Content uploaded by Matt Bergman
Author content
All content in this area was uploaded by Matt Bergman on Aug 02, 2018
Content may be subject to copyright.
Copyright © 2018 Wiley Periodicals, Inc., A Wiley Company
1 University of Louisville
Corresponding Author: Mathew Bergman, University of Louisville, College of Education & Human Development , 1905 S 1ST ST, Louis-
ville, KY 40292
Author Email: matt.bergman@louisville.edu
Author note:
This work is based on the doctoral dissertation of the first author. Versions of this paper have been presented at regional conferences through-
out the United States.
New Horizons in Adult Education
& Human Resource Development
30(3), xx-xx
The Impact of Student
Attributes and Program
Characteristics on Doctoral
Degree Completion
Glenn Gittings1
Mathew Bergman1
Kevin Rose1
Brad Shuck1
Abstract
Doctoral students typically represent a highly educated group of students that have demonstrated the academ-
ic aptitude to successfully complete multiple degrees. Yet, research has continually shown that 40%-60% of
doctoral students do not persist to graduate (Allum & Okahana, 2015; Bowen & Rudenstine, 1992). The pur-
pose of this study was to explore the possible influence of individual doctoral student characteristics as well
as doctoral program characteristics on doctoral degree completion. Tinto’s (1993) theory of doctoral attrition
was applied to explore specific variables that may assist or detract a doctoral student with their degree com-
pletion. Results suggested age, full-time employment, employment change after comprehensive exams, en-
rollment status, satisfaction with dissertation chair, and satisfaction with academic involvement all impacted
doctoral completion. Implications for theory and practice are discussed.
Keywords: Doctoral students, adult learners, persistence, educational attainment, doctoral degree comple-
tion
“Doctoral students are people who have generally succeeded at school. For some, choosing to leave
graduate school can feel like the ultimate defeat by a system in which a student has always been
successful” (Golde, 1994, p. 23).
A common misconception perpetuated by institutions is that doctoral students have mastered the practice of
navigating educational waters. By receiving a bachelor’s degree, and in most cases a master’s degree, we
might assume doctoral students would not need – and do not want – guidance or assistance throughout their
2 New Horizons in Adult Education & Human Resource Development, 30(3)
doctoral program. They have made it to the top and can traverse the doctoral experience independently. Unfor-
tunately, the long-term doctoral attrition rate nationwide has historically hovered at just above 50% (Allum &
Okahana, 2015; Ampaw & Jaeger, 2011; Council of Graduate Schools, 2008). Despite the overwhelming need
to justify completion rates (Lipschutz, 1993) and issues of knowledge development and national competitive
advantage (Begin & Gerard, 2013; National Science Foundation, 1997), institutions of higher education typi-
cally focus resources toward undergraduate students and expend less effort engaging graduate and professional
students (Pontius & Harper 2006). Consequently, the structural components surrounding doctoral completion
could include a host of programmatic, institutional, departmental, and individual-level variables. Each of these
factors alone can be a contributing piece of the completion equation, likely leading toward some combination
of completion factors that are unique to each student, program, and institution (Tinto, 2012). Some key defini-
tions help clarify the context of the doctoral student puzzle. For example, attrition has been defined as the vol-
untary or involuntary discontinuance of a student’s participation in the degree program prior to degree comple-
tion (Malmberg, 2000, p. 14), while persistence is defined as a student continuing progress toward doctoral
degree completion (Malmberg, 2000). Within the context of both attrition and persistence, doctoral student de-
gree completion is defined as successful completion of all coursework, comprehensive examinations, and grad-
uation from the doctoral program (Felder, 2010).
Adult students face a multitude of attrition risk factors due to completion deterrents including family responsi-
bilities, age, part-time enrollment status, distance from campus, lack of socialization opportunities with facul-
ty, and employment demands (Bergman, Gross, Berry, & Shuck 2014). As Meriwether (2015) explained, “just
as African American, Latino, LGBT or other student populations cannot simply be treated as a homogenous demo-
graphic group, adult learners require similar strategic focus and attention” (p. 18). Doctoral education remains an
evolving research setting that must be explored not only within the context of those individual and institutional
characteristics that contribute to attrition but also those programmatic interventions shown to positively influ-
ence doctoral student completion rates. In positioning our work, we drew inspiration from de Valero (2001)
who stated:
Given the high costs associated with graduate education, the current national climate of diminishing
resources for higher education, and an increased competition for these resources between undergradu-
ate and graduate programs, understanding and examining the factors that affect the students’ ability to
complete their degree requirements in a timely manner and considering the implications of these fac-
tors becomes crucial. (p. 341)
Drawing from this context, the purpose of our work was to identify individual doctoral student characteristics
and doctoral program institutional variables that influence doctoral student degree completion. Because of the
unique impact that doctoral programs have on research, education, leadership, policy, and professional prac-
tice, attention to this area of higher education is critical and warranted (Felder, 2010; Blair & Hayworth, 1999).
The current study attempted to build upon and add to the research literature by determining individual and doc-
toral program characteristics that may contribute to or impede doctoral student degree completion. Graduate
students experience the acquisition of an entirely new culture when entering graduate education. Graduate stu-
dents can experience, positively or negatively, a socialization process that introduces them to both the academ-
ic department and their future career (Begin & Gerard, 2013; Tinto, 1993). We were interested to know wheth-
er doctoral student demographic variables and program characteristics would have any influence on doctoral
student degree completion. Our work unfolds in the following main sections: (a) overview of relevant litera-
ture, (b) methods, (c) results, (d) discussion, and finally (e) limitations and recommendations for future re-
search.
Overview of Relevant Literature
There are many factors that could contribute to a student’s retention or attrition in an educational degree pro-
gram (Bergman, et al., 2014). According to Lampley (2001), “factors such as decreased funding, slow enroll-
ment growth, rising cost, increased competition, an increasing need for accountability, and a stronger sense of
student consumerism, may force institutions of higher education to take a closer look at how they operate” (p.
13). The National Science Foundation (1998, p. 1) stated, “The doctoral student is a precious resource in
providing the new discoveries and expert knowledge essential to the nation’s future.” Lipschutz (1993) sug-
gested as faculty and graduate deans seek financial resources from government, foundations, and private
sources, they also needed to justify their completion rates. Gillingham, Seneca, and Taussig (1991) explained
3 New Horizons in Adult Education & Human Resource Development, 30(3)
that due to the years committed to completion of the doctoral degree and the sacrifice, measured by potential
contributions to both doctoral students and society, is costly. Moreover, attrition from a doctoral program is
costly for faculty advisers and the institution (Kluever, Green, and Katz, 1997). Given the essential impact
doctoral study has on research, education, leadership, policy, and professional practice, consistently high drop-
out rates remain problematic (Bair & Haworth, 1999).
Tinto (1975), addressed a need to further develop the literature on student persistence and withdrawal behavior
at the undergraduate level that, since then, has received considerable research attention (Pascarella, 1986).
Tinto’s undergraduate model sought to explain that various characteristics influence undergraduate student
persistence. These concepts included background characteristics, initial commitments to the goal of college
graduation, social and academic integration of student within the college, and subsequent commitments to the
goal of college graduation. The model that Tinto developed, substantially contributed to the theoretical under-
standing of undergraduate student persistence and withdrawal behavior (Pascarella, 1986).
After decades of continued and substantial contribution to the topic of undergraduate persistence, Tinto offered
the beginnings of a theory on doctoral student attrition in 1993 and again in 2012 with his influential books on
student attrition. Tinto (2012 & 1993) suggested that, graduate persistence is shaped by the personal and intel-
lectual interactions that occur between students and faculty and various communities that make up the academ-
ic and social systems of the institution. Tinto (1993) explained doctoral persistence as:
reflecting an interactive series of nested and intersecting communities not only within the university,
but beyond it to the broader intellectual and social communities of students and faculty that define the
norms of the field of study at a national level. The process of doctoral persistence seems to be marked
by at least three distinct stages, namely that of transition and adjustment, that of attaining candidacy or
what might be referred to as the development of competence, and that of completing the research pro-
ject leading to the awarding of the doctoral degree. (pp. 234-235).
Tinto (2012 & 1993) further attempted to develop a longitudinal model of graduate persistence (See Figure 1),
but quickly cautioned that the process of graduate persistence cannot be easily described by one simple model.
Tinto postulated that factors of importance to attrition included: student attributes, entry goals and orientation,
institutional and program experiences, academic and social integration into a program, and research experienc-
es (Kluever, Green, & Katz, 1997). He also defined several components of the model, including (a) student
attributes (gender, age, race, ability, individual educational experiences, and social class); (b) external commit-
ments (work and family responsibilities); (c) individual goals (educational and career); (d) commitments (goal
and institutional); (e) financial resources (type and amount of financial aid); and (f) participation in graduate
school (full- or part-time attendance and on- or off-campus residence). The model and theory of doctoral per-
sistence posited by Tinto, is in no way offered as a rigid formula that serves as the only method in which to
study doctoral student retention/attrition. Tinto’s work on doctoral persistence, rather, offers the opportunity to
guide research with tools that help provide a frame of reference and allow for evaluation. Therefore, based on
this theoretical framework, we sought to address the variables that may serve to assist or detract a doctoral stu-
dent with degree completion. A visual representation of Tinto’s theoretical framework is provided below:
4 New Horizons in Adult Education & Human Resource Development, 30(3)
This information could be significant to institutions of higher education, specific academic departments within
the institutions, graduate faculty, doctoral degree program designers, and doctoral students. If specific individ-
ual characteristics or departmental programmatic interventions/coursework contribute to, or detract from doc-
toral student degree completion, the previously mentioned groups would benefit from the knowledge in order
to implement structure or behaviors that would contribute to doctoral student degree completion. Doctoral de-
gree completion will likely never get to 100%, but characteristics, structures, and programming that contribute
to higher degree attainment could assist in raising the overall percentage of doctoral degree completion.
While the decision to persist ultimately may rest on a combination of multiple factors unique to each individu-
al student, we chose to focus on graduate student socialization and support as well as student orientation. This
aligns closely with Tinto’s (1993) model of student retention within the Integration construct and provides a
structural framework with which to investigate doctoral student retention further.
Socialization, Student Support, and Orientation
Students are shaped by the various types of interactions between various individuals at multiple social layers
within the institution (Tinto, 2012, 1993). As students interact with each of these individuals and navigate the
educational institution, information about norms, values, and culture are transmitted. While this process is of-
ten informal, it can also be formalized by a variety of mechanisms. Tierney and Rhoads (1993) conceptualized
the formal process of culture transmission as socialization. Graduate student socialization is fostered through
interactions with faculty, interactions with peers, and opportunities for observation and participation in the ac-
tivities of the profession (Austin, 2002; Begin & Gerard, 2013; Brown-Wright et al., 1997; Corcoran & Clark,
1984; Poock, 2001). For example, graduate students typically have a formal relationship with a faculty advisor
or mentor to help guide the student through the curriculum and the dissertation process. An advisor/student
5 New Horizons in Adult Education & Human Resource Development, 30(3)
relationship characterized by frequent and consistent contact (Begin & Gerard, 2013; Berg & Ferber, 1983)
can have a strong influence on whether or not the student persists through the coursework, comprehensive ex-
ams, and the dissertation writing and defense process (Bair & Haworth, 1999; Malmbeg, 2000; Seagram, et al.,
1998). Because the dissertation writing process is a solitary and somewhat less structured process than other
aspects of the doctoral degree, a relationship with a faculty advisor is critical to student completion during this
stage (Campbell, 1992). Moreover, some institutions may also foster peer socialization through the implemen-
tation of doctoral cohorts, which in turn lead to supportive student networks (Dorn & Papalewis, 1997). The
social cohesion created in these cohort models of education positively influences degree completion (Maher,
Fallucca, & Halasz, 2013; Dorn & Papalewis, 1997) by providing peer networks that engender mutual ac-
countability, support, and even stress reduction (Goplerud, 1980). Further, graduate assistantships and teaching
assistantships contribute to departmental and career socialization for students, and evidence supports that assis-
tantships also increase doctoral student degree completion and career opportunities (Corcoran & Clark, 1984;
Lovitts, 2004).
But graduate assistantships may indeed serve another important purpose beyond socialization. In many colleg-
es and universities, graduate assistantship positions may come with benefits like tuition waivers, stipends,
health insurance, and other benefits. This, and other kinds of formal and informal infrastructure, we broadly
refer to as student support (Bair & Haworth, 1999; Bauer, 2004). Though many kinds of support exist, per-
haps the three most salient are financial support, departmental support, and personal support (Maher, Fallucca,
& Halasz, 2013; Bauer, 2004). As mentioned, financial support can come in the form of assistantships or even
scholarships to attend school. When this kind of support is present, students have to worry less about meeting
basic needs and can concentrate more on successful completion of their academic program (King & Chepyator
-Thompson, 1996; Pauley, 1998). Additionally, the department in which the student studies also provides sup-
port in various ways. Departments distribute important information relevant to academic and professional suc-
cess, they may host learning opportunities outside of the classroom, and they scaffold the graduate student’s
experience through processes and procedures (e.g. comprehensive exams) designed to support student degree
attainment (Begin & Gerard, 2013; Bauer, 2004; Pauley, 1998). Courses, seminars, support groups, and de-
partmental resources can provide doctoral students with much needed structure, experience, and guidance in
eliminating the sometimes-confusing process of completing the doctoral degree. Further, beyond the academic
experience, successful students often have personal supportive networks including support from faculty, peers,
friends, and family (Maher, Fallucca, & Halasz, 2013). These relationships may sometimes serve as both so-
cialization and support mechanism (e.g. faculty and peers), but supportive networks function more to encour-
age students, relieve burdens, hold students accountable, and offer inspiration (Bergman et al., 2014; Bauer,
2004; Pauley, 1998; Seagram et al., 1998). These supportive relationships lead to higher persistence rates for
doctoral students (Emerson, 1998; King & Chepyator-Thompson, 1996; Leadabrand, 1985; Malmberg, 2000).
Tinto (2012) postulated that one of the stages of persistence included a time of initial transition and adjust-
ment. Orientation programs often serve to address perception, transition, and role acquisition that graduate stu-
dents experience by introducing new doctoral students to programs, structures, and policies that aid in the tran-
sition process. Additionally, time to meet faculty and other students when transitioning were found to be posi-
tive aspects of orientation programs (Poock, 2002; Rosenblatt & Christensen, 1993; Taub & Komives, 1998).
This is an important component of socializing doctoral students to their new roles. According to Coulter et al.
(2004), graduate students’ perceived orientations were needed for the improvement of the graduate experience.
Orientation programs generally produced a positive effect on issues of graduate student adjustment and assimi-
lation to a new role and institution (Barker et al., 1997).
Method
To examine the influence of individual doctoral student characteristics and doctoral program institutional vari-
ables on doctoral student degree completion, the researchers developed a survey based on Tinto’s Doctoral
Studies Questionnaire survey and a nationally vetted instrument called the Survey of Earned Doctorates. The
Doctoral Studies Questionnaire was field-tested at Syracuse University and has been used by previous re-
searchers to study, for example, the relationship between variables that influence attrition in a doctoral pro-
gram (Lee, 2003). Similarly, the Survey of Earned Doctorates has been used in prior research (e.g. Abedi &
Benkin, 1987). Therefore, this instrument adapted from Tinto (1995) by Lee (2003) as well as the Survey of
Earned Doctorates, served as the basis for adaptation to our research.
6 New Horizons in Adult Education & Human Resource Development, 30(3)
The following directional hypotheses were tested:
H1: Students who are employed full-time will be less likely to complete the doctoral degree.
H2: Students who experience an employment status change after comprehensive exams will be less likely to
complete the doctoral degree.
H3: Students who are enrolled part-time will be less likely to complete the doctoral degree.
H4: Students who are satisfied with the departmental assistance available will be more likely to complete the
doctoral degree.
H5: Students who are satisfied with the opportunity for social involvement during their doctoral degree will be
more likely to complete the doctoral degree.
H6: Students who are satisfied with the clarity and understanding of academic procedures/requirements within
their program will be more likely to complete the doctoral degree.
H7: Students who are satisfied with the dissertation chair contact will be more likely to complete the doctoral
degree.
H8: Students who are satisfied with the opportunities for academic involvement within their program will be
more likely to complete the doctoral degree.
Participants were recruited from a pool of students enrolled in a doctoral program between the years of 1997
and 2003 at two research-intensive universities in the US Midwest. Due to differences in structure and degree
completion requirements, students pursuing professional degrees (e.g. medicine, law, dentistry, pharmacy)
were not included in this study.
The researchers utilized a two-stage mailing process that included an initial postcard and a reminder postcard
mailed three weeks later. Postcards briefly explained the purpose of the study, requested participation, and pro-
vided the web address to access the survey. Of the total 3,158 former students who were sent a postcard, 275
completed the online survey, representing an 8.7% response rate. Multiple attempts of various methods were
made to help boost the response rate, but unfortunately further improvement in the participation was unable to
be reached. Additionally, it should be noted that we calculated response rate based on the total number of stu-
dents contacted. However, it is very likely that incorrect addresses were used in the mailing process. Because
of how this information was sent, incorrect addresses were not reported back to us and could not be excluded
as part of our analysis. Thus, it is entirely probable that our response rate includes individuals who never truly
received notification of this survey.
Survey Instrument
The survey contained four sections: (1) demographics, (2) doctoral educational information, (3) doctoral stu-
7 New Horizons in Adult Education & Human Resource Development, 30(3)
dent experiences, and (4) support received. Section 1 consisted of questions related to an individual’s back-
ground and demographic information. Section 2 covered questions centered on educational experiences and
characteristics and included a subset of questions for those students that discontinued their doctoral education.
Section 3 covered questions concentrated on doctoral student experiences while pursuing their doctoral degree
and included subsets of questions on the following topics: (a) satisfaction, (b) social involvement, (c) clarity
and understanding of degree requirements, (d) interaction with dissertation chair, and (e) satisfaction with aca-
demic involvement opportunities. Section 4 consisted of questions related to various forms of support received
during the pursuit of their doctoral degree.
Although each instrument has been tested separately, processes to establish validity and reliability were con-
ducted prior to formal collection of data for this study due to the use of an amended study and combination of
instruments. A pilot study was conducted to test the validity of the revised instrument. The scaled question
related to clarity and understanding of academic requirements produced a lower Cronbach’s alpha of α = .227.
All of the remaining scaled questions produced Cronbach’s alpha results ranging from α = .656 to α = .894.
The results were judged high enough to demonstrate consistency of the scaled items in the instrument.
In order to assess the extent to which variables accurately measured the constructs of interest, the survey items
in this research were analyzed using an exploratory factor analysis based on a pilot study performed by the re-
searchers. Additionally, a panel of tenured research methodologists at the authors’ home institution reviewed
the survey instrument to determine if the survey items were content valid.
Sample
The final sample size consisted of 275 individuals who had attended doctoral programs at two research univer-
sities in the Midwest. Table 1 includes the demographic profile of the sample, including gender, ethnicity, em-
ployment status, and enrollment status.
8 New Horizons in Adult Education & Human Resource Development, 30(3)
Table 1 continued
The average age of reporting respondents was 43.41, suggesting that most respondents were adult students.
Most respondents identified as White (n = 236; 85.8%) and a majority of respondents were female (n = 158;
57.5%). We also sought to gather data on participants’ current employment status after their experience with
doctoral studies. Employment status was identified by full-time, part-time, not working, and pursuing other
education. The majority of respondents worked full-time while pursuing their doctoral studies (73.6%) while
far fewer students worked only part-time (16.5%). Only 8.8% (n = 23) of respondents reported not working at
all and 4.6% (n = 12) reported that they were pursuing other education. Enrollment status was defined as either
full time (nine or more hours of concurrent enrollment) or part time (less than nine hours of concurrent enroll-
ment) and we asked respondents to indicate whether they were completing their doctoral program either most-
ly or all full or part-time. The majority of respondents indicated enrollment at the all full-time equivalent of 9
hours or more per semester (31.7%). Closely behind, 30.6% of respondents indicated enrollment at the mostly
part-time equivalent of below 9 hours. When the categories of mostly part-time and all part-time were com-
bined, they accounted for the majority of total respondents (51.5%), indicating an overall enrollment profile of
a part-time student.
Analysis
The independent variables in this study were grouped in two key categories, and within each category, the var-
iables are further defined as more specific measurable attributes.
Individual doctoral student characteristic variables:
1. Age - Ratio scale based on self-reported date of birth by year.
2. Ethnicity - Dummy coded nominal variables of various ethnicities.
3. Gender - Nominal level variables indicating gender with ordering implied.
4. Financial Support - Nominal variables individually selected as they apply to each participant.
5. Employment - Dummy coded nominal variables of levels of employment.
6. Marital Status - Dummy coded nominal variables of status levels.
7. Dependents - Ratio scale based on self-report indicating the number of dependents.
8. Distance From Campus - Ratio scale based on self-report of miles from campus.
9. Debt Load - Interval item scale with 9-point Likert response of debt levels in dollar amounts.
10. Employment Status Change After Comprehensive Exams - Dummy coded nominal variables indicating
status change.
11. Enrollment Status - Interval item scale with 4-point Likert response of enrollment levels.
9 New Horizons in Adult Education & Human Resource Development, 30(3)
Doctoral Program Characteristic variables:
1. Orientation - Dummy coded nominal variables indicating attendance.
2. Departmental Assistance - Interval six item scale with 5-point Likert response measuring satisfaction of
elements within the department.
3. Social Involvement - Interval seven item scale with 5-point Likert response measuring elements of op-
portunity for social involvement during pursuit of the doctoral degree.
4. Dissertation Preparation Courses - Dummy coded nominal variable indicating completion of this type
of coursework.
5. Dissertation Preparation Seminars - Dummy coded nominal variable indicating participation in this
type of seminar.
6. Clarity and Understanding of Academic Program Procedures/Requirements - Interval five item scale
with 5-point Likert response measuring satisfaction of elements and offerings within the academic program.
7. Dissertation Chair Contact - Interval six item scale with 5-point Likert response measuring satisfaction
of various aspects with the dissertation chair.
8. Academic Involvement - Interval four item scale with 5-point Likert-type response measuring satisfac-
tion with opportunities for academic involvement within the doctoral program.
9. Support Groups - Dummy coded nominal variable indicating participation in this type of support
group.
10. Mentor - Dummy coded nominal variable indicating utilization of a mentor.
Logistic regression was used to determine the effect size and variance explained by the model predictor varia-
bles on the degree completion of doctoral students (Leech, Barrett, & Morgan, 2005). Ishitani (2006) defined
degree completion outcomes as having dichotomous values (e.g., whether or not students maintained continu-
ous enrollment to graduation) at discrete points in time. For the purpose of this study, doctoral student degree
completion was defined as graduation from a doctoral degree program that would include completion of all
coursework, passing doctoral comprehensive exams, and fulfillment of the dissertation requirement of the aca-
demic department. This variable categorized a subject in one of two groups (no doctoral degree completion,
doctoral degree completion) and therefore was a dichotomous variable. According to Vogt, “Regression analy-
sis seeks to explain or predict the variability of a dependent variable using information about one or more inde-
pendent variables” (p. 269). Due to the nature of the dependent variable being dichotomous, the researcher uti-
lized logistic regression for statistical analysis. Logistic regression is commonly used to predict the occurrence
of an event, in this case doctoral degree completion. As such, logistic regression was specifically chosen be-
cause of the dichotomous nature of the dependent outcome variable.
A series of logistic regression equations were initiated to develop the most parsimonious model. The first
equation included independent variables centered on individual doctoral student characteristics. The second
equation included independent variables centered on doctoral program characteristics. After completing the
two initial regression equations, the researchers removed variables that did not demonstrate significance (p
< .05). A final model was presented containing all variables indicating significance.
Results
Contrary to our initial assertions, the regression revealed that if a respondent was employed full-time while
pursuing a doctoral degree, they were significantly more likely to complete a doctoral degree (Exp(β) = 4.392,
β = 1.480, p <.05). If, however, a respondent changed employment status after comprehensive exams, they
were less likely to complete the doctoral program (Exp(β) = .218, β = -1.524, p <.05). Thus, H1 was not sup-
ported, but H2 was. Enrollment status (Exp(β) = 1.694, β = .527, p <.05) positively influenced the odds of
completing the doctoral degree, thus H3 was supported…not supported. Although not included in our original
hypotheses, we also found that an increase in respondent age (Exp(β) = 1.043, β = .042, p <.05) positively in-
fluenced the odds of completing the doctoral degree
10 New Horizons in Adult Education & Human Resource Development, 30(3)
The regression results for the variables of departmental assistance (H4), social involvement (H5), and clarity
and understanding of academic procedures/requirements (H6) were nonsignificant, therefore our hypotheses
were not supported. Respondent satisfaction with the dissertation chair was found to be a predictor of program
completion (Exp(β) = 3.012, β = 1.103, p <.05), thus supporting H7. Lastly, increased satisfaction with aca-
demic involvement predicted a decreased chance of graduating (Exp(β) = .506, β = -.682, p <.05). Therefore,
H8 was not supported.
Table 2 displays the summary of model variables from the first logistic regression (i.e., student personal char-
acteristics). Within the model, three variables indicated significance: (a) age (Exp(β) = 1.052, β = .051, p
<.05), (b) job change after comps (Exp(β) = .221, β = -1.508, p <.05), and (c) enrollment status (Exp(β) =
1.990, β = .688, p <.05).
Table 2
Table 3 displays the summary of model variables from the second logistic regression regarding doctoral pro-
gram characteristics. Within the model, two variables were significant (p < .05): (a) satisfaction with disserta-
tion chair (Exp(β) = 2.750, β = 1.102, p <.05) and (b) satisfaction with academic involvement (Exp(β) = .558, β
= -.584, p <.05).
11 New Horizons in Adult Education & Human Resource Development, 30(3)
The research team conducted a final logistic regression combining those variables indicating significance from
the two previous models. To be as comprehensive as possible, in addition to the five variables identified in
each of the previous models, variables that indicated bivariate significance where also included in the final
model.
Table 4 displays a summary of the final logistic model. Results suggested that as a doctoral student increased
in age (Exp(β) = 1.043, β = .042, p <.05) and was employed full-time (Exp(β) = 4.392, β = 1.480, p <.05), the
likelihood of completing the doctoral degree increased. The model also revealed a suppressed relationship be-
tween changed job status after comprehensive exams and persistence (Exp(β) = .218, β = -1.524, p <.05). Ad-
ditionally, a minor finding concerning enrollment status was the indicated significance between enrollment
status and doctoral degree completion (Exp(β) = 1.694, β = .527, p <.05). Results suggested that the higher the
respondent was satisfied with their dissertation chair, the likelihood of completing the doctoral degree substan-
tially increased (Exp(β) = 3.012, β = 1.103, p <.05). Finally, By controlling for other variables, the regression
revealed a suppressed relationship between the respondents’ increased satisfaction with academic involvement
and a decreased odds of doctoral degree completion (Exp(β) = -.682, β = .527, p <.05).
12 New Horizons in Adult Education & Human Resource Development, 30(3)
Discussion
The findings of the current research could provide important opportunities for doctoral programs to adjust the
structure and the culture experienced by doctoral students. Policies within departments are designed to provide
certain experiences, opportunity, and learning desired by the department (Begin & Gerard, 2013; Malmbeg,
2000). The actual practice of carrying out the departmental policies affects the individual and group experienc-
es that both students and faculty encounter throughout a doctoral program.
The two strongest findings revolve around employment status change after comprehensive exams and satisfac-
tion with academic involvement. Our analysis indicated that if a doctoral student changes employment status
after their comprehensive exams, they are less likely to persist to graduation. Doctoral student completion may
increase if program advisers and dissertation chairs consider cautioning against, if at all possible, an employ-
ment status change after completing comprehensive exams. Many individuals within a doctoral program en-
dure the needed commitment of the degree plan in order to eventually create future career opportunities. Once
students complete the comprehensive exams and enter into candidacy, the road to completion becomes a large-
ly unstructured path of independent research. It is at this point, where a student needs further goal commitment
and focus on the research requirement. Opportunities for promotion and new career roles can also become
available due to the achievement of ABD status. It is at this point where the practice and guidance of faculty,
staff, and mentors can serve to remind the student that beginning a new career opportunity can severely siphon
away time, focus, and commitment to completing the degree. This would seem to be consistent with Tinto’s
(1993) model in that any new professional role will require a stronger investment of time and energy to adjust
and perform to the desired level. The requisite decrease in time and attention dedicated to the dissertation pro-
13 New Horizons in Adult Education & Human Resource Development, 30(3)
cess seems to have a negative impact on students’ ultimate success. The results of this study can be used by
programs to demonstrate this negative effect on degree completion and caution against changing employment
status after completing the comprehensive exams.
Surprisingly, respondents who indicated increased satisfaction with academic involvement showed a decreased
chance of graduating. The indication that doctoral student completion may decrease if respondents indicated
higher satisfaction with academic involvement is a peculiar finding. We might consider that the student’s focus
and attention on the dissertation may be getting distracted due to other research interests within the depart-
ment. It also may explain that over involvement, over commitment of non-required activities and research can
sideline completion. Those students that take on too many volunteer research projects, attend a myriad of con-
ferences, and simply take on too many things not connected to completing the coursework and dissertation, are
inviting opportunities and excuses to reduce their focus on completing the doctoral degree. Tinto’s theory on
doctoral attrition seems to focus on the building of communities both socially and academically. The faculty,
staff, and mentor community the student builds could serve as a cautionary force in the area and help reign in
the student’s extracurricular activities to a manageable amount.
Implications and Recommendations
Doctoral programs should consider the implications of higher degree completion among older students and
part-time enrollment status when considering composition of incoming cohorts of doctoral students and the
structure of the coursework offered and the pattern of doctoral student enrollment status. When creating the
course scheduling and offerings, doctoral programs could focus on course times and structures that would bet-
ter appeal to older, part-time enrolled students. Although most programs have several evening coursework op-
tions, the structuring of a single day that offers two courses in one evening may better appeal to students. Start-
ing the first course in the late afternoon, followed by a small dinner break and then offering the second course
in that same evening allows students to devote only one night a week to class, yet attend two courses. This
block scheduling could be an attractive alternative in course scheduling. Weekend coursework and distance
education coursework can also allow for varied opportunities for students to attend either only on select week-
ends throughout a semester or at the convenience of their home through the use of an online format. Simply
working to offer these varied methods of scheduling may serve to remove impediments to efficiently complet-
ing the coursework component of the doctoral degree.
With age demonstrating a strong relationship with degree completion, it would benefit doctoral programs to
integrate admissions practices and services that accommodate older students. The composition of the doctoral
students included an average age of just over 43. The composition also included a high majority of students
employed full-time and a majority of students enrolled part-time. Therefore, it is obvious that the need is not only to
plan for future students of an older age, but the current student population already represents a higher range of
ages and would be well served by policies and services designed for their needs. A needs assessment of the
current population could serve as a direct opportunity to understand what this population may desire to best
succeed. Once these needs are understood, a program could then better serve the current students and the fu-
ture incoming students that will most likely be represented by an older, full-time employed, and part-time en-
rolled composition.
Doctoral programs and incoming doctoral students should consider the implications of higher degree comple-
tion among students satisfied with their dissertation chair. Careful planning and consideration of selecting a
dissertation chair in the early stages of a doctoral program could be facilitated by the program itself and could
be designed to fit into the structure and the culture of the doctoral program. The better a student is informed of
the interests and research of faculty, the better a student can select a chair that aligns to their research interest
and therefore have a stronger opportunity for higher dissertation chair satisfaction. This is consistent with the
research of Brown-Wright et al. (1997) who emphasized the important role of chair in the student mentoring
process.
The correlation results of positive influence of degree completion when students complete either dissertation
coursework or attend dissertation seminars could be important policy opportunities for doctoral programs. The
dissertation is an exercise of independent research by a student who more than likely is an inexperienced re-
searcher at best. Once the student is admitted to candidacy and begins the research component of the degree, a
high majority of the structure, previously provided through the coursework phase, disappears. Doctoral pro-
grams could provide coursework that is focused on practicing the methods of dissertation writing. The more a
14 New Horizons in Adult Education & Human Resource Development, 30(3)
student is exposed to structures, examples, and opportunities to begin guided work on the dissertation, the
more prepared the student will be to complete the daunting task of overcoming the ABD status and completing
the doctoral degree. This process of acculturation and assimilation into the technical process of dissertation
completion might provide students with the extra support they need to finish their program (Begin & Gerard,
2013). Seminars or programs offered through the department on dissertation writing could also serve as both
motivation and a tool to better equip a student with the ability to complete the doctoral degree.
Our findings regarding employment status would seem to go against a common perception that the more a stu-
dent works, the less time that student has to complete a doctoral degree and therefore leading to a higher attri-
tion rate. Table 4 indicates the high majority of respondents worked full-time while pursuing their doctoral
studies (73.6%). While the results must be considered with caution since only a minority of the respondents
were employed part-time, this may indicate a profile of a student that demonstrates a strong motivation, organ-
ization, and structure to complete a doctoral degree while remaining employed full-time.
Doctoral student completion may increase if program advisers and dissertation chairs consider cautioning
against, if at all possible, an employment status change after completing comprehensive exams. Many individ-
uals within a doctoral program endure the needed commitment of the degree plan in order to eventually create
future career opportunities. Once students complete the comprehensive exams and enter into candidacy, the
road to completion becomes an unstructured path of independent research. It is at this point, where a student
needs further goal commitment and focus on the research requirement. Opportunities for promotion and new
career roles can also become available due to the achievement of ABD status. Again, consistent with Tinto’s
(1993) work, it is at this point where the practice and guidance of faculty, staff, and mentors can serve to re-
mind the student that beginning a new career opportunity can severely siphon away time, focus, and commit-
ment to completing the degree. Any new professional role will require a stronger investment in time and ener-
gy to adjust and perform to the desired level. The results of this study can be used by programs to demonstrate
this negative effect on degree completion and caution against changing employment status after completing the
comprehensive exams.
The finding that doctoral student completion may decrease if respondents indicated higher satisfaction with
academic involvement is peculiar. This may be interpreted that possibly the student’s focus and attention on
the dissertation may be getting distracted due to other research interests within the department. It also may ex-
plain that over involvement, over commitment of non-required activities and research can sideline completion.
Those students that take on too many volunteer research projects, attend a myriad of conferences, and simply
take on too many things not connected to completing the coursework and dissertation, are inviting opportuni-
ties and excuses to reduce their focus on completing the doctoral degree. Tinto’s theory on doctoral attrition
seems to focus on the building of communities both socially and academically. The faculty, staff, and mentor
community the student builds could serve as a cautionary force in the area and help reign the student’s extra-
curricular activities in to a manageable amount.
Doctoral student completion may increase if program advisers and dissertation chairs consider recommending
a part-time enrollment status. Yet conversely, doctoral programs should not shy away from those students em-
ployed full-time that desire to pursue a terminal degree. Tinto (1993) explained, “Individuals whose education-
al and career goals are such as to require the completion of a doctorate are more likely to finish than other per-
sons whose goals are not so linked” (p. 239). This result of full-time employment status indicating higher de-
gree completion is an implication that should be considered by doctoral students contemplating the start of a
doctoral degree and doctoral programs when considering composition incoming cohorts of doctoral students
and the structure of the coursework offered to accommodate doctoral student employment status.
Limitations and Recommendations for Future Research
As with all research, the present research has limitations that should be explored. First, limiting the sample to
only a specified six-year time period may not be representative of the broader doctoral student population. In a
similar regard, selecting two research institutions in the Midwest may not allow for wide generalization to oth-
er institutions across the United States, particularly institutions with differing missions (e.g. primarily teaching
institutions) or institutions with targeted student populations (e.g. HBCUs). The study also eliminated students
pursuing a professional degree in the areas of Law, Medicine, Pharmacy, and Dentistry. Thus, it is possible
that elimination of these degrees and the constraining factors imposed on the design could have affected the
outcome of the study. While the results do not inform the larger literature base on doctoral education, the study
15 New Horizons in Adult Education & Human Resource Development, 30(3)
does provide insight to those factors that influence doctoral student persistence toward graduation in general,
and presents fruitful opportunities for future research. For example, future research might consider the role of
doctoral students at HBCU’s or primarily Hispanic Serving Institutions (HIS) and compare and contracts find-
ings with the outcomes from this study. More, it might be interesting to study professional degree attainment,
alongside the PhD, or look more closely Carnegie classification as a factor in degree attainment.
Second, the survey used for this study was administered online, and while the online environment provides for
greater access to a population at a lower cost, it reduces the ability of the researchers to control the conditions
in which the survey is administered, holistically speaking. Because of this, it is possible that some students self
selected out because they did not have working knowledge of how to use a computer, or how to use the Inter-
net (although, we suspect this population might be small considering students were enrolled in earning a termi-
nal degree). It is perhaps more likely that of those who received the postcard in the mail, some did not have
access to the Internet or access to the Internet was inconvenient for them (e.g. requiring a necessary trip to a
public library). Notwithstanding, it is possible that this unintentionally affected who could take the survey and
thus, systematically excluded a portion of the population who could not obtain access to complete online sur-
veys. Despite the low probability, the chance remains that administrating the survey online could have biased
the results.
Finally, the response rate of the survey population was 8.7%, therefore interpretation of the results must be
done with caution. We recognize that this very low response rate presents a problem in interpreting the results
of our study. While we do not wish to diminish the importance of this consideration, we offer several ideas as
to why this might have occurred. Baruch and Holtom (2008) found that response rates are decreasing over
time and can be particularly problematic with populations that are difficult to reach. They suggested that in
cases where response rates are very low, as in our study, some explanation and further detail be given to help
the reader interpret the results judiciously. In that vein, we offer the following thoughts. We hypothesize that
the low response rate might have been due to the way in which the survey was administered. Because post-
cards were sent to the sample, it is possible that incorrect addresses were obtained. It is also likely that the na-
ture of completing the survey was not convenient enough. In a typical online survey format, participants simp-
ly click on a link and are taken to the survey. In our procedure, participants were required to input the web ad-
dress from the postcard into their browser and then complete the survey. Lack of time, incorrect transcription,
and other factors may have contributed to the low response rate. We also see this population as being particu-
larly difficult to reach. Again, we emphasize that university records that are a decade old may not be as fruit-
ful as more up to date records. Additionally, students who never completed their degrees may not have a sense
of goodwill toward the institution and may not be as cooperative in completing surveys. Future research
should work to increase the response rate and might do so by working with more recent student populations, or
local student populations that would allow for the research team to send the survey request using University
affiliated email, and therefore increase the response rate. This could be done with unique collaborations with
offices of Institutional Research, Student Affairs, and/or school of Graduate Education.
References
Abedi, J., & Benkin, E. (1987). The effects of students’ academic, financial, and demographic variables on
time to the doctorate. Research in Higher Education, 27, 3-14.
Ampaw, F. D. & Jaeger, A. J. (2012). Completing the three stages of doctoral education: An event history
analysis. Research in Higher Education, 53(6), 640-660.
Austin, A. E. (2002). Preparing the next generation of faculty: Graduate school as socialization to the academ-
ic career. Journal of Higher Education, 73, 94-122.
Bair, C. R., & Haworth, J. G. (1999, November). Doctoral student attrition and persistence: A meta-synthesis
of research. Paper presented at the annual meeting of the Association for the Study of Higher Education,
San Antonio, TX.
Barker, S., Felstehausen, G., Couch, S., & Henry, J. (1997). Orientation programs for older and delayed-entry
graduate students. NASPA Journal, 35, 57-68.
Baruch, Y., & Holtom, B. C. (2008). Survey response rate levels and trends in organizational research. Human
Relations, 61(8), 1139-1160.
16 New Horizons in Adult Education & Human Resource Development, 30(3)
Bauer, E.R. (2004). An examination of the effect of departmental factors on student completion of doctoral re-
quirements. Unpublished doctoral dissertation, University of Massachusetts Boston. MA.
Begin, C., & Gerard, L. (2013). The role of supervisors in light of the experience of doctoral students. Policy
Futures In Education, 11(3), 267-276.
Berg, H. M., & Ferber, M. A. (1983). Men and women graduate students: Who succeeds and why? Journal of
Higher Education, 54, 629-648.
Bergman, M., Gross, J.P.K., Berry, M., Shuck, B. (2014). If life happened but a degree didn’t: Examining fac-
tors that impact adult student persistence. The Journal of Continuing Higher Education, 62(2), 90-101.
Brown-Wright, D. A., Dubick, R. A., & Newman, I. (1997). Graduate assistant expectation and faculty percep-
tion: Implications for mentoring and training. Journal of College Student Development, 38, 410-416.
Campbell, R. (1992). A study of the completion and non-completion of the doctor of education degree in edu-
cational leadership at the University of Delaware. Unpublished doctoral dissertation, University of Dela-
ware, Newark.
Corcoran, M., & Clark, S. M. (1984). Professional socialization and contemporary career attitudes of three fac-
ulty generations. Research in Higher Education, 20, 131-153.
Coulter, F. W., Goin, R. P., & Gerard, J. M. (2004). Assessing graduate students' needs: The role of graduate
student organizations. Educational Research Quarterly, 28(1), 15-26.
Cuetara, J., & LeCapitaine, J. (1991). The relationship between dissertation writing experiences and doctoral
training environments. Education, 112, 233-241.
de Valero, Y. F. (2001). Departmental factors affecting time-to-degree and completion rates of doctoral stu-
dents at one land-grant research institution. Journal of Higher Education, 72, 341-367.
Dorn, S. M., & Papalewis, R. (1995). Educators earning their doctorates: Doctoral student perceptions regard-
ing cohesiveness and persistence. Education, 116, 305-314.
Dorn, S. M., & Papalewis, R. (1997, March). Improving doctoral student retention. Paper presented at the an-
nual meeting of the American Educational Research Association, Chicago, IL.
Emerson, J. (1998). An investigation of the characteristics that facilitate or impede completion of inquiry, a
nontraditionally organized doctoral program in Educational Administration. Unpublished doctoral disser-
tation, Columbia University, New York.
Felder, P. (2010). On doctoral student development: Exploring faculty mentoring in the shaping of African
American doctoral student success. Qualitative Report, 15(2), 455-474.
Golde, C. M. (1994, November). Student descriptions of the doctoral student attrition process. Paper presented
at the annual meeting of the Association for the Study of Higher Education, Tucson, AZ.
Golde, C. M. (1996). How departmental contextual factors shape doctoral student attrition. Unpublished doc-
toral dissertation, Stanford University, Palo Alto, CA.
Golde, C. M. (2000). Should I stay or should I go? Student descriptions of the doctoral attrition process. Re-
view of Higher Education, 23, 199-227.
Golde, C. M. (2005). The role of the department and discipline on doctoral student attrition: Lessons from four
departments. Journal of Higher Education, 76, 669-700.
Golde, C. M., & Dore, T. M. (2001). At cross purposes: What the experiences of today’s doctoral students re-
veal about doctoral education. Philadelphia: Pew Charitable Trusts.
Goplerud, E. N. (1980). Social support and stress during the first year of graduate school. Professional Psy-
chology, 11, 283-289.
17 New Horizons in Adult Education & Human Resource Development, 30(3)
Green, K. E., & Kluever, R. C. (1997, March). The dissertation barriers scale. Paper presented at the annual
meeting of the American Educational Association, Chicago.
Hatley, R. V., & Fiene, J. R. (1995, April). Enhancing doctoral student progress and improving dissertation
quality: A success scenario. Paper presented at the annual meeting of the American Educational Re-
search Association, San Francisco.
Heiss, A. M. (1967). Berkeley doctoral students appraise their academic programs. Educational Record, 30-
44.
Ishitani, T. T. (2006). Studying attrition and degree completion behavior among first-generation college stu-
dents in the United States. The Journal of Higher Education, 77(5), 861-885. Retrieved August 2, 2011,
from Research Library. (Document ID: 1137308921)
King, S. E., & Chepyator-Thompson, J. R. (1996). Factors affecting the enrollment and persistence of Afri-
can-American doctoral students. Physical Educator, 53, 170-180.
Kluever, R. C., Green, K. E., & Katz, E. (1997, March). Dissertation completers and non-completers: An
analysis of psycho-social variables. Paper presented at the annual meeting of the American Research
Association, Chicago.
Lampley, J. (2001). Service quality in higher education: Expectations versus experiences of doctoral stu-
dents. College and University, 77(2), 9-14.
Leadabrand, J. A. (1985). Doctoral candidate persistence in community and human resources: A replication
of the Pascarella and Terenzini studies based on Tinto’s concept. Unpublished doctoral dissertation, The
University of Nebraska-Lincoln, Lincoln.
Lee, B. (2003). Factors affecting non-completion of doctoral degrees as evidenced by students labeled all
but the dissertation (ABD). Unpublished doctoral dissertation, Tennessee State University, Nashville.
Leech, N. L., Barrett, K. C., & Morgan, G. A. (2005). SPSS for intermediate statistics: Use and
interpretation. Mahwah, N.J: Lawrence Erlbaum.
Lipschutz, S. S. (1993). Enhancing success in doctoral education: From policy to practice. New Directions
for Institutional Research, 80, 69-80.
Lovitts, B. E. (2004). Research on the structure and process of graduate education. In D. H. Wulff & A. E.
Austin (Eds.), Paths to the professoriate: Strategies for entering the preparation of future faculty (pp.
115-136). San Francisco: Jossey-Bass.
Lovitts, B. E., & Nelson, C. (2000). The hidden crisis in graduate education: Attrition from Ph.D. programs.
Academe, 86(6), 44-51.
Maher, M., Fallucca, A., & Halasz, H. M. (2013). Write on! Through to the Ph.D.: Using writing groups to
facilitate doctoral degree progress. Studies In Continuing Education, 35(2), 193-208.
Malmberg, E. (2000). Retention and attrition of doctoral candidates in higher education. Unpublished doc-
toral dissertation, University of North Texas, Denton.
Meriwether, J. L. (In press). My unexpected learning about adult learners: Engaging adult learners through
intentional advocacy. About Campus.
National Science Foundation (1997). Proceedings from the Division of Science Resource Studies: Workshop
on Graduate School Attrition. Arlington, VA.
Pauley, R. (1998). A study of factors relating to the attrition from West Virginia University-Marshall Uni-
versity-West Virginia Graduate College Cooperative Doctoral Program administered by the West Vir-
ginia Board of Trustees. Unpublished doctoral dissertation, West Virginia University, Morgantown.
18 New Horizons in Adult Education & Human Resource Development, 30(3)
Pontius, J. L., & Harper, S. R. (2006). Principles for good practice in graduate and professional student en-
gagement. New Directions for Student Services, 115, 47-58.
Poock, M. (2001). A model for integrating professional development in graduate education. The College Stu-
dent Journal, 35, 345-355.
Poock, M. C. (2002). Graduate student orientation: Assessing needs and methods of delivery. Journal of Col-
lege Student Development, 43, 231-245.
Rosenblatt, H. S., & Christensen, C. (1993). "Welcome to the whole family": A graduate student orientation.
College Student Journal, 27, 502-505.
Seagram, B. C., Gould, J., & Pyke, S. W. (1998). An investigation of gender and other variables on time to
completion of doctoral degrees. Research in Higher Education, 39, 319-335.
Stallone, M. N. (2003). Factors associated with student attrition and retention in an educational leadership
doctoral program. Unpublished doctoral dissertation, Texas A&M University-Kingsville and Texas
A&M University-Corpus Christi, Kingsville.
Taub, D., & Komives, S. R. (1998). A comprehensive graduate orientation program: Practicing what we
preach. Journal of College Student Development, 39, 394-398.
Tinto, V. (1993). Leaving college: Rethinking the causes and cures of student attrition (2nd ed.). Chicago:
University of Chicago Press.
Tinto, V. (2012). Completing College: Rethinking Institutional Action (1st ed.) Chicago: University of Chica-
go Press.
Vogt, W. P. (2005). Dictionary of statistics & methodology: A nontechnical guide for the social sciences (3rd
ed.). Thousand Oaks, CA: Sage.
Wagner, D. V. (1986). Selected personality characteristics and situational factors as correlates of comple-
tion and non-completion of the doctoral dissertation. Unpublished doctoral dissertation, University of
Michigan, Ann Arbor.
Wood, D. J. (1978). Internal and external indicators of attrition in the college of education doctoral program
at the University of Toledo from 1962 through 1976. Unpublished doctoral dissertation, University of To-
ledo, Toledo, OH.
19 New Horizons in Adult Education & Human Resource Development, 30(3)
Appendices
20 New Horizons in Adult Education & Human Resource Development, 30(3)
