support the predicted positive correlation
between internship experience and subsequent professional performance.
Understanding this result should have important implications for those planning
careers as tax professionals in public accounting. The advancement rate fndings
are also an important indication to both educators and CPA frms of the value of a
well-planned and well-managed internship program.
Intuitively, it would appear that individuals with the same educational background
plus professional internship experience would be promoted earlier than those
individuals with similar backgrounds but without internship experience. Also, prior
research on tax accountant performance shows this to be true (Siegel et al., 1992a,
2010). However, there are key distinctions between these studies and the present
research. First, this study has increased control over the sample selection, resulting
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0 12 24 36 48 60 72
C
u
m
u
l
a
t
i
v
e
P
r
o
b
a
b
i
l
i
t
y
Months
Internship
Non-Internship
Figure 2.
Survival curve-promotion
velocity to manager
279
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
in a more homogeneous sample. Second, it uses a statistical methodology that
improves the testing process by accounting for participants entering and exiting the
sample at different time periods. Somers and Birnbaum (1999) would have
researchers also statistically compare the results of traditional versus survival
analysis. This could further improve our understanding of survival analysis’
contribution to the role internships play in employee performance.
The present study reports no signifcant differences in turnover rates between
the two tax accountant groups. These results are at variance with prior internship
studies using analysis of variance (Siegel and Rigby, 1988; Siegel et al., 2010) and
Table III.
Life tables – turnover rate
Interval atart
(months)
No.
entering
a
No.
censored
b
No.
exposed
to risk
c
No.
leaving
d
Proportion
leaving
Cumulative
probability
e
Standard
error
f
Panel A – Internship Group (N ? 151)
0 151 0 151.0 0 0.0000 0.0000 0.0000
12 151 0 151.0 11 0.0728 0.0728 0.0000
24 130 0 130.0 14 0.1077 0.1727 0.0491
36 106 1 105.5 22 0.2085 0.3452 0.0652
48 78 3 76.5 9 0.1176 0.4222 0.0716
60 66 7 62.5 0 0.0000 0.4222 0.0721
72 59 42 38.0 0 0.0000 0.4222 0.0721
84 42 42 21.0 0 0.0000 0.4222 0.0721
95 56
Panel B – Non-internship Group (N ? 121)
0 121 0 121.0 1 0.0083 0.0083 0.0000
12 120 23 108.5 5 0.0461 0.0540 0.0031
24 92 19 82.5 14 0.1697 0.2146 0.0102
36 59 14 52.0 16 0.3077 0.4562 0.0203
48 29 9 24.5 5 0.2041 0.5672 0.0274
60 15 7 11.5 2 0.1739 0.6425 0.0303
72 6 6 3.0 0 0.0000 0.6425 0.0340
78 43
Notes:
a
Subjects still employed at the start of a particular interval (year) and not terminating
employment in a prior period – for example, in panel A, the number entering the interval starting after
12 months (151) is the number entering the preceding interval (151) minus those censored (0) or
terminating (9) during the preceding interval;
b
subjects lost to follow-up during the interval because of
not yet terminating at the data collection date;
c
the average number of subjects exposed to the
possibility of terminating during a particular interval, that is, the number entering that interval minus
one-half the number censored that interval;
d
the number terminated divided by the number exposed –
an estimate of the conditional probability of terminating during the interval (i.e., the probability of
terminating in the interval, conditional on not terminating previously;
e
the unconditional probability
of terminating up to and including the current interval; calculated as one minus the accumulated
product of the conditional probabilities of not terminating in all previous years – for example, in panel
A, the cumulative probability of terminating after 36 months (0.3452) is the complement of the
accumulated product of the complements of the proportions terminating in each interval up to and
including the current interval: 0.3452 ? 1 – (1 ? 0.0000)(1 ? 0.0728)(1 ? 0.1077)(1 ? 0.2085);
f
the
standard error of the cumulative probability estimates
ARJ
27,3
280
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
using survival analysis (Spiceland et al., 1992). These studies reported signifcant
differences in turnover rates between internship and non-internship groups. As
discussed earlier, perhaps the study results would be different if the analysis were
extended to a longer time period. Also, those tax accountants with internships are
more likely to be in greater demand due to their experience and perhaps leave the
frm for a more lucrative position. Hagen and Siegel (1995) argue that contrasting
results have also been found in other turnover studies (Wright, 1988; Siegel et al.,
1988), suggesting that other factors, such as socialization, outside employment
opportunities and different stress levels, may infuence the retention rate. Future
research might address these conficting results by controlling or specifcally
considering these other factors.
The association between personality types and professional performance should
also be examined. Research has shown that stress levels and voluntary and
involuntary turnover often relate to personality types and exogenous factors such as
mergers, time constraints and management changes (Satava, 1996; Choo, 1995).
These studies report that stress level is a signifcant variable relating to turnover.
Several limitations and recommendations regarding our results should be
considered. As noted earlier, a positive correlation between internship experience
and promotion velocity does not prove a causal effect. Other factors, such as those
discussed above, may play an integral role in the promotion and retention of tax
professionals. For example, the internship group may be more highly motivated
toward professional work and/or have different personality traits that are conducive
to professional tax careers. These individuals may be more aggressive and mature
than their non-internship counterparts. To the extent that these characteristics are
factors in promotion velocity and frm retention, the performance of tax accountants
with internship experience might be upwardly biased.
Our sample was drawn for a specifed geographic area and from regional frms.
Therefore, the results may not be generalizable to other regions and other types of
frms, such as large international or small local CPA frms. Changing market
conditions within the profession and student time constraints may also infuence the
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0 12 24 36 48 60 72 84
C
u
m
u
l
a
t
i
v
e
P
r
o
b
a
b
i
l
i
t
y
Months
Internship
Non-Internship
Figure 3.
Survival curve – turnover
rate
281
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
nature and success of internship experiences. The present study contributes to our
understanding of the effect of an internship on tax professionals’ performance, and
shows the potential of survival analysis as an appropriate method for shedding
further light on this topic.
Note
1. Degrees of freedom (df) are calculated as: df ?2 groups (internship and non-internship) – 1.
References
Akers, M., Giacomino, D. and Trebby, J. (1997), “Designing and implementing an accounting
assessment program”, Issues in Accounting Education, Vol. 12 No. 2, pp. 259-280.
American Accounting Association (1952), “Report of the committee on internship programs”, The
Accounting Review, Vol. 27 No. 3, pp. 361-323.
American Accounting Association and American Institute of Accountants Committee on Faculty
Residency and Internship Programs and Committee on Accounting Personnel (1955),
“Statement of standards and responsibilities under public accounting internship
programs”, The Accounting Review, Vol. 30 No. 2, pp. 206-210.
American Institute of Certifed Public Accountants, Task Force on the Committee on Education
and Experience Requirements for CPAs (1978), Education Requirements for Entry into the
Accounting Profession, AICPA, New York, NY.
Anderson-Gough, F., Grey, C. and Robson, K. (1988), Making up Accountants: the Organizational
and Professional Socialization of Trainee Chartered Accountants, Ashgate, London.
Beard, D.F. (1998), “The status of internships/cooperative education experiences in accounting
education”, Journal of Accounting Education, Vol. 16 No. 3, pp. 507-516.
Beard, D.F. (2007), “Assessment of internship experiences and accounting core competencies”,
Accounting Education, Vol. 16 No. 2, pp. 207-220.
Beck, E.B. and Halim, H. (2008), “Undergraduate internships in accounting: what and how do
Singapore interns learn fromexperience?”, Accounting Education, Vol. 17 No. 2, pp. 151-172.
Benke, R.L. and Rhodes, J.G. (1980), “The job satisfaction of higher level employees in large
certifed public accounting frms”, Accounting Organizations and Society, Vol. 5 No. 2,
pp. 187-201.
Bertolini, M., Borgia, C. and Siegel, P.H. (2010), “The social skill preferences of tax professionals in
CPA frms: a FIRO-B analysis”, The Journal of Applied Business Research, Vol. 26 No. 2,
pp. 105-114.
Blank, M.B., Siegel, P.H. and Rigsby, J.T. (1991), “Determinants of international CPA frm
orientation among accounting students”, The British Accounting Review, Vol. 23 No. 4,
pp. 281-300.
Broaddus, B. (2010), “Graduate accounting internships: counting on a transformation”, Baylor
Business Review, Vol. 28 No. 2, pp. 24-27.
Bröcheler, V., Maijoor, S. andVanWitteloostuijn, A. (2004), “Auditor humancapital andaudit frm
survival: the Dutch audit industry in 1930–1992”, Accounting, Organizations and Society,
Vol. 29 No. 7, pp. 627-646.
Choo, F. (1995), “Auditors’ judgment performance under stress: a test of the predicted relationship
by three theoretical models”, Journal of Accounting, Auditing & Finance, Vol. 10 No. 3,
p. 611.
Cluskey, G.R. and Vaux, A.C. (1997), “Is seasonal stress a career choice of professional
accountants?”, Journal of Employment Counseling, Vol. 34 No. 1, pp. 7-19.
ARJ
27,3
282
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Commission on Professional Accounting Educator (1983), A Post-baccalaureate Education
Requirement for the CPA Profession, National Association of State Boards of Accounting,
New York, NY.
Cox, D.R. and Oaks, D. (1984), Analysis of Survival Data, Chapman and Hall, London.
Cummings, R. and Tataman, W. (2007), “Accounting internships: a student driven approach”,
Journal of Business and Economic Research, Vol. 5 No. 12, pp. 1-12.
Dalton, D.R., Hill, J.W. and Ramsay, R.J. (1997), “Women as managers and partners: context
specifc predictors of turnover in international public accounting frms”, Auditing: A
Journal of Practice & Theory, Vol. 16 No. 1, pp. 29-50.
DeMong, R.F., Lindgren, J.H. and Perry, S.E. (1994), “Designing an assessment program for
accounting”, Issues in Accounting Education, Vol. 9 No. 1, pp. 11-27.
Duncan, J. and Schmutte, J. (2006), “Change in accounting program: the impact of infuences and
constraints”, The Accounting Educators Journal, Vol. 16, pp. 52-81.
English, D.M. and Koeppen, D.R. (1993), “The relationship of accounting internships and
subsequent academic performance”, Issues in Accounting Education, Vol. 8 No. 4,
pp. 292-299.
Feldman, D.C., Folks, W.R. and Turnley, W.H. (1998), “The socialization of expatriate interns”,
Journal of Managerial Issues, Vol. 10 No. 4, pp. 403-418.
Hagen, J. and Siegel, P.H. (1995), “Mentoring for tax professionals: an examination of stress during
a CPA frm merger”, paper presented at the American Accounting Association Annual
Meeting, Orlando, FL.
Hall, M. and Smith, D. (2009), “Mentoring and turnover intentions in public accounting frms: a
research note”, Accounting, Organization and Society, Vol. 34 Nos 6/7, pp. 695-704.
Herring, H.C. III and Izard, C.D. (1992), “Outcomes assessment of accounting majors”, Issues in
Accounting Education, Vol. 7 No. 1, pp. 1-17.
Johnstone, K. and Biggs, S.F. (1998), “Problem-based learning: introduction, analysis, and
accounting curricula implications”, Journal of Accounting Education, Vol. 16 Nos 3/4,
pp. 407-427.
Kleinman, G., Siegel, P.H. and Eckstein, E. (2001), “Mentoring and learning: the case of CPA
frms”, Leadership & Organization Development Journal, Vol. 22 No. 1, pp. 22-34.
Knechel, W.R. and Snowball, D. (1987), “Accounting internships and subsequent
academic performance: an empirical study”, The Accounting Review, Vol. 65 No. 4,
pp. 799-807.
Kolb, D.A., Boyatzis, R.E. and Mainemelis, C. (2000), “Experiential learning theory: previous
research and new directions”, in Sternberg, R.J. and Zhang, L.F. (Eds), Perspectives on
Cognitive, Learning, and Thinking Styles, Lawrence Erlbaum, NJ.
Lowe, R.E. (1965), “Public accounting internships”, The Accounting Review, Vol. 40 No. 4,
pp. 839-846.
McCombs, G. and VanSyckle, L. (1994), “Accounting internships: a win-win arrangement”, The
National Public Accountant, Vol. 39 No. 5, pp. 21-23.
Maletta, M., Anderson, B. and Angelini, J. (1990), “Experience, instruction and knowledge
acquisition: a study in taxation”, Journal of Accounting Education, Vol. 17 No. 4,
pp. 351-366.
Morita, J.G., Lee, T.W. and Mowday, R.T. (1989), “Introducing survival analysis to organizational
researchers: a selected application to turnover research”, Journal of Applied Psychology,
Vol. 74 No. 2, pp. 280-292.
283
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Motzer, R.J., Mazumdar, N., Block, J., Berg, W., Amsterdam, A. andFerrara, J. (1999), “Survival and
prognostic stratifcation of 670 patients with advanced renal cell carcinoma”, Journal of
Clinical Oncology, Vol. 17 No. 8, pp. 2530-2540.
National Association of Colleges and Employers (2009), “Research brief: 2009 experiential
education survey”, Bethlehem, PA, available at: https://naceweb.org/
uploadedFiles/intern.pdf (accessed 17 December 2012).
National Association of Colleges and Employers (2011), “Research brief: 2011 experiential
education survey”, Bethlehem, PA, available at: https://naceweb.org/
uploadedFiles/intern.pdf (accessed 17 December 2012).
O’Shaughnessy, J. and Naser-Tavakolian, N. (2000), “Internal auditing internships and
subsequent academic performance: some empirical evidence”, Internal Auditing, Vol. 15
No. 6, pp. 16-23.
Ostroff, C. and Kozlowski, S.W. (1993), “The role of mentoring in the information gathering
processes of newcomers during early organizational socialization”, Journal of Vocational
Behavior, Vol. 42 No. 2, pp. 170-183.
Pasewark, W.R., Strawser, J. and Wilkerson, J. (1989), “An empirical examination of the effect of
previous internship on interviewing success”, Journal of Accounting Education, Vol. 7 No. 1,
pp. 25-40.
Pencina, M. and D’Agostino, R.B. (2004), “Overall C as a measure of discrimination in survival
analysis: model specifc population value and confdence interval estimation”, Statistics in
Medicine, Vol. 23 No. 13, pp. 2109-2123.
SAS Institute (2009), SAS/Stat 9.2 Users Guide: Survival Analysis, SAS Institute, Cary, NC.
Satava, D. (1996), “Personality types of CPA: national versus local frms)”, Journal of Psychological
Type, Vol. 36 No. 2, pp. 36-41.
Schmutte, J. (1986), “Accounting internships: the state of the art”, Journal of Accounting
Education, Vol. 4 No. 1, pp. 227-236.
Siegel, P.H. and Rigby, J.T. (1988), “The relationship of accounting internships and subsequent
professional performance”, Issues in Accounting Education, Vol. 3 No. 2, pp. 423-432.
Siegel, P.H. and Rigby, J.T. (1998), “Institutionalization and structuring of certifed public
accountants: an analysis of the development of education and experience requirements
for certifed public accountants”, Journal of Management History, Vol. 4 No. 2,
pp. 81-93.
Siegel, P.H., Persellin, M.B. and Rigby, J.T. (1992a), “The relation between graduate tax education
and professional performance”, The Journal of the American Taxation Association, Vol. 14
No. 2, pp. 134-143.
Siegel, P.H., Rigsby, J.T. and Leavins, J. (1992b), “An analysis of the relative contribution of
experience/educations to the professional development of auditors”, Advances in
Accounting, Vol. 10, pp. 143-158.
Siegel, P.H., Rutledge, R.W. and Hagen, J.M. (1997), “The mentor relationship within the public
accounting frm: its impact on tax professionals’performance”, Advances in Taxation,
Vol. 9, pp. 175-199.
Siegel, P.H., Blackwood, B.J. and Landy, S.D. (2010), “Tax professional internships and
subsequent professional performance”, American Journal of Business Education, Vol. 3
No. 5, pp. 51-60.
Siegel, S. (1956), Nonparmetric Statistics for the Behavioral Sciences, McGraw-Hill, NewYork, NY.
ARJ
27,3
284
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Somers, M.J. and Birnbaum, D. (1999), “Survival versus traditional methodologies for studying
employee turnover: differences, divergences and directions for future research”, Journal of
Organizational Behavior, Vol. 20 No. 2, pp. 273-284.
Spiceland, J.D., Siegel, P.H. and George, C.R. (1992), “Educational preparation of auditors,
promotion time, and turnover: a survival analysis”, Advances in Accounting, Vol. 10,
pp. 61-76.
Thompson, J.H. (2011), “A comparative empirical analysis of characteristics associated with
accounting internships”, International Journal of Business, Humanities and Technology,
Vol. 1 No. 1, pp. 54-68.
Viator, R.E. (1999), “An analysis of formal mentoring programs and perceived barriers to
obtaining a mentor at large public accounting frms”, Accounting Horizons, Vol. 13 No. 1,
pp. 37-53.
Viator, R.E. (2001), “The association of formal and informal public accounting mentoring with role
stress and related job outcomes”, Accounting, Organizations and Society, Vol. 26 No. 1,
pp. 73-93.
Viator, R.E. and Scandura, T.A. (1991), “A study of mentor-protégé relationships in large public
accounting frms”, Accounting Horizons, Vol. 5 No. 3, pp. 20-30.
Wright, A. (1988), “The comparative performance of MBAs vs. undergraduate accounting majors
in public accounting”, The Accounting Review, Vol. 63 No. 1, pp. 123-136.
Further reading
Siegel, P.H. and Rigby, J.T. (1989), “An analysis of the development of education and experience
requirements for CPAs”, Research in Accounting Regulation, Vol. 3, pp. 45-68.
Appendix
Corresponding author
Dennis Ortiz can be contacted at: [email protected]
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
Table AI.
Summary of participants
(N ?272)
Internship Group Non-internship Group Totals
N 151 121 272
Gender:
Male 82 58 140
Female 69 63 132
CPA Firm Region:
Southeast 72 59 131
Mid-South 79 62 141
285
Tax
professionals’
internship
experience
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
1
:
2
1
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_579662344.pdf