Laissez-faire or guidance? Effective supervision of bachelor theses

Abstract

Bachelor thesis supervision involves a supporting goal and an assessment goal, requiring more guidance versus more autonomy and freedom for students, respectively. This paper tests the hypotheses that the final grade of undergraduate bachelor thesis is positively related to a supervisor’s (a) guidance given to and (b) affiliation with the bachelor thesis student/project. To assess these hypotheses, undergraduate students at the FHNW School of Business have been surveyed for three years about their perception and satisfaction with bachelor thesis supervision. This data (n = 189) was combined with student grades before the thesis and their final thesis grades. Our results show supervision to have a measurable impact on the bachelor thesis outcome and most effective when focused on guidance related to goal definition and methodological support. In other areas supervision can, to a certain degree, follow a more laissez-faire supervision style.

Full text

Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=cshe20
Studies in Higher Education
ISSN: 0307-5079 (Print) 1470-174X (Online) Journal homepage: https://www.tandfonline.com/loi/cshe20
Laissez-faire or guidance? Effective supervision of
bachelor theses
Felix Strebel, Stefan Gürtler, Beat Hulliger & Johan Lindeque
To cite this article: Felix Strebel, Stefan Gürtler, Beat Hulliger & Johan Lindeque (2019): Laissez-
faire or guidance? Effective supervision of bachelor theses, Studies in Higher Education, DOI:
10.1080/03075079.2019.1659762
To link to this article: https://doi.org/10.1080/03075079.2019.1659762
© 2019 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group
Published online: 24 Sep 2019.
Submit your article to this journal
View related articles
View Crossmark data

Laissez-faire or guidance? Effective supervision of bachelor theses
Felix Strebel , Stefan Gürtler , Beat Hulliger and Johan Lindeque
School of Business, University of Applied Sciences and Arts Northwestern Switzerland FHNW, Olten, Switzerland
ABSTRACT
Bachelor thesis supervision involves a supporting goal and an assessment
goal, requiring more guidance versus more autonomy and freedom for
students, respectively. This paper tests the hypotheses that the final
grade of undergraduate bachelor thesis is positively related to a
supervisor’s (a) guidance given to and (b) affiliation with the bachelor
thesis student/project. To assess these hypotheses, undergraduate
students at the FHNW School of Business have been surveyed for three
years about their perception and satisfaction with bachelor thesis
supervision. This data (n= 189) was combined with student grades
before the thesis and their final thesis grades. Our results show
supervision to have a measurable impact on the bachelor thesis
outcome and most effective when focused on guidance related to goal
definition and methodological support. In other areas supervision can, to
a certain degree, follow a more laissez-faire supervision style.
KEYWORDS
Supervision; affiliation;
guidance; bachelor thesis;
quantitative research
Introduction
My supervisor guided me through the process and was interested in my topic. I’ve learned a lot and the outcome
is convincing.
I saw my supervisor once at the beginning and once after submitting the paper. I’ve learned a lot and the
outcome is convincing.
Both the above quotes represent typical personal experiences with writing an undergraduate
bachelor thesis and illustrate the contradiction that raised our curiosity and was the starting point
of this study. We seek to understand how two such starkly different supervision experiences can
be associated with a positive outcome, by answering two guiding research questions:
© 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://
creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the
original work is properly cited, and is not altered, transformed, or built upon in any way.
CONTACT Felix Strebel felix.strebel@fhnw.ch School of Business, University of Applied Sciences and Arts Northwestern
Switzerland FHNW, Riggenbachstrasse 16, 4600 Olten, Switzerland
STUDIES IN HIGHER EDUCATION
https://doi.org/10.1080/03075079.2019.1659762

(1) To what extent can the supervisor contribute to the outcome of a thesis?
(2) On what aspects of the thesis supervision should the supervisor focus?
To explain the observed contradiction, we assumed that not just the extent of supervision is
important, but also the focus of the supervision. There could be areas where supervision is necessary
and helpful, whereas in other areas supervision could be demotivating and have a negative impact
on the learning experience.
We answer these questions by studying the supervision of undergraduate bachelor thesis projects
at the FHNW School of Business over a three-year period. The focal bachelor thesis projects are com-
pleted individually or in teams of two students, at the end of a three-year undergraduate degree in one
of four-degree programmes at the School of Business; a German language Business Administration
(BA), English language International Management (IM) and German and English language versions
of the Business Information and Technology (BIT) degrees, respectively.
The bachelor thesis project represents the culmination of students’studies and they are expected
to take primary ownership and responsibility for managing this research project that they complete
for an external client, typically a business organization. Each bachelor thesis project has a single aca-
demic supervisor that is there to offer guidance and support, as well as intervene if projects begin to
go awry. Students are expected to organize the scheduling of supervision meetings and set agendas
for any meetings with their supervisors. Students are also expected to professionally manage their
relationship with their clients in the spirit of a consulting project.
Students complete a number of assignments individually and in groups during their degree in
preparation for the bachelor thesis, as well as receiving methodological and academic writing train-
ing in a dedicated course. The work of the students on the bachelor thesis project is assessed based
on the submitted final bachelor thesis project report at the end of the project and primarily assessed
on the quality of the academic elements of the research, with minor aspects of the process, the inde-
pendence and contribution of the students and the interaction with all involved stakeholders also
included as a part of the assessment framework.
Our study proceeds as follows, we first develop a conceptual foundation for our study, which sup-
ports our hypotheses development, and then explain our research design. Our analytical approach
includes both deductive and inductive elements. Exploration of the features and relations of the
data and a model-building phase lead us to a final parsimonious regression model of optimal
quality. After discussing our results, we conclude by drawing some key implications of our work
for the supervision of bachelor theses.
Supervision and BSc thesis student performance: hypotheses development
In most European countries, the final assessment in a degree programme is a thesis (Meeus, Van
Looy, and Libotton 2004), in which students individually perform a supervised piece of empirical
research. This applies to Bachelor’sandMaster’s degree programmes, whereby ‘the Bachelor’s
thesis is in most cases a derivative of the Master’sthesis’(Meeus, Van Looy, and Libotton 2004,
300). The bachelor thesis typically requires a more applied ‘market research’, serving as a
bridge between education and the job market (Gunneng and Ahlstrand 2002), while the
master thesis is on balance more ‘academic’in nature. Given the prominent gatekeeping function
of an undergraduate bachelor thesis at the end of the academic curriculum and its unique didac-
tic setting, it has ‘an essential role in education, irreplaceable by other means of teaching’(Han
2014, 120).
Successfully completing a bachelor thesis depends on a plethora of factors. Since a ‘thesis is a far
bigger project than most students will ever have undertaken before, it requires more independent
study, more self-motivation’(Murray 2011, 2). At the same time writing a thesis also requires ‘the
development of new skills associated with the comprehension of a large volume of information, criti-
cal analysis and the development of an academic writing style. Many students need support in one or
2F. STREBEL ET AL.

all of these key areas’(Business et al. 2013 citied in Sharma 2017, 3). This entirely unfamiliar situation
calls for a supervisor who consciously chooses an appropriate supervision role.
Yet Çetinkaya and Yılmaz (2017) diagnose a shortage of research on supervision in undergraduate
programmes. This may be due to the fact that higher degrees call for an even more elaborate super-
vision, when compared to undergraduate thesis projects. Interestingly though Schulze and Lessing
(2003) observed in their study on postgraduate supervision a general mismatch between students’
aspirations and supervisors’coaching, leaving students under-supported. While Bloom et al. (2007)
recommend the relationship between the supervisor and the candidate in postgraduate studies to
be based on genuine care, since it is the essential component in determining degree completion.
The momentum of good student–supervisor relations is underpinned by the fact that ‘students com-
mented that the relationship with their chairperson was vital to their success’(Marshall, Klocko, and
Davidson 2017, 82). We argue this equally applies to bachelor theses.
Supervision approaches and quality assessment
Not only the role of the supervisor, but also the process of supervision has to be assessed with care
(Wisker 2004; Wisker et al. 2008). Given the fuzzy nature of the thesis supervision process, it is not
particularly obvious what supervision style is most appropriate:
One dimension of quality in supervision may be related to a scientific perspective, for example, familiarity with the
academic demands on theory and method. Other dimensions may be related to a learning perspective, referring
to the pedagogic purpose of the task and to a societal perspective, mainly referring to students’employability.
From the students’point of view there may also be a dimension related to a social perspective, related to the
degree of service and consideration for different students’needs. Andersson and Person (2002), quoted in Holm-
berg (2006, 208)
Furthermore, the style and degree of supervision may vary considerably between different
research fields (Egan et al. 2009). In the end, supervision would appear to be a juggling act
between the supervisor’s role as a coach, who regards a thesis ‘as a joint responsibility and the
supervisor as a trainer …, responsible for even the work climate’(Holmberg 2006,213),asacon-
sultant, ‘i.e. a resource that the students may use if they want to’(Holmberg 2006, 213) and as a
‘second mother’,whotakes‘responsibility for the total situation in a way similar to parenthood’
(Holmberg 2006, 213).
Due to this multidimensional nature of the supervision approach, the quality of supervision is not
easy to assess. Kleijn et al. (2012) suggest conceptualize it in three parts, the (1) final grade of the
work, the (2) perceived supervisor’s input and the (3) student’s satisfaction. Student personal satisfac-
tion is suggested as an indicator of supervision quality, because of the highly personalized nature of
the supervision process. ‘Student satisfaction can include satisfaction with the student’s own role, sat-
isfaction with their own rate of progress, satisfaction with their supervisor, etc.’(Kleijn et al. 2012, 928).
In particular, dissatisfied students are less likely to finish their thesis (Ives and Rowley 2005). We
assume that students’satisfaction with the guidance is positively related to the final thesis result,
as it is considered an essential element of the student–supervisor relationship and the student–super-
visor relationship is considered to be vital for the outcome of the thesis.
Grading, guidance and affiliation
Most studies conceptualize the student–supervisor relationship with a measure of the proximity of
the supervision (affiliation, intimacy, supporting and so on) and a measure of guidance (direction,
structuring, influence and so on) (Kleijn et al. 2012). Good supervision may enhance the quality of
a thesis and lead to a higher grade by having an appropriate level of proximity and guidance, i.e.
an interest in the student’s progression on the one hand and the intention to influence this pro-
gression on the other (Mainhard et al. 2009; Kleijn et al. 2012). Meeus, Van Looy, and Libotton
STUDIES IN HIGHER EDUCATION 3

(2004) take a notable opposite position by emphasizing the importance of independent learning:
Independent learning means:
leaving as many choices as possible up to the student …[i.e.] the subject, field of action or basic competence
forming the basis of the thesis; which goal, which problem or line of questioning will be at the forefront; the per-
sonal supervisor; the number of meetings with the supervisor; when the meetings with the supervisor take place.
Meeus, Van Looy, and Libotton (2004, 305)
This requirement regarding independence puts student and supervisor in a permanent ‘field of
tension between the demand for a clearer task description, on the one hand, and the need for auton-
omy …, on the other’(Holmberg 2006, 313). We develop a conceptual model of the influences on
the outcome of a bachelor thesis in terms of guidance and affiliation, while controlling for other influ-
ences, see Figure 1.
This tension also refers to the fact that a thesis combines a learning goal and an assessment goal
(Todd, Smith, and Bannister 2006). This makes the supervisor’s role a two-fold one –supervisor expres-
sis verbis and assessor, guiding the student through the research on the one hand –i.e. exercising
influence on the student’s activities –and assessing the quality of this research on the other. Both
goals are rather vast and
it is unclear how it can be decided to what extent a student is indeed able to do research, and to what extent the
student has learned from doing the …thesis, and more importantly what the student has learned. (Kleijn et al.
2012, 927)
What is clear is that too much guidance will eventually lead to assessment problems, ‘since the super-
visor has also influenced that quality. This may result in the supervisor grading his or her own work,
instead of the thesis of a student’(Kleijn et al. 2012, 926; Manathunga 2007).
An essential element of guidance resides in the definition of goals. There may be a consensus on
some general goals to be achieved, yet the specific goals of a thesis and the responsibilities in goal-
setting are always to be clarified –is it the academic department, the supervisor(s), the principal who
handed in the topic (cf. chapter 4 of this topic) or maybe the student him- or herself that carries this
responsibility? This point out the importance of goal alignment between all parties involved (Ander-
son, Day, and McLaughlin 2006; Halse and Malfroy 2010).
The discussion of coaching aspects primarily focuses on the amount of guidance and is often
unspecific, in the sense that a generic feeling of the students is measured, as e.g. ‘My supervisor
wants me to do things his/her way’(Kleijn et al. 2012, 930). Less attention is paid to the contextual
dimension of supervision –referring e.g. to the question of which coaching inputs have the greatest
influence on the success of a bachelor thesis. In our analysis we will have a closer look at the different
elements of guidance and if some have a stronger explanatory power for the final grade. This allows
us to formulate the following hypothesis:
H1: The final grade of the bachelor thesis is positively related to the quality of supervisor guidance.
Turning to the relational aspects of thesis supervision, Kleijn et al. (2012, 928) report ‘that students
who perceive more affiliation from their supervisor receive higher final grades, are more satisfied, and
perceive their supervisor to have made a larger contribution to their learning’.Affiliation –which is
rooted in the motivational disposition and the capacity of a supervisor (Sun 2015)–may take several
forms: personal caring (Firestone and Rosenblum 1988), a sense of responsibility for the academic
achievements of students (Nir 2002), an identification with students and a willingness to dedicate
a considerable amount of time and effort to them (Menzies 1995). Many studies report positive
effects of these relational components on student learning (Gill and Reynolds 1999; Janisch and
Johnson 2003) and even on academic achievements (Housego 1999; Solomon 2007).
All aspects considered, Sun (2015, 616) concludes that ‘teacher commitment is positively corre-
lated with students’achievement’. From a student’s point of view, it is the perceived supervisor’s con-
tribution to work, being related to his or her involvement and control, that is positively influencing
4F. STREBEL ET AL.

the learning outcome (Shuell 1988). According to Kleijn et al. (2012), the final grade largely depends
upon the type and level of the supervisor–student relationship. Affiliation, i.e. the interpersonal proxi-
mity between student and supervisor and of any other stakeholder involved, produces a positive
linear effect on the grade –the more affiliation, the better the grade. We assume that this is also
the case for the student satisfaction with the affiliation. This suggests the following hypothesis:
H2: The final grade of the bachelor thesis is positively related to the degree of supervisor affiliation.
Methods
Study context
The focal Swiss University of Applied Sciences School of Business offers four bachelor degree pro-
grammes, Business Administration in German, Business Administration in English (International Man-
agement), Business Information Technology in both a German and English language degree. The
English version of the Business Information and Technology degree programme is new and had
the first graduates in 2017. Just one student of this degree programme has completed the survey
in 2017. Therefore, we will consider the BIT programmes as one programme in the analysis. There
were no significant pedagogical interventions or improvements on any of the degrees.
The bachelor students work on an assignment from a company or an organization for their bache-
lor’s thesis. The assignments are business related and can include the whole range of subjects under
Figure 1. Conceptual model of influences on BSc thesis performance.
STUDIES IN HIGHER EDUCATION 5

Business Administration or Business Information Technology degrees. Some topics are timeless, e.g.
marketing concepts, cost calculations and so on, whereas other topics arise in waves, e.g. balanced
scorecard, social media concepts and so on. The duration of the thesis is three to five months. Fulltime
students write their thesis at the end of three years of studies, part-time students after four years of
studies.
Three parties are involved in the bachelor thesis. The representative of the company or organiz-
ation (client), the student or students and the supervisor. The client pays a fee, to ensure that the
client has an interest in the topic and contributes to the project. The client sets the overall goal of
the bachelor thesis, but specific details are negotiated between the three involved parties, with
the supervisor supporting the student(s) in leading this negotiation. The experience of the supervisor
in judging feasibility of the bachelor thesis is critically important in this negotiation process.
The bachelor thesis projects studied in this paper were supervised by a total of 82 supervisors
that were required to hold a minimum qualification of a master degree to be able to supervise a
bachelor thesis. New supervisors are given an introduction to the bachelor thesis project and docu-
ments, such as the guidelines and assessment framework, to ensure a common understanding of
the projects and supervision role. The supervisor evaluates the bachelor thesis project together
with the client, no further examiner is involved in the grading. In case of disagreement with the
client about the grade, the supervisor has the final word to ensure an academic evaluation. Super-
visors can supervise theses for all four degree programmes (BA (DE), IM (EN) and BIT (EN/DE)), but
usually have a core programme for which they supervise theses. Individual and group theses are
treated equally, that means that the same expectations and grading criteria are used for group
and individual theses. The thesis regulations include ethical considerations in particular concerning
surveys and interviews.
Sample and procedure
In the years 2014, 2015 and 2017, a satisfaction survey was administered to the bachelor students
who had just finished their thesis. In 2016, the survey was not conducted due to longer holidays
of the responsible person. Table 1 shows the number of observations per year in terms of survey
response rates and thesis type, i.e. whether the person answering was working alone or in a team
of two men, two women or man and woman.
Table 1. Overview of sample.
Year
2017 2015 2014 Total
Survey responses
Numbers of thesis 178 111 100 389
Responses 80 49 60 189
Response rate 45% 44% 60%
Team setting and year
Man 25 21 32 78
Woman 17 18 11 46
Two men 13 7 10 30
Two women 13 2 7 22
Man and woman 12 1 0 13
BA BIT IM Total
Team setting and study programme
Man 18 46 14 78
Woman 22 12 12 46
Two men 12 0 18 30
Two women 10 0 12 22
Man and woman 11 0 2 13
Sum 73 58 58 189
6F. STREBEL ET AL.

Pooling over the three years there are 189 observations in total, where an observation refers to a
student that filled in a questionnaire, with annual response rates varying between 44% and 60%. To
test the legitimacy of the pooling, we use heteroscedasticity robust standard errors and include the
survey year in the regression analysis. The students can decide if they want to work alone or in a team
of two on their thesis. The largest group is a man working alone, with 78 theses, while the second
largest group, with 46 theses, is of a woman working alone. When working in a team students
prefer to work with a fellow student of the same gender, there are 30 observations of all men
teams and 22 observations of all women thesis teams. Only 13 observations are of mixed-gender
bachelor thesis teams, with one man and one woman. An observation in our study is a student
from a team that has filled out the survey. It is possible that two students that worked together
have both answered the questionnaire. They will have the same bachelor thesis grade though the
individual judgement about the support of the supervisor and the satisfaction with the support
may differ as well as other potential covariates like gender and age. The pairing for the bachelor
thesis can be checked via the birthdate and the gender of students. In total, there is clustering for
the same bachelor thesis among 13 pairs, where both students completed the satisfaction survey,
corresponding to 26 of the 189 students in total.
The three curricula involved in the study are presented in Table 1. The thesis in the BIT curriculum
is always an individual study and has the lowest proportion of women as students.
The survey was kept short in order to attract as many participants as possible. The survey
included four groups of questions, addressing the (1) perceived supervision support at the
start of the thesis project (with defining the objectives and elaborating the practical approach),
(2) supervisor availability during the thesis process measured in terms of reaction time to
requests, (3) student perception of the meetings (helpful, patient, motivational,
convincing, correct, understanding, setting of appropriate requirements, objective), and (4)
student satisfaction with the supervision process (in terms of subject-specific support of the
supervising lecturer, the methodological support of the supervising lecturer, the practice orien-
tation of the supervising lecturer, and the overall supervision experience). To understand differ-
ences in satisfaction dimensions and to avoid feedback from the bachelor thesis grade to the
satisfaction variables, since the survey was carried out after the students knew their bachelor
thesis grade, the overall satisfaction was not considered in the analysis. The full questionnaire
can be found in the Appendix.
The survey was not anonymized, which allowed us to match responses to institutional records of
students, including the average grades of the students before the thesis, student age, student gender
and the final bachelor thesis grade.
Variables
The variables that are considered in the analysis are described in Table 2.Variable1‘bachelor
thesis’grade is taken from institutional records. Variables 2–7 directly correspond to questions
in the questionnaire. Variable 8 meetings is derived from 8 items evaluating the meetings
between students and supervisors in the questionnaire. While variables 10–16 are drawing
from institutional records.
Dependent variable
The bachelor thesis grade operationalizes the success of a bachelor thesis, see Figure 1 and Table 2,
the dependent variable for the study. The grading system in Switzerland is based on a scale from 1.0
to 6.0; 6.0 is the highest grade possible. A 4.0 is satisfactory or put differently a pass, while a grade of
5.0 is seen as a good result and 6.0 exceptionally good. The scale of the variable btgrade is in 10th of a
full grade and thus variable btgrade has a possible range from 1 to 6 with one decimal; most students
receive a final grade between 4.0 and 6.0, see Figure 2.
STUDIES IN HIGHER EDUCATION 7

Independent variables
Guidance is operationalized by five variables respectively measuring supervisor support with the
definition of specific realistic goals, supervisor support with the elaboration of the practical approach,
satisfaction of the student with subject-specific, methodological and practice orientation support, see
Figure 1. A principal component analysis showed that the variables used for the construction of gui-
dance are well aligned on the first principal component. Guidance is the mean per observation of the
variables aims, methods, sat_subject, sat_methods and sat_practice. Missing values are omitted
when taking the mean. Variable guidance has a range of 1–4, where 1 indicates low student satisfac-
tion with guidance and 4 indicates high satisfaction with guidance.
Affiliation is measured by two variables, the supervisor responsiveness/availability in terms of reac-
tion time and the student perception of the meetings. The multiple response questions on the meet-
ings were coded from the initial dichotomies (helpful, patient, motivating, convincing, correct,
understanding, reasonable) into a variable 8, meetings, by setting missing answers for a dichotomy
to 0, positive answers to 1 and negative answers to −1 and, finally, summing these dichotomies.
Availability is coded as a scale from 1 to 4, where 4 means the best availability. The correlation
between availability and meetings is moderate (0.317) but clearly significant. Finally, affiliation is
the sum of availability and meetings/4 such that the potential range of affiliation is from −1to6,
with 6 representing the highest degree of affiliation.
Figure 2. Histogram bachelor thesis grades.
Table 2. Overview of variables.
Var Var name Description
Dependent variable(s)
Thesis. 1 btgrade Final grade of the bachelor thesis
Independent variables(s)
Guidance 2 aims Support with the definition of specific realistic goals
3 methods Support with the elaboration of the practical approach
4 sat_subject Satisfaction with subject-specific support
5 sat_methods Satisfaction with methodological support
6 sat_practice Satisfaction with practice orientation of support
Affiliation 7 availability In terms of reaction time
8 meetings Derived from meetings qualifications in the survey
9 sat_overall Overall satisfaction (not used in regression due to correlations)
Controls
10 avgrade Average grade prior to bachelor thesis (average if working with partner)
11 avage Age of the student (Average age if working with partner)
12 Team setting Gender and group formation.
1: one man, 2: one woman, 3: two men, 4: two women, 5: man and woman
13 Curriculum Curriculum (BA, IM, BIT), originally 9, but aggregated to 3 later
14 year 2014, 2015 or 2017
8F. STREBEL ET AL.

Control variables
Several previous studies that have measured the value of the supervision on the outcome considered
students as a uniform group (e.g. Kleijn et al. 2012; Mainhard et al. 2009). We added control variables
to our model that focus on student attributes, see Figure 1. The student academic ability is measured
with the average grade a student or with the average grade a pair of students achieved prior to the
thesis project. We added student age as a proxy for experience. The team setting distinguishes
between a man and a woman working alone and teams consisting of two men, two women or
man and woman. The three degree programmes Business Administration, International Management
and Business Information Technology are coded in variable curri and the year of the cohort when the
bachelor thesis was written is also a recorded. Adding these controls allows us to answer the question
if supervision has a measurable impact on the result with greater confidence, by more actively con-
sidering student attributes in the model.
Analysis and validity
Descriptive and inferential analysis
The analysis of the data used descriptive and exploratory methods to understand the nature of the data
(Cook and Weisberg 1999). The inferential analysis was completed in two steps. First, a regression
model including all covariates and the variables guidance and affiliation was established, providing
a base model for the analysis. Diagnostic plots are used to find outliers and leverage points. Then a
model-building process eliminated variables with low predictive power and lead to a parsimonious
regression model. After the regression analysis with the constructed variables guidance and affiliation,
the component variables of these two constructs were analysed in a post-hoc analysis to see whether
particular constituents of guidance or affiliation could be identified as the main drivers for the success
of a bachelor thesis. In both steps we adopted a model-building process, using ‘all subsets’regression,
leading to the identification of a parsimonious model with good explanatory power. This type of search
through a large number of potential models using the Bayesian Information Criterion (BIC)
1
has good
potential in finding a globally best model (Fox 2015; Stock and Watson 2007, 552–554).
Heteroscedasticity robust standard errors were used for the tests on the coefficients. A transform-
ation of btgrade was also applied to check whether a less heteroscedastic representation would point
to other variables as predictors in the base model. This was not the case (Stock and Watson 2007,
326–328).
Tests for robustness of analysis
Methodological concerns addressed in the analysis process include an evaluation of the distributions
of the individual variables, linearity of the data collected, and multicollinearity (Stock and Watson
2007, 206–210). The distribution of the final grades shows left skewness, see Figure 2, with 5 bachelor
theses that have been graded below 4. On the other hand, four students got the maximal grade of
6.0. The main part of the grades is concentrated between 5.0 and 5.7 with a median of 5.4.
Minor deviations of the distribution of the curricula in the sample compared with the population
also occur (see Table 2). Overall, the deviations were deemed rather small and it is not necessary to
adopt a weighting scheme. The satisfaction on all measured dimensions was in general high with
only a few students that were dissatisfied. The distribution of the answers to question 1 on the
support provided when defining specific and realistic goals of the thesis, i.e. variable aims, is shown
in Figure 3. The bar charts of all other variables exhibit a similar monotone increasing pattern or at
least left skewness.
The normalplot of the residuals at the base model shows some asymmetry but no clear outliers.
The residual plot does not show a clear nonlinearity or outliers, but indicates decreasing variability
with increasing predicted value (Stock and Watson 2007, 124). Therefore, the tests for the coefficients
will use heteroschedasticity robust standard errors. Figure 4 shows the scatterplot, using the
STUDIES IN HIGHER EDUCATION 9

hexbinplot approach, of the dependent variable btgrade versus the variable guidance. The plot
shows a rather clear and generally linear relationship but also considerable scatter and heteroscedas-
ticity (Stock and Watson 2007,92–96).
Students with the maximum grade of 6.0 can be rather dissatisfied with the received guidance and
rate the guidance received with low 1.8 points on a scale from 1 to 4. On the other hand students with
a grade of 3.8, that is just a pass, can be highly satisfied with the received guidance and rate guidance
with the highest possible marks. The possible objection that giving a good bachelor thesis grade will
lead automatically to students that are satisfied with the received guidance is thus not supported by
the data.
Results
The results section presents and discusses the descriptive statistics and correlation results of the
analysis, then presents the results of the regression analysis for the concepts guidance and affiliation
and finally provides a post-hoc analysis to identify the most parsimonious model for explaining
student performance in BSc thesis projects.
Descriptive statistics and correlations
The number of observations varies slightly across the models because of missing values. The final
grade of the bachelor thesis (variable btgrade) has a mean of 5.264 and a standard deviation of
Figure 3. Histogram bar chart variable aims.
Figure 4. Hexbinplot BTGrade versus guidance.
10 F. STREBEL ET AL.

0.496. The average grade of bachelor theses over a period of 12 years is 5.07. It seems that the stu-
dents that participated in the satisfaction survey were slightly better than the average; see the dis-
tribution of the grades in Figure 2.
The correlation between the average grade of all subjects before the thesis and the bachelor thesis
is visible (see Figure 5). Nevertheless, the variance is considerable and even with a rather low average
grade before the thesis, a high bachelor thesis grade is possible, e.g. with an average grade of 4.57 a
bachelor thesis grade of 5.9 has been achieved. On the other hand, a good average grade before the
thesis does not automatically lead to a good bachelor thesis grade, e.g. with an average grade before
the thesis of 5.36 students have subsequently achieved a 4.4 in their bachelor thesis.
The Spearman correlations of btgrade, guidance, and affiliation are given in Table 3. Obviously, the
correlation of 0.665 between guidance and affiliation makes it difficult to distinguish between the
influence of guidance and affiliation on btgrade, though the correlation of guidance with btgrade
is stronger than between affiliation and btgrade.
The variables directly derived from the questionnaire and the average grade have highly signifi-
cant correlations. However, the satisfaction variables are, as usual, also correlated strongly among
themselves and the question of what variables are actually the most important ones must be
answered by a multiple regression analysis.
Regression analysis
The full model with btgrade as the dependent variable (outcome) and with all potential independent
variables (predictors)
2
yields an overall adjusted R
2
of 0.298 with 187 observations (2 observations
were deleted due to missing values). Using a search through all possible subsets of the variables
and the BIC criterion, all variables except avgrade, team setting, guidance and affiliation can be
dropped from the model without a loss of predictive power. In the team setting, we used the
largest number of observations, namely one man working alone (n= 78) as the base line. Then, we
evaluated if the findings for the other compositions, one woman, two men, two women or man
and woman together, differ from the base line. The normal plot of the residuals showed mild asym-
metry, but no outliers and visible but not strong heteroscedasticity. Overall the model showed a good
Figure 5. Bachelor thesis grade versus average grade before the bachelor thesis. Empty circles indicate pairs.
Table 3. Correlations.
btgrade Guidance Affiliation
btgrade 1.000 0.414 0.319
guidance 0.414 1.000 0.665
affiliation 0.319 0.665 1.000
STUDIES IN HIGHER EDUCATION 11

fit with an adjusted R
2
of 0.319. The model is presented in Table 4 and is called the base model. The t-
test for the individual variables was also carried out with heteroscedasticity robust standard errors,
but since they are only marginally different from the normal t-tests we show the latter ones in
Table 4.
The grade of the bachelor thesis is influenced by the average grade of the students prior to the
thesis and as a rule of thumb academically strong students tend to write a good bachelor thesis.
This is not surprising, of course, but must be taken into account in order not to overstate the
influence of other variables, in particular, guidance and affiliation. Variable guidance is significant
at the 5% level and adds about 0.166 to btgrade when raising guidance by 1.0 on a scale from
one to four. Variable affiliation is just significant at the 5% level and adds about 0.117 to btgrade
for every 1.0 increase on a scale from −1 to 6 in affiliation. Only groups composed of two women
or a woman and man working together are significantly better than a single man, adding about a
0.270 or 0.258 to btgrade, respectively, see Figure 6.
To illustrate the potential influence of supervision on student performance for the bachelor thesis,
a minimum and maximum impact scenario on the weakest academic student is presented. The
lowest average grade before the bachelor thesis is an average grade (avgrade) of 4.413 out of 6.0
on the Swiss scale, a sound pass. Taking lowest avgrade (=4.413) and the minimal observed values
for the other variables (team setting = base line (single man), guidance = 1, affiliation = 0) as a base-
line, the bachelor thesis grade btgrade would be predicted as 4.123 by our model. However, a 95%
prediction interval for that btgrade is [3.273, 4.971], a considerable range! A student with the minimal
average grade but with a supervisor with maximal guidance (=4) and maximal affiliation (=3) would
have a predicted btgrade of 4.970.with a 95% prediction interval [4.144, 5.797]. Therefore, the super-
visor may, at least according to our model, have a considerable effect on the final thesis grade, this
change in grade is qualitatively equivalent in the Swiss system to barely passing (4.0) and a good per-
formance (5.0).
Table 4. Regression analysis.
btgrade
Predictors Estimates SE p
(Intercept) 0.220 0.595 .712
avgrade 0.809 0.144 <.001
woman 0.059 0.071 .408
two men 0.026 0.102 .798
two women 0.270 0.095 .005
man and woman 0.258 0.146 .079
guidance 0.166 0.079 .037
affiliation 0.117 0.059 .048
Observations 187
R
2
/adjusted R
2
0.345/0.319
Figure 6. Boxplot BTgrade versus team setting.
12 F. STREBEL ET AL.

Post-hoc analysis
Analysing the constituents of the two constructed variables guidance and affiliation and searching for
a parsimonious model, besides avgrade only two variables remain significant: aims and sat_methods.
Figure 7 shows the BIC-plot of a search over all subsets of variables.
The model with avgrade, aims and sat_methods yields an adjusted R
2
of 0.304, see Table 5. The
normalplot of residuals shows some skewness but no outliers. The heteroscedasticity of residuals is
visible. The model coefficients are given in Table 5, variable avgrade is highly significant and aims and
sat_methods are significant.
3
This suggests that bachelor thesis supervisors can focus their efforts on supporting students with
the development of the aims of their study and on ensuring students have appropriate support with
the methodological aspects of their research.
Discussion
In our literature review, we have seen the tension between the learning and assessment goal of the
thesis, as a result of the learning goal asking for support and the assessment goal for more freedom
(Todd, Smith, and Bannister 2006; Delamont, Parry, and Atkinson 1998; Kleijn et al. 2012). In the
attempt to deal with this tension, previous researchers have made different arguments, some have
stated the importance of independent learning (e.g. Meeus, Van Looy, and Libotton 2004), while
others have emphasized the importance of personal caring and identification with the students
(e.g. Nir 2002; Menzies 1995).
Lee (2008), Mainhard et al. (2009) and Kleijn et al. (2012) have used a model with two dimen-
sions for the description of the student–supervisor relationship. One dimension considers the pro-
vision of guidance and direction, whereas the other dimension considered the proximity or
affiliation with the topic and/or the student. This model was the basis for our hypothesis, that
the bachelor thesis outcome is positively related to both the quality of guidance and to the
Table 5. Regression analysis for post-hoc analysis.
btgrade
Predictors Estimates SE std. Beta p
(Intercept) 0.804 0.778 .302
avgrade 0.727 0.142 0.381 <.001
aims 0.126 0.074 0.198 .091
sat methods 0.132 0.060 0.227 .029
Observations 186
R
2
/adjusted R
2
0.315/0.304
Figure 7. BIC-plot post-hoc model.
STUDIES IN HIGHER EDUCATION 13

degree of affiliation. We have been able to test this hypothesis thanks to the pooling of three
waves of student satisfaction surveys 189 observations from students about the satisfaction
with the supervision have been collected. The analysis used the achieved bachelor thesis grade
as the outcome of bachelor thesis supervision.
The model-building process in our paper has considered all possible models of all available cov-
ariates. A base model including (1) the average grade before beginning the bachelor thesis, (2) the
team setting (gender and team formation), (3) the guidance and (4) the affiliation as potential pre-
dictors turns out to be relatively stable and explains about a third of the variability of the final bache-
lor thesis grade. The inclusion of gender and team formation opened up an interesting field of future
enquiry. The expectation in the literature is that in undergraduate studies women have slightly better
average grades and less dispersion in the distribution of grades than their fellow male students
(Barrow, Reilly, and Woodfield 2009, 592). The final bachelor thesis grades in our study provide
further evidence for this pattern.
When considering a man or woman working alone on a thesis there is no significate difference in
observed outcomes. Turning to findings for teams, the evidence for the impact of different gender
compositions of teams are controversial. Some research has found women dominated teams to out-
perform teams dominated by men (Fenwick and Derrick 2001), while others have found the opposite
(Rogelberg and Rumery 1996; Apesteguia, Azmat, and Ibriberri 2012). Our findings suggest that a
team of a man and woman and two women teams have a favourable impact on the bachelor
thesis grade, whereas two men working together seem to have a negative dynamic. Mixed-
gender teams have been found to be favourable for group cohesion and team processes,
Table 6. Which factors have an impact on the bachelor thesis grade.
Base model Post-hoc analysis
Explanatory power Explanatory power
Variables Yes No Yes No
Guidance, including: X
.Support with aims definition X
.Support elaboration practical approach X
.Satisfaction with subject-specific support X
.Satisfaction with the methodological support X
.Satisfaction with the practice orientation X
Affiliation, including: X
.Availability in terms of reaction time X
.Perception of the meetings X
Controls
.Average grade before the thesis X X
.Age of student X X
.Team setting (gender and group formation) X X
.Curriculum/study programme X X
.Year that the thesis was written X X
Goodness of fitR
2
= 0.345;
adjusted R
2
= 0.319
R
2
= 0.315;
adjusted R
2
= 0.304
14 F. STREBEL ET AL.

whereas the impact on performance has remained controversial (Lee and Farh 2004;Bearand
Woolley 2011). Due to the limited number of observations, e.g. only 13 observations for man and
woman working together, we also do not wish or seek to interpret too much into these results
and discuss them with caution.
Kleijn et al. (2012), Mainhard et al. (2009) and Hon Kam (1997) have considered the extent of gui-
dance and affiliation and its impact on the outcome of the thesis. In our post-hoc analysis, we did not
focus on the extent but on where guidance and affiliation is provided. The BIC analysis, which takes
into account the variance reduction by the model, but also gives a penalty for the number of variables
in the model, identified the most parsimonious model as including, (1) the average grade before the
bachelor thesis, (2) the satisfaction with the support at the stage where the aims of the bachelor
thesis are agreed and (3) the satisfaction with the support with the project methodology. We sum-
marize the findings about what has an impact on the final bachelor thesis result in Table 6.
The initial base model might be seen as a summary of all available variables, as just a few variables
are not included. Besides the explanatory power of the variables for the bachelor thesis grade, there
might also be a relationship between them. Student attributes (controls) may influence the affiliation
of the supervisor, a strong affiliation might give direction and positively influence a student’s willing-
ness to accept guidance. The post-hoc model has less variables and does potentially overstate their
predictive power but on the other hand gives clearer hints where to draw attention in the supervision
process.
Conclusion
We conclude by discussing the contribution of our study to the academic literature, highlight its main
limitations, suggest future opportunities for research and close by drawing specific recommendations
for practice.
Contribution
We provide evidence that the average grade before the bachelor thesis, the satisfaction with the
support at the stage where the aims of the bachelor thesis must be defined and the satisfaction
with the support in methodology is the most parsimonious explanation for BSc thesis grade
achievement. Our findings suggest that guidance by the elaboration of a practical approach
and satisfaction in areas such as subject-specific support, practice orientation, availability and
the supervision meetings do not have an explanatory power for the quality of the thesis in our
study.
Coming back to our research question: ‘To what extent can the supervisor contribute to the
outcome of a thesis?’We can clearly state that the supervision has a measurable impact on the bache-
lor thesis result. Supervision and the capacity of the student are the only predictors that are left in the
most parsimonious model. The standardized beta coefficient, which measures the strength of the
effect of each individual independent variable to the dependent variable, shows for the capability
of the student 0.381 (average grade before the thesis) and for the support of the supervisor with
the aims definition 0.198 and for the methodological support 0.227. These findings suggest that
the supervision and the capability of the student have a comparable importance. Potentially, super-
vision can explain up to almost one grade of bachelor thesis grade.
Many authors have already stated that the student–supervisor relationship is important e.g. Lee
(2008) or Bloom et al. (2007). We included controls such as the capacity of the student in terms of
average grades before the bachelor thesis and we were not sure if supervision would still have an
explanatory power for the bachelor thesis grade. It still has and we are surprised of the extent of
the explanatory power. Therefore, our quantitative findings allow us to confirm the previous
studies that were often conducted with a qualitative approach or without control variables (e.g. Mar-
shall, Klocko, and Davidson 2017; Kleijn et al. 2012).
STUDIES IN HIGHER EDUCATION 15

Limitations
One limitation of our study is that the student knew their bachelor thesis grade when filling in the
questionnaire. This may have had an influence on the satisfaction evaluation by the student.
However, filling in the questionnaire after the bachelor thesis but before the publication of the
grade may also influence the satisfaction evaluation when a student wants to avoid a negative
impact on the bachelor thesis grade. Of course, the information of the survey would not have
been disclosed to the supervisors but this may not have avoided such an effect fully.
A further problem is the high correlation among the satisfaction ratings. The satisfaction with the
definition of the aims and the satisfaction with the support in methodology may be seen as proxy
variables for the other satisfaction ratings. And, in fact, the overall satisfaction would be the strongest,
but the least informative, predictor and thus has been left out of the models. Still the modelling
process clearly pointed to these two variables as the most important satisfaction variables.
Opportunities for future research
We see two natural continuations of our research. One approach would go deeper. What do the stu-
dents actually need when it comes to the aims definition and the methodological support? What
does a supervisor need to know if he or she would like to give valuable inputs in these two areas?
The other approach would go broader. Are there other elements that have a predictive power, e.g.
the type of topic or the engagement of the client?
Practical implications
The practical implications of our study are related to how supervisors should approach the supervi-
sion of bachelor theses. Our findings provide specific recommendations for where to place the
emphasis in supervision. We found evidence that effective supervision should focus on two elements:
(1) The support at the beginning of the thesis, when the aims of the bachelor thesis are defined.
(2) The support of students in the methodology by the supervisor.
Both elements have an almost equal weight in predictive power of the bachelor thesis grade. It
might be that these two elements are overweighed due to the reduction of variables and the
concept of affiliation and guidance should not be neglected.
The role of the supervisor, e.g. Holmberg (2006, 213) and the nature of the supervision is often
discussed. Whereas some authors emphasize the importance of independent learning, e.g. Meeus,
Van Looy, and Libotton (2004) others state the importance of the identification with the topic and
student and willingness to invest time and effort in the supervision, e.g. Menzies (1995) or Nir
(2002). Our findings could allow a balanced supervision. It suggests guidance where it has an
impact and, to a certain degree, laissez-faire and independent learning in other areas.
Our findings suggest that a convincing bachelor thesis outcome with a supervisor that appears
once at the beginning and then give his feedback after the work is completed can only have two
explanations: Either the supervisor gives both great support with the aims definition and the meth-
odology in a very short time or he, she has a good student.
Notes
1. BIC uses a penalty for the number of parameters in the model in addition to the residual sum of squares.
2. The model contains avgrade, year, curriculum, avage, composits, guidance and affiliation.
3. Note that the heteroscedasticity robust standard errors are larger and corresponding p-values are larger for aims
and sat_methods but significant at the 5% level.
16 F. STREBEL ET AL.

Acknowledgements
We would like to thank our reviewer and editor for a very positive and developmental review process, which greatly
benefited our paper. We would also like to thank Teresa Atzeni for her support in developing our dataset, Denise Maillard
for improving our tables and figures and Markus Freiburghaus for his support during the whole process.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Felix Strebel http://orcid.org/0000-0002-6606-5136
Stefan Gürtler http://orcid.org/0000-0002-9594-1078
Beat Hulliger http://orcid.org/0000-0001-5252-8606
Johan Lindeque http://orcid.org/0000-0002-8303-2715
References
Anderson, C., K. Day, and P. McLaughlin. 2006.“Mastering the Dissertation: Lecturers’Representations of the Purposes
and Processes of Master’s Level Dissertation Supervision.”Studies in Higher Education 31 (2): 149–68. doi:10.1080/
03075070600572017.
Andersson, G., and A. Person. 2002.Coaching och Handledning av Grupper. Lund: Studentlitteratur.
Apesteguia, J., G. Azmat, and N. Ibriberri. 2012.“The Impact of Gender Composition on Team Performance and Decision
Making: Evidence from the Field.”Management Science 58 (1): S78–93. doi:10.1287/mnsc.1110.1348.
Barrow, M., B. Reilly, and R. Woodfield. 2009.“The Determinants of Undergraduate Degree Performance: How Important Is
Gender?”British Educational Research Journal 35 (4): 575–97. doi:10.1080/01411920802642322.
Bear, J. B., and A. W. Woolley. 2011.“The Role of Gender in Team Collaboration and Performance.”Interdisciplinary Science
Reviews 36 (2): S146–53. doi:10.1179/030801811X13013181961473.
Bloom, J. L., A. E. Propst Cuevas, J. W. Hall, and C. V. Evans. 2007.“Graduate Students’Perceptions of Outstanding
Graduate Advisor Characteristics.”NACADA Journal 27 (2): 28–35. doi:10.12930/0271-95l 7-27.2.28.
Business, B. O., V. Elena, D. John, and B. Santina. 2013.“Project Work Writing.”Journal of International Education 6: 22–34.
Çetinkaya, E. Ş., and S. Yılmaz. 2017.“How BA Students Perceive Graduation Thesis Writing Process: A Qualitative Inquiry.”
Journal of Narrative and Language Studies 5 (8): 46–58.
Cook, R. D., and S. Weisberg. 1999.Applied Regression Including Computing and Graphics. New York: Wiley. doi:10.1002/
9780470316948
Delamont, S., O. Parry, and P. Atkinson. 1998.“Creating a Delicate Balance: The Doctoral Supervisor’s Dilemmas.”Teaching
in Higher Education 3 (2): 157–72. doi:10.1080/1356215980030203.
Egan, R., D. Stockley, B. Brouwer, D. Tripp, and N. Stechyson. 2009.“Relationships Between Area of Academic
Concentration, Supervisory Style, Student Needs and Best Practices.”Studies in Higher Education 34 (3): 337–45.
doi:10.1080/03075070802597143.
Fenwick, G. D., and J. N. Derrick. 2001.“Effect of Gender Composition on Group Performance.”Gender, Work &
Organization 8 (2): 205–25. doi:10.1111/1468-0432.00129.
Firestone, W. A., and S. Rosenblum. 1988.“Building Commitment in Urban High Schools.”Educational Evaluation and
Policy Analysis 10 (4): 285–99. doi:10.3102/01623737010004285.
Fox, J. 2015.Applied Regression Analysis and Generalized Linear Models. 3rd ed. Los Angeles: Sage.
Gill, S., and A. Reynolds. 1999.“Educational Expectations and School Achievement of Urban African American Children.”
Journal of School Psychology 37 (4): 403–24. doi:10.1016/S0022-4405(99)00027-8.
Gunneng, H., and E. Ahlstrand. 2002.Quality Indicators in Final Theses in Higher Education –A Comparative Pilot Study.
Linköping: Linkoöping University Faculty of Arts and Sciences.
Halse, C., and J. Malfroy. 2010.“Retheorizing Doctoral Supervision as Professional Work.”Studies in Higher Education 35 (1):
79–92. doi:10.1080/03075070902906798.
Han, Y. 2014.“An Analysis of Current Graduation Thesis Writing by English Majors in Independent Institute.”English
Language Teaching 7 (1): 120–7. doi:10.5539/elt.v7n1p120.
Holmberg, L. 2006.“Coach, Consultant or Mother: Supervisors’Views on Quality in the Supervision of Bachelor Theses.”
Quality in Higher Education 12 (2): 207–16. doi:10.1080/13538320600916833.
Hon Kam, B. 1997.“Style and Quality in Research Supervision: The Supervisor Dependency Factor.”Higher Education 34
(1): S81–103. doi:10.1023/A:1002946922952.
Housego, B. 1999.“Outreach Schools: An Educational Innovation.”Alberta Journal of Educational Research 45 (1): 85–101.
STUDIES IN HIGHER EDUCATION 17

Ives, G., and G. Rowley. 2005.“Supervisor Selection or Allocation and Continuity of Supervision: PhD Students’Progress
and Outcomes.”Studies in Higher Education 30 (5): 535–55. doi:10.1080/03075070500249161.
Janisch, C., and M. Johnson. 2003.“Effective Literacy Practices and Challenging Curriculum for At-Risk Learners: Great
Expectations.”Journal of Education for Students Placed at Risk 8 (3): 295–308. doi:10.1207/S15327671ESPR0803_1.
Kleijn, R.A.M. de, M. T. Mainhard, P. C. Meijer, A. Pilot, and M. Brekelmans. 2012.“Master’s Thesis Supervision: Relations
Between Perceptions of the Supervisor-Student Relationship, Final Grade, Perceived Supervisor Contribution to
Learning and Student Satisfaction.”Studies in Higher Education 37 (8): 925–39. doi:10.1080/03075079.2011.556717.
Lee, A. 2008.“How are Doctoral Students Supervised? Concepts of Doctoral Research Supervision.”Studies in Higher
Education 33 (3): S267–281. doi:10.1080/03075070802049202.
Lee, C., and J.-L. Farh. 2004.“Joint Effects of Group Efficacy and Gender Diversity on Group Cohesion and Performance.”
Applied Psychology 53 (1): S136–154. doi:10.1111/j.1464-0597.2004.00164.x.
Mainhard, T., R. van der Rijst, J. van Tartwijk, and T. Wubbels. 2009.“A Model for the Supervisor-Doctoral Student
Relationship.”Higher Education 58 (3): 359–73. doi:10.1007/s10734-009-9199-8.
Manathunga, C. 2007.“Supervision as Mentoring: The Role of Power and Boundary Crossing.”Studies in Continuing
Education 29 (2): 207–21. doi:10.1080/01580370701424650.
Marshall, S. M., B. Klocko, and J. Davidson. 2017.“Dissertation Completion: No Longer Higher Education’s Invisible
Problem.”Journal of Educational Research and Practice 7 (1): 74–90. doi:10.5590/jerap.2017.07.1.06.
Meeus, W., L. Van Looy, and A. Libotton. 2004.“The Bachelor’s Thesis in Teacher Education.”European Journal of Teacher
Education 27 (3): 299–321. doi:10.1080/0261976042000290813.
Menzies, T. 1995.“Teacher Commitment in Colleges of Applied Arts and Technology: Sources, Objects, Practices and
Influences.”Unpublished Doctoral diss., OISE/University of Toronto, Toronto.
Murray, R. 2011.How to Write a Thesis. 3rd ed. New York: Open University Press.
Nir, A. 2002.“School-based Management and its Effect on Teacher Commitment.”International Journal of Leadership in
Education 5 (4): 323–41. doi:10.1080/13603120210134616.
Rogelberg, S. G., and S. M. Rumery. 1996.“Gender Diversity, Team Decision Quality, Time on Task, and Interpersonal
Cohesion.”Small Group Research 27 (1): S79–90. doi:10.1177/1046496496271004.
Schulze, S., and A. C. Lessing. 2003.“Lecturers Experience of Postgraduate Supervision in Distance Education Context.”
South African Journal of Higher Education 17 (2): 159–68. doi:10.4314/sajhe.v17i2.25309.
Sharma, P. 2017.“Project Work Writing in BBS Fourth Years Course: Procedures and Practices.”The Journal of Nepalese
Business Studies 10 (1): 1–7. doi:10.3126/jnbs.v10i1.19128.
Shuell, T. J. 1988.“The Role of the Student in Learning from Instruction.”Contemporary Educational Psychology 13 (3): 276–
95. doi:10.1016/0361-476X(88)90027-6.
Solomon, C. 2007.“The Relationships among Middle Level Leadership, Teacher Commitment, Teacher Collective Efficacy,
and Student Achievement.”PhD diss., University of Missouri, Columbia, MI. Publication No. AAT 3322742.
Stock, J. H., and M. W. Watson. 2007.Introduction to Econometrics. 2nd ed. Boston: Pearson/Addison Wesley.
Sun, J. 2015.“Conceptualizing the Critical Path Linked with Teacher Commitment.”Journal of Educational Administration
53 (5): 597–624. doi:10.1108/JEA-05-2013-0063.
Todd, M., K. Smith, and P. Bannister. 2006.“Supervising a Social Science Undergraduate Dissertation: StaffExperiences
and Perceptions.”Teaching in Higher Education 11 (2): 161–73. doi:10.1080/13562510500527693.
Wisker, G. 2004.The Good Supervisor: Supervising Postgraduate and Undergraduate Research for Doctoral Theses and
Dissertation. Basingstoke: Palgrave Macmillan.
Wisker, G., K. Exley, M. Antoniou, and P. Ridley. 2008.Working One-To-One with Students: Supervising, Coaching, Mentoring,
and Personal Tutoring. London: Routledge.
18 F. STREBEL ET AL.

Appendix. Questionnaire
1 As and when required the supervising lecturer supported you
I agree I partly agree I partly disagree I disagree
with the definition of specific, realistic goals
with the elaboration of a practical approach
2 How do you assess the availability of the supervising lecturer
Οvery good (reaction time: Ø two working days)
Οgood (reaction time: Ø three working days)
Οsufficient (reaction time: Ø four working days)
Οinsufficient (reaction time: Ø more tHan four working days)
3 How do you characterize the meetings with your supervising lecturer?
(multiple choice possible)
Οhelpful Οnot very helpful
Οpatient Οnot very patient
Οmotivational Οnot very motivational
Οconvincing Οnot very convincing
Οcorrect Οincorrect
Οunderstanding Οnot very understanding
Οsetting of appropriate requirements Οsetting of not very
appropriate requirements
Οobjective Οbiased
Οfurther: ……….
4 How satisfied are you with the following aspects?
Very
satisfied
Satisfied Rather not
satisfied
Not
satisfied
with the subject-specific support of the supervising lecturer
with the methodological support of the supervising lecturer
with the practice orientation of the supervising lecturer
with the overall satisfaction
STUDIES IN HIGHER EDUCATION 19