You are on page 1of 26

Small Bus Econ (2020) 55:257–282

https://doi.org/10.1007/s11187-019-00143-y

The impact of entrepreneurship education on university


students’ entrepreneurial skills: a family embeddedness
perspective
Davide Hahn & Tommaso Minola & Giulio Bosio &
Lucio Cassia

Accepted: 5 February 2019 / Published online: 12 March 2019


# Springer Science+Business Media, LLC, part of Springer Nature 2019

Abstract To provide individuals with entrepreneurial paper takes advantage of quasi-experimental research on a
skills and prepare them to engage in entrepreneurial activ- sample of 427 university students who participated to two
ities, universities offer entrepreneurship education (EE) consecutive waves of the Global University Entrepreneur-
courses. However, the growing number of studies on EE ial Spirit Students’ Survey (GUESSS). The study finds that
impact offers mixed and apparently contradictory results. both types of EE contribute to students’ entrepreneurial
The present study contributes to this literature by indicat- skills; however, the impact of EE in compulsory courses is
ing the type of EE (elective vs. compulsory) and the contingent on students’ perceptions of parents’ perfor-
characteristics of students’ exposure to an enterprising mance as entrepreneurs.
family as two complementary boundary conditions that
contribute to explain the outcomes of EE. To do so, the Keywords University entrepreneurship . Student
entrepreneurship . Entrepreneurship education . Family
D. Hahn : T. Minola (*) : L. Cassia embeddedness . Global University Entrepreneurial Spirit
Department of Management, Information and Production Students’ Survey (GUESSS)
Engineering, University of Bergamo, Via Pasubio 7b,
24044 Dalmine, BG, Italy
e-mail: tommaso.minola@unibg.it JEL classifications I23 . J24 . L26

D. Hahn
e-mail: davide.hahn@unibg.it 1 Introduction

L. Cassia Individuals endowed with entrepreneurial skills, who


e-mail: lucio.cassia@unibg.it are able to identify and act upon new business opportu-
D. Hahn : T. Minola : G. Bosio : L. Cassia nities (Politis 2005), drive economic growth and societal
Center for Young and Family Enterprise, University of Bergamo, well-being by spurring the exploitation and commercial-
Via Salvecchio 19, 24127 Bergamo, BG, Italy ization of knowledge (Audretsch and Belitski 2013). By
providing individuals with these skills, universities can
G. Bosio make a relevant contribution to society (Audretsch
e-mail: giulio.bosio@unibg.it 2014; Laukkanen 2000). With the purposes of offering
students opportunities to engage in entrepreneurial
G. Bosio learning (Hahn et al. 2017), universities have been pro-
Department of Management, Economics and Quantitative
Methods, University of Bergamo, Via dei Caniana 2, viding entrepreneurship education (EE) extensively
24127 Bergamo, BG, Italy (Kuratko 2005; Mustar 2009; Rauch and Hulsink
2015). This has attracted the interest of a vibrant
258 D. Hahn et al.

research stream aimed at evaluating EE’s impact on attending elective courses explicitly self-select into EE
students (Lyons and Zhang 2018; Naia et al. 2014; (von Graevenitz et al. 2010) and generally have a stron-
Rideout and Gray 2013). ger specific interest for entrepreneurship as a career op-
However, judicious evaluations on the extent to which tion (Rauch and Hulsink 2015). In contrast, the partici-
EE can actually produce the valuable outcomes universi- pation to compulsory courses is mandatory for students
ties aim for have produced contrasting results that report to complete the studies they have chosen (Karimi et al.
positive (Karlsson and Moberg 2013; Rauch and Hulsink 2016) that require them to learn what entrepreneurship is
2015; Sànchez 2011), non-significant (Oosterbeek et al. about (Rauch and Hulsink 2015). Albeit these motiva-
2010), and even negative effects (Chang and Rieple 2013). tional aspects can specifically affect the results produced
This has been attributed both to empirical and conceptual by EE (Karimi et al. 2016; Martin et al. 2013; Rauch and
inaccuracies (Nabi et al. 2017). Specifically, relatively little Hulsink 2015), research on this is scant.
is known on the boundary conditions under which entre- Second, the paper considers a social dimension of EE
preneurship education is more effective (Bae et al. 2014; audience by taking into account students’ embeddedness
Martin et al. 2013). A nuanced appraisal of the contextual in a family (Aldrich and Cliff 2003; Sieger and Minola
dimensions under which EE and learning processes occur 2017) and particularly in an enterprising one (Jaskiewicz
is deemed as necessary to advance the scholarly knowl- et al. 2015; Minola et al. 2016a) for both theoretical and
edge effectively in this field (Nabi et al. 2017). In partic- practical reasons. From a theoretical perspective, as ac-
ular, with EE becoming more and more diffused through- knowledged by a flourishing stream of research (Criaco
out campuses, it tends to reach individuals who differ in et al. 2017; Laspita et al. 2012), an enterprising family
terms of demographics, prior entrepreneurial exposure, represents a social context that offers a unique learning
and commitment. The heterogeneous profiles of partici- environment (Hamilton 2011), which complements uni-
pants to entrepreneurship courses, whose outcomes have versity EE (Bae et al. 2014). From a practical standpoint,
been assessed in extant EE impact studies, represent a family-based businesses are ubiquitous, especially in the
potential reason for heterogeneous results, leading to con- European context, and a large fraction of university stu-
trasting findings and limited theoretical advancement dents have been exposed to entrepreneurship in their
(Martin et al. 2013; Naia et al. 2014). In fact, the audience families before receiving EE (Salvato et al. 2015). De-
of EE is deemed to assume a particular importance in spite some research efforts in such direction (e.g., Fayolle
determining the outcomes of EE observed by scholars and Gailly 2015; Peterman and Kennedy 2003; von
(Fayolle and Gailly 2015; Lyons and Zhang 2018; von Graevenitz et al. 2010), very little is known about the
Graevenitz et al. 2010) because the benefits of EE depend channels through which an exposure to an enterprising
to large extent on the learner (Béchard and Grégoire 2005). family works, also because such exposure is usually
For example, it has been suggested that entrepreneurial proxied by the measure Bhaving entrepreneurs as
skills are particularly difficult to teach unless the student parents.^ To overcome this limitation, this paper disen-
actively takes part to the learning process (Honig 2004; tangles exposure to an enterprising family into two di-
Mueller and Anderson 2014). Despite this intuition, extant mensions that the literature indicates as particularly rele-
literature is still lacking evidence that explores the hetero- vant (Criaco et al. 2017; Hamilton 2011; Mungai and
geneity of university students as a contingency that influ- Velamuri 2011): perceived parents’ performance in en-
ences learning and obtained skills from EE (Fayolle 2013; trepreneurship (PPE) and students’ actual working expe-
Nabi et al. 2017; Naia et al. 2014). rience for their family enterprise.
To address such limitation, this paper offers a deep In sum, this study explores the moderated model
examination of students’ profile in evaluating the out- where the relationship between EE in different courses
comes of EE. In particular, it follows recent recommen- (elective vs. compulsory) and entrepreneurial skills is
dations (e.g., Nabi et al. 2017; Naia et al. 2014; Martin contingent on students’ exposure to an enterprising fam-
et al. 2013) and explores two boundary conditions that, if ily (PPE and work experience). To test such framework,
taken together, describe the profile of participants to EE the study draws on a longitudinal sample of 427 univer-
and their predisposition to learn from educational inter- sity students from different study fields and education
ventions. First, the characteristics of students attending levels in eight European countries. The sample is obtain-
EE depend on the type of course attended (Karimi et al. ed by selecting students who participated to two consec-
2016; Nabi et al. 2017). More specifically, students utive waves (2013 and 2016) of the Global University
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 259

Entrepreneurial Spirit Students’ Survey (GUESSS) and Moberg 2013; McGee et al. 2009). For this purpose, in the
who did not attend EE courses before 2013. This empir- last decade, EE has been increasingly incorporated in
ical setting provides a valuable opportunity to rely on a higher education curricula at different levels (bachelor,
quasi-experimental design in order to explore the bound- master, and doctoral) and in different study fields (not
ary conditions of EE. Quasi-experimental designs allow only business and economics, but also social and natural
researchers to assess the effect of training interventions sciences and engineering; Mustar 2009; Shinnar et al.
by examining the differences pre- and post-intervention 2009; Souitaris et al. 2007). The assumption underlying
in students not randomly assigned to entrepreneurship such growing diffusion of entrepreneurship courses in
education and to a control group (Rauch and Hulsink universities is that students attending EE will learn entre-
2015). This approach has been recently recommended to preneurial skills (Hahn et al. 2017; Karlsson and Moberg
test the impact of EE more rigorously (Rideout and Gray 2013) of which they can take advantage during their future
2013) particularly in exploring the effect on skills and careers, as employees, self-employed workers, or entrepre-
abilities (Karlsson and Moberg 2013). neurs (Laukkanen 2000; Leitch et al. 2012).
The results reveal that, when separating elective and However, scholarly research that attempted to quan-
compulsory EE, only the former has a general positive tify the outcomes of EE has cast some doubts on this
effect on students’ entrepreneurial skills. On the other assumption, by questioning what might be the actual
hand, our result shows that the extent to which students impact of EE on university students’ entrepreneurial
benefit from taking EE in compulsory courses is contin- skills (Hahn et al. 2017; Nabi et al. 2017). Indeed, EE
gent to their prior exposure to an enterprising family; in impact studies show contrasting results: in some cases,
particular, we observe that only at higher level of PPE they show a positive effect of EE on entrepreneurial
students in compulsory courses benefit from EE to a skills (Karlsson and Moberg 2013; Sànchez 2011), but
significant extent. some other research reveals mixed (Volery et al. 2013),
The findings contribute to extant literature on how non-significant (Oosterbeek et al. 2010), and even neg-
universities can support the proliferation of entrepreneurial ative effects (Chang and Rieple 2013). While most stud-
individuals in an entrepreneurial society (Audretsch 2014; ies suggest a positive correlation between EE and entre-
Audretsch and Belitski 2013) by conceptualizing and ex- preneurial skills, those studies that comply with rigorous
ploring the role of individuals with their university and methodological standards to identify the effect of EE
social contexts as critical boundary conditions. The paper (e.g., using control groups or longitudinal research
also adds to literature on EE and university entrepreneur- designs; Rideout and Gray 2013) offer more controver-
ship by offering a more nuanced view of the contingencies sial results (Bae et al. 2014; Martin et al. 2013; Nabi et al.
under which university efforts to stimulate entrepreneurial 2017). Among these studies, Souitaris et al. (2007) found
learning among students are more likely to succeed. More- no significant effect of EE on students’ perceived ability
over, it provides further insights about the role played by to become entrepreneurs; the authors suggest that the
the family in the transmission of entrepreneurship by non-significant effect found in their study occurs due to
conceptualizing and empirically disentangling different the sample they have adopted, composed of elite univer-
channels for the influence of an enterprising family. sity students who had high self-confidence in entrepre-
neurship even before attending the course. Oosterbeek
et al. (2010) and Volery et al. (2013) advanced another
2 Literature review possible explanation for the non-significant effect of EE
on self-assessed entrepreneurial skills: after the course,
2.1 The controversial impact of entrepreneurship students have obtained a more realistic perspective on
education on university students’ entrepreneurial skills what it takes to perform entrepreneurial tasks, thus revis-
ing their self-confidence in entrepreneurial skills. In a
Universities can make a key contribution to societal devel- qualitative study, Chang and Rieple (2013) support this
opment by creating a context that equips individuals with argument. They found that, during an entrepreneurship
entrepreneurial skills (Audretsch 2014; Audretsch and course, students’ self-assessed skills even declined be-
Lehmann 2005; Shah and Pahnke 2014), which allow cause students were exposed to challenges that made
them to perform the variety of tasks required to identify them realize the inadequacy of their skills to face the
and act upon new business opportunities (Karlsson and hurdles encountered during the entrepreneurial process.
260 D. Hahn et al.

On a similar note, those studies using robust research transmitted by instructors in the classroom, in order to
designs that found a positive effect of entrepreneurship acquire entrepreneurial skills, students need to take a
courses—in the direction also recommended by Rideout proactive role in their learning process (Haase and
and Gray (2013)—contain some caveats and suggest Lautenschläger 2011; Honig 2004; Neck and Greene
that some boundary conditions might bias their results 2011). Beyond motivation, it has been recently shown
in favor of educational interventions (Lyons and Zhang that individuals would react differently to EE based on
2018; Rauch and Hulsink 2015). For example, Karlsson their pre-existing experiences (von Graevenitz et al.
and Moberg (2013), whose study revealed a positive 2010), which endow them with abilities, human and
effect of EE on self-assessed entrepreneurial skills, also social capital, and a legitimacy perception of entrepre-
found that such increase was remarkably lower for neurship (Peterman and Kennedy 2003). These help
students who had already mastered entrepreneurial skills students to learn and elaborate on EE effectively in order
before starting the course. Fayolle and Gailly (2015) to gain entrepreneurial skills that prove effective in
found that after a six-month entrepreneurship course, performing entrepreneurial tasks (Hahn et al. 2017). In
students increased their confidence in performing entre- this respect, the type of entrepreneurship course (Nabi
preneurial tasks, but such increase was remarkably low- et al. 2017; Rauch and Hulsink 2015) and the family
er for students with entrepreneurs in their families. embeddedness of the participants (Fayolle and Gailly
Rauch and Hulsink (2015), who adopted a quasi- 2015) have been indicated as valuable dimensions in
experimental research design to assess the effect of an exploring the audience as a boundary condition of EE.
entrepreneurship course, recognized that the positive
effect on perceived entrepreneurial skills they found 2.2 The type of entrepreneurship education course
could be upward biased by the elective nature of the as boundary condition
course. Accordingly, they raised some concerns about
the generalizability of their findings. EE can be offered to university students in different
In sum, the conflicting results of this research have types of courses. In some curricula, attending EE is
induced scholars to explore the reasons for such hetero- required to obtain the degree (i.e., compulsory courses),
geneity in EE impact (Fayolle and Gailly 2015; Hahn while in others, attending an entrepreneurship course
et al. 2017; Karimi et al. 2016). In other words, to move represents one of the various options available to stu-
the scholarly discussion ahead, rather than merely asking dents in order to gain course credits necessary for grad-
whether EE is effective or not, recent calls have indicated uation (i.e., elective courses). To date, in the attempt to
that the circumstances in which EE takes place matter, assess the impact of EE on university students, most
and have advocated exploratory studies on the boundary papers have analyzed the outcomes of single programs
conditions under which EE leads to the desired outcomes or have condensed different types of courses together. In
(Fayolle 2013; Martin et al. 2013; Nabi et al. 2017). In doing so, they could not explore the type of course
particular, scholars have argued that the contradictory (elective rather than compulsory) as a boundary condi-
outcomes revealed by EE impact studies can be ex- tion that can influence the results students obtain from
plained if the audience of entrepreneurship is better taken EE. A notable exception is offered by a study of Karimi
into account (Martin et al. 2013; Naia et al. 2014; Toutain et al. (2016). However, these authors did not employ
et al. 2017). control groups of students who did not take EE. Without
Literature on students’ entrepreneurial learning pro- such methodological refinement, the researchers could
cess and its outcomes (Dodd and Hynes 2012; Mueller not completely isolate the effect associated with EE in
and Anderson 2014; Neck and Greene 2011) provides the two types of courses. Recognizing this, instead, is
valuable suggestions in this respect (Leitch et al. 2012; important because EE in compulsory and elective
Walter and Dohse 2012; Wilson et al. 2007). It has been courses is addressed to different profiles of participants,
argued that, for students to actually benefit from partic- and such differences can be crucial in explaining the
ipation to EE, they need to be predisposed to and gen- contrasting outcomes observed in different impact stud-
uinely interested in entrepreneurship before attending ies (Karimi et al. 2016; Rauch and Hulsink 2015; Nabi
the course (Karimi et al. 2016; Mueller and Anderson et al. 2017).
2014). While to learn hard facts about business the First, while participants to elective EE show a spe-
learner could be a passive recipient of information cific preference towards learning entrepreneurship by
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 261

self-selecting directly in EE courses (Karimi et al. 2016), audience taking entrepreneurship courses (Hahn et al.
in compulsory courses, classes are made up of a mix of 2017; Lyons and Zhang 2018; Martin et al. 2013). This
students with varying degrees of interest in entrepre- paper considers family embeddedness and, more specif-
neurship. These students have self-selected into ically, exposure to an enterprising family as a boundary
studies—typically business studies—in which entrepre- condition that contributes to explaining the contradicto-
neurship courses complement the variety of subjects ry findings of EE impact studies. BExposure to an en-
taught (von Graevenitz et al. 2010). Some authors terprising family^ refers to the existence of an entrepre-
(Rauch and Hulsink 2015) even argue that participation neurial activity, in which the student’s family exerts
to EE in elective courses is generally conceived for significant influence in decision-making (Jaskiewicz
students who have interest in entrepreneurship as a et al. 2015; Minola et al. 2016a) and is therefore to some
possible career and who aim to acquire the skills to extent responsible for the development of entrepreneur-
become entrepreneurs, while EE in compulsory courses ial dynamics (Rogoff and Heck 2003). Family-
usually Bdoes not try to create entrepreneurs per se, but controlled firms are widely diffused in any world econ-
rather to teach participants what entrepreneurship is omy (Memili et al. 2015) and represent a key context
about^ (Rauch and Hulsink 2015: 200). As a result, affecting students’ social and cognitive development. In
students in elective EE courses are more likely to as- their family, university students can be exposed to en-
sume a more active posture in learning activities than trepreneurship before being enrolled in EE (Peterman
students forced to take a course, which makes it easier to and Kennedy 2003). At the same time, such exposure
observe positive outcomes of EE (Lyons and Zhang can be source of stress and negative perceptions about
2018; Rauch and Hulsink 2015). entrepreneurship (Criaco et al. 2017; Sieger and Minola
Second, there are reasons to believe that non- 2017). Unfortunately, most literature on student entre-
significant (Souitaris et al. 2007)—or even negative preneurship and EE treat exposure to entrepreneurship
(Chang and Rieple 2013)—influences of EE on stu- in the family as a dichotomic dimension (have vs. not
dents’ confidence in their entrepreneurial skills can be have), hence overlooking its nuanced nature.
explained by the elective nature of EE. In such context, Prior exposure to an enterprising family represents an
students might select the course confident in their entre- important contingency that influences the impact of EE
preneurial aptitudes and motivated to reinforce them, for (Fayolle and Gailly 2015; Peterman and Kennedy
example, because they had already received EE in 2003), because in entrepreneurship courses, students
which they learnt about entrepreneurship. However, evaluate the learning experiences in the classroom also
during an entrepreneurship course, they can be exposed on the basis of their previous knowledge and beliefs
to experiences and knowledge that challenge their be- about entrepreneurship (Chang and Rieple 2013; von
liefs about entrepreneurship, leading to decline in their Graevenitz et al. 2010). Both these can be influenced by
confidence in their skills (Chang and Rieple 2013). an enterprising family, which is inherently an effective
Also, the course might not be sufficient to satisfy their entrepreneurial learning environment (Hamilton 2011).
expectations (Hytti et al. 2010). Besides, exposure to an enterprising family might affect
In sum, it appears that the type of course represents a a student’s desire for an entrepreneurial career, in gen-
boundary condition that is worth exploring in order to eral and in the family firm (Block et al. 2013; Laspita
obtain a more nuanced understanding of the impact of et al. 2012). Also, through exposure to close individual
EE (Nabi et al. 2017). Accordingly, this paper tests the involvement in entrepreneurship, students are more like-
effect produced by EE in compulsory and elective ly to recognize and conform to entrepreneurial role
courses on university students’ perceived entrepreneur- models (Chlosta et al. 2012) and to perceive entrepre-
ial skills. neurship as a more legitimate career. Such desirability
and legitimacy perceptions could determine students’
2.3 Students’ family embeddedness as boundary roles as active rather than passive learners and hence
condition impact the effectiveness of EE in different types of
contexts (i.e., compulsory rather than elective).
Besides and together with the type of course, scholars Besides considering exposure to an enterprising fam-
assessing the impact of EE have been advised to con- ily as such, this paper builds on literature on the inter-
sider the specific entrepreneurial background of the generational transmission of entrepreneurship, which
262 D. Hahn et al.

invites scholars to account for the specific nature of such and taking advantage of the information provided in
exposure (Criaco et al. 2017; Mungai and Velamuri the classroom to further develop their entrepreneurial
2011). Also in the context of EE, Peterman and skills (Hahn et al. 2017; Lyons and Zhang 2018).
Kennedy (2003) and von Graevenitz et al. (2010) dis- Conversely, students with experience in their parents’
tinguish between positive and negative exposure to en- entrepreneurial activity might benefit less from EE
trepreneurship in investigating its effect on the results because, compared to other students who have never
obtained by high school students from an elective en- been exposed to entrepreneurship, they might already
trepreneurship course. Therefore, to gain a more nu- be aware of entrepreneurship and have less to learn
anced understanding of the influence of exposure to an (Karlsson and Moberg 2013; Peterman and Kennedy
enterprising family, it is important to disentangle the 2003; von Graevenitz et al. 2010). In light of these
specific mechanisms through which such exposure contrasting arguments, we include in our framework
works. students’ experience in parents’ family business as a
First, the paper considers the PPE. Considering this relevant dimension of their exposure to an enterpris-
aspect is important because the extent to which students ing family.
perceive their parents’ entrepreneurial activity as suc- The research framework that guides the study is
cessful has been shown to influence their inclination represented in Fig. 1. The effect of EE on university
towards entrepreneurship (Criaco et al. 2017; Mungai students’ entrepreneurial skills in elective and compul-
and Velamuri 2011). Indeed, successful self-employed sory courses is tested. How this effect is moderated by
parents can act as role models transferring values about students’ exposure to an enterprising family, which is
entrepreneurship to their children (Chlosta et al. 2012; disentangled into PPE and work experience in the par-
Hoffmann et al. 2015) and can use their influence as ents’ family business, is also tested.
close family members to stimulate students’ interest
towards entrepreneurship (Criaco et al. 2017). Having
self-employed parents as role models, students might 3 Data
become attracted to entrepreneurship, aim to learn more
about it, and end up doing so. Therefore, PPE could play 3.1 Research design
an important role in determining whether the exposure
to self-employed parents represents an obstacle or pre- This study relies on a quasi-experimental research
disposes students towards actively participating and design, whose adoption has been encouraged to
learning from EE. rigorously identify the actual impact of educational
Second, the paper considers the experience obtain- interventions on students who have been exposed to
ed by students in their parents’ entrepreneurial activ- EE (Rauch and Hulsink 2015; Rideout and Gray
ity. By working in their parents’ firm, students have 2013; Volery et al. 2013). A quasi-experimental
the opportunity to accumulate hands-on business ex- design is particularly appropriate to test the effect
periences (Criaco et al. 2017), which can influence produced by EE interventions (Karlsson and Moberg
the extent to which they learn from EE (Fayolle and 2013), when students cannot be randomly assign to
Gailly 2015). Such first-hand entrepreneurial experi- the educational intervention (EE), as in true experi-
ence can assist them in better absorbing, processing, ments (Rauch and Hulsink 2015). In a quasi-

Fig. 1 Graphical representation of the research framework


The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 263

experimental design, students are assigned to a classroom. The students reached by the survey be-
group that is not exposed to any educational inter- long to different fields of study (i.e., business and
vention (i.e., the comparison or control group) and economics, natural and social sciences) and different
to groups that are exposed to different educational education levels (e.g., undergraduate, graduate).
interventions. In this study, the intervention consists After the data collection carried out in 2016, the
in elective and compulsory EE. Accordingly, stu- GUESSS team built a longitudinal dataset by matching
dents are clustered into three different groups: (i) a the responses provided by students who participated to
control group of students who have not been ex- the two consecutive surveys in order to build a longi-
posed to EE, neither before 2013 nor between tudinal database. The complete longitudinal GUESSS
2013 and 2016; (ii) students who have not been 2013–2016 dataset records the answers from 1383
exposed to EE before 2013 but received EE in students from 21 countries. In order to reduce the cul-
compulsory courses between 2013 and 2016; and tural heterogeneity in the sample, this paper focuses its
(iii) students who have not been exposed to EE analysis only on European countries. After excluding
before 2013 but who received EE in elective courses those countries whose number of complete responses
between 2013 and 2016. The effect of educational was below 5, the sample contains students from the
interventions is identified by observing the changes following countries: Austria, Germany, Switzerland,
of the outcome variables of interest (i.e., entrepre- Italy, Spain, Estonia, Hungary, and Poland. Table 1
neurial skills) after the exposure to EE and in ab- reports the breakdown of students in the final sample
sence of EE. Moreover, as described in the section by country.
below, an appropriate set of control variables is The longitudinal nature of the dataset provides a
introduced to account for initial differences between valuable opportunity to identify the effects caused by
groups, thereby mitigating the bias introduced by the participation to entrepreneurship courses. To further
non-random selection of individuals into different take advantage of this aspect, the analysis included only
groups (Rauch and Hulsink 2015). questionnaires of students who in the 2013 survey re-
ported not having attended EE before. The sample also
3.2 Sample excluded the few students who attended EE in both
elective and compulsory EE. By doing so, it was possi-
The empirical analysis of this paper is based on a ble to compare the change in entrepreneurial skills from
sample of university students who participated to two 2013 to 2016 between students who participated to
consecutive waves of the Global University Entre- (compulsory or elective) EE for the first time in the
preneurial Spirit Students’ Survey (GUESSS)1: in period from 2013 and 2016 and students who had never
2013 and in 2016. The Swiss Research Institute of received EE. Finally, the sample excluded respondents
Small Business and Entrepreneurship at the Univer- for which it was not possible to build the variables of
sity of St. Gallen (KMU-HSG) in Switzerland coor- interest. The final sample size consists of 427 student
dinates the GUESSS project at a global level. At the entrepreneurs.
country level, a representative of the project has the
task of distributing the survey and collecting the data
Table 1 Breakdown of
from the universities of that country. Students who the final sample by Country No. of students
answered the survey were reached through a non- country
random process in which universities were autono- Austria 35
mous in defining the breadth of classes and schools Estonia 8
involved in the survey. Students were invited to an- Germany 151
swer the questionnaire through different channels, Hungary 51
such as in social networks, via email, or in the Italy 48
Poland 24
1
A full description of the GUESSS project is available at the website Spain 71
www.guesssurvey.org. Several works based on the GUESSS project
have already been published in entrepreneurship journals: see, for Switzerland 39
example, Bergmann et al. (2016); Minola, Donina, and Meoli (2016); Total 427
and Sieger and Minola (2017).
264 D. Hahn et al.

3.3 Measures 3.3.3 Moderation variables

3.3.1 Dependent variables In the moderated model, the relationship between EE


and entrepreneurial skills is conditional on students’
The dependent variable is entrepreneurial skills16. To exposure to an enterprising family. In a first step, this
identify and act upon new business opportunities, entre- is described through the raw measure self-employed
preneurs need the skill to perform a variety of tasks: parents, a binary variable coded as 1 if at least one of
identifying the need for new products or services, de- student’s parents was self-employed, according to stu-
signing and commercializing these products or services dents’ responses to the 2013 survey.
as innovation, building networks to access external ex- In a second step, instead of considering the crude
pertise and resources, and persuading others to believe measure of self-employed parents, students’ expo-
in their business idea (Karlsson and Moberg 2013; sure to an enterprising family is unpacked; two
McGee et al. 2009). Accordingly, to measure students’ moderating variables are built that together capture
entrepreneurial skills in 2016, they were asked to eval- the mechanisms through which exposure to an en-
uate their competence in performing the following tasks terprising family produces its influence. The first of
on a 7-point Likert scale (1 = very low competence, 7 = these two variables is PPE, measured through the
very high competence) inspired by extant self-efficacy scale adopted in Criaco et al. (2017), whose five
and creativity scales (Chen et al. 1998; Kickul et al. items together constitute a multidimensional and
2009; Zhou and George 2001): (1) BIdentifying new subjective measure of performance commonly
business opportunities^; (2) BCreating new products employed in family business research (Eddleston
and services^; (3) BManaging innovation within a firm^; et al. 2008). The creation of this variable relies on
(4) BBeing a leader and communicator^; (5) BBuilding the responses of students with self-employed parents
up a professional network^; (6) BCommercializing a to the 2013 survey, who assessed the performance of
new idea or development^; and (7) BSuccessfully man- their parents’ entrepreneurial activity on a 7-point
aging a business.^ The items of this scale loaded on a Likert scale (1 = worse; 7 = better) relative to their
single factor, while the reliability of the scale was tested competitors in terms of five dimensions (i.e., sales
by computing Cronbach’s alpha, which resulted in 0.91, growth, market share growth, profit growth, job
above the recommended value of 0.7 (Nunnally 1978). creation, and innovativeness). The items of this
The scores of the seven items were thus averaged to scale load on a single factor and are reliable (α =
compute the dependent variable, entrepreneurial 0.82). For the regression equations, PPE is standard-
skills16. ized, a procedure commonly employed to avoid
multicollinearity issues when continuous variables
are multiplied to create interaction terms (Aiken
3.3.2 Independent variables and West 1991). For students without self-
employed parents, PPE is multiplied by 0, having
To test the relationship between EE and entrepreneur- no influence on the dependent variables of the re-
ial skills, in a first step, the independent variable EE gression equation. Second, the variable experience
is built, a dummy variable that takes the value of 1 if in the parents’ entrepreneurial activity is built, by
the student attended EE in the period between 2013 measuring by the number of months students
and 2016, as reported in the 2016 GUESSS survey (0 worked in their parents’ entrepreneurial activity (0
otherwise). In a second step, the EE variable is for students without self-employed parents), as re-
unpacked to distinguish between elective EE and ported in the 2013 survey. As for PPE, experience is
compulsory EE. Both are measured through binary standardized before being interacted with the inde-
variables that take the value of 1 if the students pendent variables.
attended EE in a compulsory or elective course, re-
spectively, in the period between the two consecutive 3.3.4 Control variables
waves of the survey (0 otherwise). These two vari-
ables were also built by using information collected First, to observe the change of students’ entrepre-
in the 2016 wave of the GUESSS survey. neurial skills due to the EE, the lagged value of
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 265

entrepreneurial skills (entrepreneurial skills13), ob- joining their current university, a dummy variable is
tained from students’ assessment in the 2013 wave included that captures students’ reason for choosing
of the survey,2 is included as control variable to the their university. It takes the value of 1 if the student
regression equations. To build this variable, the stated in the 2013 wave of the survey that the strong
same seven items used in the 2016 wave of the entrepreneurial reputation of the university was the
survey are used, obtaining a Cronbach’s alpha of most important reason behind the choice of the univer-
0.90. sity (0 otherwise).
Second, taking advantage of students’ responses to Third, a set of demographic and student-related var-
the 2013 GUESSS survey, a range of individual-level iables are controlled for, in line with extant EE impact
variables are controlled as they might be correlated studies (e.g., Bergmann et al. 2016; Criaco et al. 2017;
with students’ interest and predisposition towards en- Hahn et al. 2017). As men have been found to be more
trepreneurship and affect the results they obtain from attracted towards entrepreneurship than women (Grilo
taking (or not taking) entrepreneurship courses. Stu- and Thurik 2008), gender is controlled for and the
dents’ entrepreneurial intentions, measured through variable male is created, coded with a dummy where 1
the items developed by Liñán and Chen (2009), are indicates male and 0 female. Since age may influence
controlled for. Since items load on a single factor and individuals’ predisposition to learn (Minola et al. 2014;
the data confirm the reliability of the scale (Cronbach’s Wilson and McCrystal 2007) and take advantage of EE
alpha = 0.97), entrepreneurial intention was computed (Peterman and Kennedy 2003), the model contains age,
as the average of the six items of the scale, which measured in years, as a control variable. As generic
students evaluated in the 2013 wave of the survey on education may contribute to entrepreneurial skills
a 7-point Likert scale. Students’ nascent entrepreneur (Hamilton 2011), students’ level of education is also
status in 2013 is also controlled for. Following previ- controlled for. To do so, a dummy variable is created,
ous research (Bergmann et al. 2016; Karlsson and coded 1 for undergraduate (bachelor) students and 1 for
Moberg 2013; McGee et al. 2009; Souitaris et al. graduate (master) students; more advanced levels of
2007), nascent entrepreneurs are defined as those stu- studies (PhD/doctorate, faculty/post doc, or MBA/
dents who were actually trying to start their entrepre- executive education) represent the reference category.
neurial activity and had actually performed some con- The study also controls for students’ field of study, since
crete gestation activities to do so (Bird et al. 2012; interest and aptitudes towards entrepreneurship might
Carter et al. 1996). The control variable nascent entre- vary across students of different disciplines (Shinnar
preneurial activity was thus coded as a binary variable. et al. 2009). To do so, the fields of study are grouped
It takes the value of 1 for students who in the 2013 in three broad areas: (1) business and economics, (2)
wave of the survey stated that they were Btrying to start natural sciences, (3) social sciences. A dummy variable
your own business/to become self-employed^ and had is created for each study area with the exception of
also completed at least one of the following gestation Bsocial sciences,^ which has been considered as the
activities: written a business plan; started product/ reference category. The study also controls for the extent
service development; started marketing or promotion to which students perceive entrepreneurship to be en-
efforts; purchased material, equipment, or machinery couraged and supported at their university. To do so, the
for the business; attempted to obtain external funding; variable climate is included, which was measured on the
applied for a patent, copyright, or trademark; regis- 3-item, 7-point Likert scale used by Bergmann, Geissler,
tered the business; sold product or service. To control Hundt, and Grave (2018; α = 0.88).
for students’ interest towards entrepreneurship before Fourth, students’ national context is controlled
for, by dividing the European countries to which
2
the respondents belong into three groups of coun-
The inclusion of the lagged values of the dependent variables raises
tries with relative cultural and linguistic similarity
concerns related to the potential autocorrelation in the error terms. The
Cumby and Huizinga (1992) statistic is applied to test for autocorrela- (cf. Liñán and Fernandez-Serrano 2014): Central
tion of order 1 in the residuals under the null hypothesis of no auto- Europe (Austria, Germany, Switzerland), South Eu-
correlation. The statistic rejects the null hypothesis of entrepreneurial rope (Italy, Spain), and East Europe (Estonia, Hun-
skills being serially uncorrelated (χ2 = 169.062, p < 0.001) and shows
that serial correlation between the dependent variable and its lagged gary, and Poland). A dummy variable is created for
value is correctly specified at degree 1. East Europe (equal to 1 if the student is from East
266 D. Hahn et al.

Europe, 0 otherwise) and Central Europe (equal to 1 Model 1 contains only control variables. As ex-
if the student is from Central Europe, 0 otherwise), pected, students’ entrepreneurial skills in 2016 are
with South Europe representing the category of positively associated with the students’ entrepre-
reference. neurial skills in the first wave of the survey. The
Finally, since the students of our sample are positive and statistically significant coefficient asso-
nested in different universities, we controlled for ciated with entrepreneurial intention in 2013 indi-
unobserved university characteristics, which might cates that students with higher entrepreneurial inten-
affect the dependent variable entrepreneurial skills. tions on average experience a greater increase in
To do so, we introduced university fixed effects. their entrepreneurial skills. In addition, the positive
and significant coefficient shows that, all else being
equal, older students on average had higher entre-
4 Results preneurial skills in 2016. Similarly, male students
report, on average, higher levels of entrepreneurial
The mean values, standard deviations, and correlations skills than females. Interestingly, controlling for uni-
of all variables used in the empirical analysis are report- versity effects, students of business and economics
ed in Table 2, while Table 3 displays descriptive statis- and students who chose their school because of its
tics of the sample, split into students who never received entrepreneurial reputation increase their entrepre-
EE, who received EE in elective courses, and who neurial skills to a greater extent.
received EE in compulsory courses. Model 2 introduces the independent variable EE (tak-
The means reported in Table 2 show that 13.3% ing together elective and compulsory courses), which
of the students in the sample took EE in compulsory results not significant. However, in Model 3, which
courses, while 10.3% took EE in elective course. On distinguishes between course types, the coefficients of
average, the self-assessed entrepreneurial skills have EE in elective courses (β = 0.311, p < 0.05) yield positive
a slight decrease from 2013 (mean = 4.3) to 2016 and significant results, while the coefficient of compul-
(mean = 4.1). Almost two-thirds of the students were sory EE is positive albeit not significant. This suggests
undergraduates, and almost half of the sample stud- that the outcomes of EE can be better understood by
ies in the natural sciences discipline. Almost one- disentangling the impact of elective and compulsory
fourth of the sample has self-employed parents. The courses. Overall, the results in Models 2 and 3 suggest
presence of self-employed parents is not significant- that the impact of EE in elective courses goes beyond the
ly correlated with students’ participation to elective school effect; in contrast, the impact of EE (with elective
and compulsory courses. and compulsory courses taken together) and of EE in
In the following regression analyses, the effect of EE compulsory courses cannot be disentangled from the
on entrepreneurial skills and the moderation of students’ school effect in this specification.4
exposure to an enterprising family are tested by using
moderated hierarchical OLS regression (Cohen and 4
In support to this idea, the model was also run with robust standard
Cohen 1983). Since the data contain individual-level errors, not clustered by university. In this model, the coefficient of EE
was positive and statistically significant (β = 0.203, p < 0.10). More-
observations nested within universities, as in Minola
over, we run a model without university fixed effects, and it was found
et al. (2016b), standard errors were clustered by univer- that the coefficients of EE in elective (β = 0.262, p < 0.05) and
sity in order to deal with potential correlation of errors at compulsory (β = 0.224, p < 0.10) courses were both positive and
the university level (Rabe-Hesketh and Skrondal 2008). statistically significant. To gain further insights into the different effects
produced by EE in elective and in compulsory EE, we also separately
Results are displayed in Table 4. It was verified that focused on compulsory EE vs. no EE and on elective EE vs. no EE: in
variance inflation factors were all below 10 to ensure the first case, entrepreneurial skills were regressed on EE in compul-
that multicollinearity does not cause major concerns to sory courses, leaving students with no EE as baseline and excluding
from the sample students who received EE in elective courses; in the
the analyses (Kennedy 2008).3 latter case, entrepreneurial skills were regressed on EE in elective
courses, leaving students with no EE as baseline and excluding from
3
VIF values above 5 were found only for the dummy variables the sample students who received EE in compulsory courses. In both
bachelor and master because together they cover 96% of the sample cases, the university fixed effects were included. In these specifica-
and are thus highly correlated (− 0.90). By dropping the variable tions, the coefficient of EE in elective (β = 0.358, p < 0.05) remains
bachelor and keeping undergraduate students as reference category, positive and significant, while the coefficient of compulsory EE re-
VIFs remained below 2 with results substantially unchanged. mains positive albeit not significant.
Table 2 Construct-level descriptive statistics and correlation of constructs

Variable Means SD (1) (2) (3) (4) (5) (6) (7) (8)

1. Entrepreneurial skills16 4.112 1.290 1


2. Elective EE 0.103 0.304 0.1064 1
3. Compulsory EE 0.133 0.340 0.0736 − 0.133 1
4. Self-employed parents 0.222 0.416 − 0.0046 − 0.0146 0.0053 1
5. PPE 3.848 0.522 0.0227 − 0.0473 0.0549 0.0171 1
6. Experience 1.309 7.973 0.0262 − 0.0306 − 0.0291 0.3073 0.0499 1
7. Nascent entrepreneurial activity 0.035 0.184 0.1935 − 0.0228 0.0747 − 0.0715 − 0.0126 − 0.0314 1
8. Entrepreneurial intention 3.073 1.773 0.4313 0.0629 0.1213 0.0432 0.0149 − 0.0783 0.3309 1
9. Entrepreneurial skills13 4.276 1.262 0.6292 0.1039 0.053 0.0449 0.0406 − 0.0012 0.2004 0.5305
10. Age 25.251 3.526 0.0224 − 0.1138 − 0.1061 − 0.0093 − 0.0878 0.0115 0.0009 − 0.1249
11. Male 0.405 0.491 0.1738 0.0498 − 0.0153 − 0.1318 0.0072 0.0536 0.1016 0.143
12. Bachelor 0.611 0.488 0.0697 0.0491 0.087 − 0.0701 − 0.0103 − 0.0806 0.0478 0.2479
13. Master 0.342 0.475 − 0.0455 − 0.0332 − 0.0797 0.0655 0.059 0.0514 − 0.0571 − 0.2154
14. Business and economics 0.246 0.431 0.1398 0.1284 0.0957 0.0476 0.1367 − 0.0078 0.0387 0.0984
15. Natural sciences 0.525 0.500 0.0072 − 0.0784 0.0841 − 0.0996 − 0.0635 0.0499 − 0.0221 0.0955
16. Entrepreneurial reputation 0.030 0.172 0.1176 0.0296 0.0908 0.0363 0.0513 0.2054 0.2623 0.103
17. Climate 3.806 1.441 0.2319 0.1455 0.0815 0.012 0.0316 − 0.0027 0.1052 0.1888
18. South Europe 0.279 0.449 0.0862 0.0127 − 0.0443 − 0.0562 − 0.0514 − 0.0694 0.0516 0.2679
19. East Europe 0.194 0.396 − 0.0629 0.106 0.3118 − 0.0351 0.0375 − 0.0324 0.0027 0.0461
20. Central Europe 0.527 0.500 − 0.0276 − 0.0954 − 0.2074 0.0783 0.0164 0.088 − 0.0485 − 0.2772

Variable (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20)

1. Entrepreneurial skills16
2. Elective EE
3. Compulsory EE
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family...

4. Self-employed parents
5. PPE
6. Experience
7. Nascent entrepreneurial activity
8. Entrepreneurial intention
9. Entrepreneurial skills13 1
10. Age − 0.0487 1
11. Male 0.1247 0.0266 1
12. Bachelor 0.0565 − 0.2625 0.0906 1
13. Master − 0.0213 0.173 − 0.0619 − 0.9038 1
267
268 D. Hahn et al.

(20)
In Model 4, the variable self-employed parents and its

1
interaction with EE (in elective and compulsory courses

− 0.5184
grouped together) are added to test moderation. This
variable’s coefficient is not statistically significant. Model
(19)

1
5 further interacts the term self-employed parents sepa-
rately with EE in elective and compulsory courses, and it

− 0.3053
− 0.656
is observed that the coefficients of the interaction terms
(18)

1 compulsory EE × self-employed parents and elective EE ×


self-employed parents are both not statistically significant.
− 0.018

− 0.021
0.0469
Models 6 and 7, instead of having self-employed
(17)

parents as moderator, introduce, respectively, PPE


1

and experience, as well their interactions with EE, as


− 0.0493

− 0.0505

more fine-grained proxies of students’ exposure to an


0.1196
0.087

enterprising family: EE × PPE is not statistically sig-


(16)

nificant; also the interaction between EE and experi-


ence is not significant. Models 8 and 9 then consider
− 0.2163
0.1713
0.0049

0.0897

the interaction of PPE and experience with EE in


0.062
(15)

elective vs. compulsory courses. Model 8 tests the


1

moderation of PPE, distinguishing EE in elective vs.


− 0.0194
− 0.0517
− 0.5999

compulsory courses. Here, the coefficient of the term


0.0571

0.0618
0.102
(14)

compulsory EE × PPE is positive and significant (β =


1

0.240, p < 0.10). Model 9, which tests the moderation


effect of experience, shows that the interaction terms
− 0.0128
− 0.0157

− 0.0737
− 0.1158
− 0.1135

− 0.117
0.1589

with elective and compulsory EE are not statistically


(13)

significant. Finally, Model 10 displays the full model.


It shows that EE in elective courses is positive and
− 0.1494
− 0.0088
− 0.0265

statistically significant (β = 0.436, p < 0.001). This


0.1542

0.0639
0.1164

N = 427. Absolute values of pairwise correlations above 0.10 significant at p < 0.05
0.11
(12)

implies that having taken EE in elective courses is


associated with an increase by 0.345 of the standard
deviation of students’ entrepreneurial skills16. More-
− 0.0171

− 0.0317
− 0.0684
0.1934
0.0759
0.1265
0.1041

over, even though the direct effect of EE in compul-


(11)

sory EE is not statistically significant per se, its inter-


actions with self-employed parents and PPE are sig-
− 0.2198
− 0.2178
− 0.1525
− 0.1117

nificant (respectively, β = − 0.684, p < 0.10 and β =


0.3112
− 0.12
− 0.09
(10)

0.452, p < 0.05). In order to discuss the practical im-


pact of PPE in determining the results of EE in com-
pulsory courses, the marginal effect of compulsory EE
− 0.0029
− 0.0094
0.0299

0.3363
0.0131
0.1295

0.111

on the dependent variable was computed, conditional


(9)

on the level of PPE. The results of this exercise indi-


cate that the effect of compulsory EE is not significant
at lower (i.e., at 1 standard deviation below the mean
16. Entrepreneurial reputation

value of PPE) and average levels of PPE; it becomes


14. Business and economics

positive and significant only at higher levels of PPE


(β = 0.646, p < 0.01; i.e., at 1 standard deviation above
Table 2 (continued)

15. Natural sciences

20. Central Europe


18. South Europe

the mean value of PPE). This implies that, ceteris


19. East Europe

paribus, at higher levels of PPE, EE in compulsory


17. Climate

courses increases students’ entrepreneurial skills by


Variable

0.501 of the standard deviation of students’ entrepre-


neurial skills16.
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 269

Table 3 Descriptive statistics for the sample split on the basis of participation to EE

Variables EE = 0 (N = 324) Elective EE = 1 (N = 44) Compulsory EE = 1 (N = 57)

Mean SD Mean SD Mean SD

Entrepreneurial skills16 3.995 1.306 4.516 1.208 4.353 1.211


Self-employed parents 0.213 0.410 0.204 0.408 0.228 0.423
PPE 3.847 0.548 3.775 0.499 3.921 0.334
Experience 1.395 8.726 0.590 2.060 0.719 4.920
Nascent entrepreneurial activity 0.028 0.164 0.022 0.150 0.070 0.258
Entrepreneurial intention 2.921 1.760 3.401 1.859 3.620 1.608
Entrepreneurial skills13 4.168 1.303 4.662 0.970 4.446 1.204
Age 25.574 3.588 24.068 3.187 24.298 3.023
Male 0.395 0.490 0.477 0.505 0.386 0.491
Bachelor 0.589 0.493 0.681 0.471 0.719 0.453
Master 0.358 0.480 0.295 0.461 0.245 0.434
Business and economics 0.200 0.401 0.409 0.497 0.351 0.481
Natural sciences 0.521 0.500 0.409 0.497 0.631 0.487
Entrepreneurial reputation 0.015 0.123 0.045 0.210 0.070 0.258
Climate 3.672 1.413 4.424 1.426 4.105 1.426
South Europe 0.287 0.453 0.295 0.461 0.228 0.423
East Europe 0.123 0.329 0.318 0.471 0.509 0.504
Central Europe 0.590 0.493 0.386 0.492 0.263 0.444

To facilitate the interpretation of these results, squares approach with instrumental variables (Bascle
Fig. 2 plots the relationship between EE in compul- 2008) was used. To be eligible as an instrument, a
sory courses and entrepreneurial skills for students variable should be significantly correlated with the po-
who report lower and higher levels of PPE (i.e., 1 tentially endogenous variable (i.e., EE) but not with the
standard deviation below and above the mean val- dependent variable (i.e., entrepreneurial skills 16 ;
ue). The figure shows that the students’ self-assessed Kennedy 2008). Students’ participation to EE was in-
entrepreneurial skills sharply increase only at higher strumented with the variation in percentage points be-
levels of PPE. Overall, these results underscore the tween 2008 and 2013 in the employment rate of high-
importance of taking into account PPE in assessing the tech and knowledge-intensive sectors in the same re-
compulsory EE-entrepreneurial skills relationship. gions of the universities (NUTS 2 level) included in the
sample, as reported in Eurostat. To measure the instru-
mental variable at the regional level, a period of multiple
4.1 Robustness checks
years before students answered to the first survey was
considered; doing so provides a measure of the average
First, since students in the sample have selected them-
local influences during the years preceding students’
selves into elective EE or into schools that offer com-
selection of university.6 It is reasonable to assume that
pulsory EE, it is necessary to mitigate the concern that
this variable predicts students’ self-selection into elec-
the results are driven by unobservable variables that
tive EE or into schools that offer EE as compulsory, but
drive such selections.5 To do so, a two-stage least
that it is exogenous to students’ increase of entrepre-
5
Applying the Oster (2017) procedure (cf. Lyons and Zhang 2018), neurial skills. In fact, even though employment
support is found to the need to control for potential endogeneity due to
6
unobservables (correlated both to selection into EE and to outcomes of To measure variables at the local level that might affect students’
EE). Additionally, both the Wu-Hausman test and Durbin-Wu- entrepreneurial learning, periods of multiple years have already been
Hausman statistics reject the null hypothesis that the EE indicator is used in empirical models aimed at quantitatively assessing the impact
not exogenous. of EE (cf. Hahn et al. 2017).
Table 4 Linear regression with entrepreneurial skills16 as dependent variable
270

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10

Entrepreneurial skills13 0.545*** 0.540*** 0.541*** 0.541*** 0.542*** 0.540*** 0.540*** 0.544*** 0.541*** 0.548***
(0.062) (0.062) (0.063) (0.062) (0.065) (0.062) (0.061) (0.064) (0.063) (0.067)
Nascent entrepreneurial activity 0.119 0.114 0.141 0.0951 0.128 0.119 0.149 0.173 0.155 0.189
(0.233) (0.230) (0.228) (0.240) (0.233) (0.239) (0.236) (0.243) (0.231) (0.248)
Entrepreneurial intention 0.0722** 0.0720** 0.0711** 0.0747** 0.0766** 0.0716** 0.0715** 0.0665* 0.0726** 0.0723**
(0.033) (0.033) (0.034) (0.035) (0.034) (0.034) (0.033) (0.036) (0.034) (0.036)
Age 0.041** 0.041** 0.042** 0.041** 0.043** 0.042** 0.041** 0.044** 0.043** 0.046**
(0.019) (0.019) (0.019) (0.019) (0.019) (0.019) (0.019) (0.020) (0.019) (0.020)
Male 0.213** 0.210** 0.204** 0.201** 0.196** 0.210** 0.201** 0.197** 0.198** 0.166*
(0.089) (0.089) (0.088) (0.088) (0.087) (0.090) (0.091) (0.088) (0.090) (0.091)
Bachelor 0.034 0.006 0.017 0.000 0.017 − 0.004 − 0.007 0.025 0.025 0.044
(0.190) (0.191) (0.195) (0.194) (0.200) (0.194) (0.185) (0.201) (0.192) (0.205)
Master 0.099 0.064 0.080 0.069 0.090 0.051 0.046 0.079 0.083 0.105
(0.213) (0.222) (0.223) (0.230) (0.230) (0.230) (0.221) (0.231) (0.219) (0.236)
Business and economics 0.408* 0.367 0.381* 0.359 0.380* 0.359 0.373 0.372* 0.380* 0.373*
(0.218) (0.220) (0.217) (0.218) (0.212) (0.218) (0.223) (0.213) (0.218) (0.213)
Natural sciences 0.153 0.140 0.156 0.129 0.151 0.140 0.136 0.157 0.153 0.149
(0.219) (0.217) (0.215) (0.216) (0.212) (0.218) (0.218) (0.215) (0.216) (0.216)
East Europe − 0.16 − 0.279 − 0.251 − 0.286 − 0.21 − 0.286 − 0.27 − 0.251 − 0.243 − 0.186
(0.385) (0.340) (0.350) (0.341) (0.330) (0.338) (0.339) (0.354) (0.350) (0.314)
Central Europe 0.136 0.124 0.142 0.133 0.177 0.114 0.117 0.123 0.133 0.166
(0.336) (0.316) (0.321) (0.320) (0.310) (0.315) (0.314) (0.326) (0.320) (0.311)
Entrepreneurial reputation 0.534* 0.467 0.502* 0.48 0.530* 0.457 0.373 0.491* 0.456 0.513
(0.293) (0.288) (0.273) (0.297) (0.279) (0.293) (0.310) (0.272) (0.302) (0.307)
Climate 0.030 0.025 0.021 0.025 0.021 0.025 0.027 0.022 0.021 0.020
(0.053) (0.053) (0.053) (0.053) (0.054) (0.053) (0.052) (0.052) (0.054) (0.054)
EE 0.203 0.209 0.207 0.220
(0.148) (0.136) (0.153) (0.147)
Elective EE 0.311* 0.346** 0.317* 0.325** 0.436***
(0.165) (0.156) (0.161) (0.156) (0.160)
Compulsory EE 0.124 0.204 0.09 0.128 0.194
(0.169) (0.156) (0.167) (0.169) (0.158)
D. Hahn et al.
Table 4 (continued)

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10

Self-employed parents − 0.087 − 0.011 − 0.042


(0.115) (0.120) (0.132)
EE × self-employed parents − 0.0176
(0.295)
Elective EE × self-employed parents − 0.154 − 0.361
(0.367) (0.487)
Compulsory EE × self-employed parents − 0.4 − 0.684*
(0.371) (0.376)
PPE 0.0193 0.00546 0.00381
(0.052) (0.050) (0.052)
EE × PPE 0.00796
(0.127)
Experience 0.0141 0.0343 0.0417
(0.027) (0.033) (0.038)
EE × experience 0.233
(0.139)
Elective EE × PPE 0.034 − 0.0515
(0.175) (0.137)
Compulsory EE × PPE 0.240* 0.452**
(0.127) (0.184)
Elective EE × experience 0.101 0.515
(0.214) (0.472)
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family...

Compulsory EE × experience 0.00489 − 0.0757


(0.072) (0.101)
Constant − 1.268 − 1.237 − 1.41 0.38 0.298 0.373 0.409 0.323 0.336 0.215
(0.952) (0.933) (0.961) (0.731) (0.735) (0.704) (0.722) (0.729) (0.735) (0.743)
Observations 427 427 427 427 427 427 427 427 427 427
R-squared 0.523 0.526 0.527 0.527 0.529 0.526 0.529 0.529 0.528 0.535

Robust standard errors in parentheses. University fixed effects are included in each regression
***p < 0.01, **p < 0.05, *p < 0.1
271
272 D. Hahn et al.

Fig. 2 Effect of EE in 5
compulsory courses on
entrepreneurial skills16 at lower
and higher levels of PPE

Entrepreneurial Skills16
4

3
No EE Compulsory EE

Low PPE High PPE

opportunities in high-tech sectors affect the attrac- 1.242, p < 0.01). The coefficient remains positive
tiveness of EE because they raise opportunity costs and statistically significant even if a more prudent
for undertaking innovative projects as self-employed and strict definition is adopted of the nascent entre-
workers (Kacperczyk and Marx 2016), they are like- preneur status, that is, if only concrete activities that
ly to be uncorrelated with the acquisition of entre- are close to venture creation are considered (i.e.,
preneurial skills, which requires some form of first- BPurchased material, equipment, or machinery for
hand exposure to entrepreneurship (Cope 2005; Hahn the business^; BAttempted to obtain external
et al. 2017), such as direct experience of entrepre- funding^; BApplied for a patent, copyright, or
neurship or a close relationship with entrepreneurs in trademark^; BRegistered the business^; BSold prod-
the circle of family and friends (Peterman and uct or service^), rather than activities that may only
Kennedy 2003). By using this instrument, the esti- be associated with tasks typically required of stu-
mated values of the participation to EE courses were dents during EE courses (e.g., write a business
computed, and those values were used to estimate a plan). These results further support the importance
regression model for the entrepreneurial skills vari- for EE to increase students’ self-assessed entrepre-
able, as displayed in Table 5.7 Results show that the neurial skills, if its aim is to encourage entrepreneur-
overall effect of EE is positive and significant, there- ial activity at university.
by supporting a causal identification of the EE im- Third, to check to what extent the results are
pact on entrepreneurial skills. contingent on the European context, the baseline
Second, as an additional robustness check, anal- model is run again on the whole sample of 646
ysis shows that self-assessed entrepreneurial skills students,8 which also comprehends 18 respondents
represent a perception that has concrete impact on from Russia and 201 from Brazil. Results are report-
students’ behaviors (Fayolle and Gailly 2015). Ta- ed in Table 7. The notable differences with respect
ble 6 displays odd ratios of the logit regression to the baseline results arising from the inclusion of
model with nascent entrepreneur status in 2016 as non-European countries are twofold. First, the effect
outcome variable. The coefficient for entrepreneur- of EE in compulsory courses yields significant and
ial skills16 is positive and strongly significant (β = positive results; second, experience in parents’

7
In the first stage, using a probit estimator, the EE indicators on the
8
instrument are regressed to obtain predicted values that contain the As described in Section 3.1, the complete longitudinal GUESSS
variation in the EE indicators uncorrelated with the error terms. Then, 2013–2016 dataset records the answers from 1383 students from 21
in the second stage, the dependent variable, entrepreneurial skills, is countries, but the present analysis includes only questionnaires of those
regressed on predicted values defined in the first stage, applying a two- students who in the 2013 survey reported not having attended EE
stage least squares estimator. Concerning the instrument relevance, before. The sample also excludes students who attended EE in both
first-stage estimates suggest that the instrument and the EE indicators elective and compulsory EE, students from countries whose number of
are negatively correlated and that the degree of correlation rises as complete responses was below 5, and respondents for whom it was not
university fixed effects are introduced in the first-stage regression. possible to build the variables of interest.
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 273

Table 5 2SLS estimation with entrepreneurial skills16 as depen- Table 6 Logistic regression with nascent entrepreneurship in
dent variable 2016 as dependent variable

Entrepreneurial skills13 0.339*** Entrepreneurial skills16 1.202***


(0.123) (0.394)
Nascent entrepreneurial activity13 0.288 Nascent entrepreneurial activity13 0.552
(0.267) (1.013)
Entrepreneurial intention 0.047 Entrepreneurial intention 0.531**
(0.043) (0.249)
Age 0.064*** Age − 0.00375
(0.022) (0.0836)
Male 0.155
Male 0.407
(0.108)
(0.595)
Bachelor − 0.973*
Master 0.390
(0.559)
(0.654)
Master − 1.152*
Business and economics − 0.554
(0.680)
(0.867)
Business and economics − 1.669
Natural sciences − 1.555
(1.105)
(1.015)
Natural sciences − 0.765
(0.489) Entrepreneurial reputation 0.164
Entrepreneurial reputation − 1.386 (1.409)
(1.000) East Europe 2.031**
Climate − 0.203 (0.889)
(0.128) Central Europe 1.187*
East Europe − 2.042** (0.685)
(0.932) EE − 0.0801
Central Europe 0.820 (0.640)
(0.506) Constant − 11.69***
EE 2.197* (2.986)
(1.152)
N = 427; robust standard errors in parentheses
Self-employed parents − 0.694**
***p < 0.01, **p < 0.05, *p < 0.1
(0.316)
PPE 0.198*
(0.108)
Experience 0.248** future research, but extant literature provides some
(0.112) arguments that explain why different outcomes of
Constant 4.682*
EE can be observed in countries with different
(2.826)
socio-economic contexts (Walter and Block 2016).
Observations 427
In developing countries, such as Brazil, the entre-
R-squared 0.53
preneurial career is often seen as a way to overcome
All models are estimated with a two-stage least squares approach unemployment; accordingly, EE will focus on teach-
where the exclusion restriction in the first stage is the variation in ing students the basic skills to become self-
the employment rate in high-tech and knowledge-intensive sectors
employed (Dodd and Hynes 2012). Therefore, EE
at the NUT2 level between 2008 and 2013. Robust standard errors
in parentheses. Estimation with university fixed effects in compulsory courses is more likely be effective
***p < 0.01, **p < 0.05, *p < 0.1 because it raises the interest of larger portions of
students looking at self-employment as a viable job
where other working opportunities are scarce. On
family business negatively moderates the effect of the other hand, in developing areas, EE tends to be
EE in elective courses. Such differences warrant less concerned with imparting the more advanced
Table 7 Linear regression with entrepreneurial skills16 as dependent variable for the whole sample
274

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10

Entrepreneurial skills13 0.507*** 0.501*** 0.502*** 0.501*** 0.503*** 0.500*** 0.501*** 0.503*** 0.502*** 0.506***
(0.050) (0.050) (0.050) (0.050) (0.051) (0.050) (0.050) (0.050) (0.050) (0.051)
Nascent entrepreneurial activity 0.049 0.039 0.056 0.028 0.038 0.037 0.047 0.079 0.051 0.077
(0.186) (0.186) (0.184) (0.188) (0.183) (0.188) (0.187) (0.185) (0.185) (0.185)
Entrepreneurial intention 0.102*** 0.103*** 0.103*** 0.105*** 0.107*** 0.103*** 0.103*** 0.0995*** 0.105*** 0.103***
(0.035) (0.035) (0.034) (0.035) (0.035) (0.035) (0.035) (0.035) (0.035) (0.035)
Age 0.018 0.018 0.019 0.018 0.019 0.018 0.019 0.019 0.019 0.021
(0.016) (0.016) (0.016) (0.016) (0.016) (0.016) (0.016) (0.016) (0.016) (0.016)
Male 0.203** 0.201** 0.197** 0.198** 0.193** 0.202** 0.200** 0.192** 0.189** 0.174*
(0.090) (0.090) (0.090) (0.090) (0.090) (0.090) (0.090) (0.090) (0.090) (0.091)
Bachelor − 0.160 − 0.168 − 0.164 − 0.165 − 0.160 − 0.182 − 0.167 − 0.169 − 0.165 − 0.158
(0.189) (0.189) (0.191) (0.190) (0.192) (0.190) (0.189) (0.192) (0.191) (0.194)
Master − 0.058 − 0.072 − 0.066 − 0.064 − 0.058 − 0.085 − 0.072 − 0.073 − 0.065 − 0.057
(0.184) (0.184) (0.186) (0.185) (0.187) (0.185) (0.184) (0.187) (0.186) (0.190)
Business and economics 0.229 0.189 0.183 0.186 0.185 0.191 0.189 0.191 0.192 0.211
(0.142) (0.144) (0.144) (0.144) (0.144) (0.144) (0.145) (0.145) (0.145) (0.147)
Natural sciences − 0.052 − 0.066 − 0.059 − 0.070 − 0.055 − 0.061 − 0.064 − 0.051 − 0.055 − 0.037
(0.128) (0.129) (0.129) (0.129) (0.130) (0.128) (0.129) (0.130) (0.130) (0.131)
Entrepreneurial reputation 0.488** 0.446** 0.463** 0.451** 0.479** 0.448** 0.435** 0.443** 0.433** 0.433**
(0.196) (0.195) (0.190) (0.195) (0.189) (0.197) (0.199) (0.194) (0.195) (0.196)
Climate 0.036 0.034 0.031 0.033 0.031 0.034 0.034 0.030 0.032 0.032
(0.035) (0.035) (0.035) (0.036) (0.036) (0.036) (0.036) (0.036) (0.036) (0.037)
East Europe 0.194 0.134 0.093 0.117 0.053 0.122 0.146 0.117 0.112 0.142
(0.390) (0.394) (0.384) (0.398) (0.383) (0.398) (0.400) (0.389) (0.391) (0.394)
Central Europe 0.439 0.454 0.429 0.446 0.393 0.450 0.462 0.445 0.438 0.439
(0.373) (0.374) (0.368) (0.377) (0.366) (0.374) (0.377) (0.370) (0.373) (0.374)
South Europe 0.122 0.134 0.098 0.121 0.046 0.134 0.142 0.115 0.106 0.073
(− 0.401) − 0.399 − 0.393 − 0.401 − 0.390 − 0.398 − 0.403 − 0.395 − 0.398 (0.396)
Russia 0.430 0.341 0.311 0.319 0.244 0.340 0.355 0.349 0.313 0.279
(0.433) (0.438) (0.426) (0.443) (0.432) (0.438) (0.443) (0.428) (0.432) (0.439)
EE 0.158* 0.174 0.160* 0.159*
(0.095) (0.106) (0.095) (0.095)
Elective EE 0.276** 0.332** 0.273** 0.274** 0.298**
(0.126) (0.143) (0.123) (0.126) (0.146)
Compulsory EE (0.155) 0.239** 0.127 0.157 0.237**
D. Hahn et al.
Table 7 (continued)

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model 10

(0.110) (0.115) (0.113) (0.110) (0.117)


Self-employed parents − 0.031 0.036 0.025
(0.124) (0.123) (0.132)
EE × self-employed parents − 0.049
(0.221)
Elective EE × self-employed parents − 0.190 − 0.069
(0.299) (0.314)
Compulsory EE × self-employed parents − 0.372 − 0.691**
(0.265) (0.286)
PPE 0.026 0.012 0.014
(0.045) (0.045) (0.045)
EE × PPE − 0.041
(0.089)
Experience 0.007 0.023 0.023
(0.035) (0.036) (0.037)
EE × experience 0.034
(0.115)
Elective EE × PPE − 0.035 − 0.028
(0.109) (0.106)
Compulsory EE × PPE 0.094 0.275**
(0.116) (0.130)
Elective EE × experience − 0.181*** − 0.170***
(0.067) (0.086)
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family...

Compulsory EE × experience 0.043 0.041


(0.084) (0.095)
Constant − 0.704 − 0.719 − 0.799 0.785 0.722 0.790 0.772 0.758 0.751 0.642
(0.633) (0.631) (0.629) (0.645) (0.639) (0.633) (0.635) (0.636) (0.634) (0.645)
Observations 646 646 646 646 646 646 646 646 646 646
R-squared 0.500 0.502 0.504 0.503 0.506 0.503 0.502 0.505 0.506 0.511

Robust standard errors in parentheses. University fixed effects are included in each regression
***p < 0.01, **p < 0.05, *p < 0.1
275
276 D. Hahn et al.

skills required of growth- and innovation-oriented research. More specifically, this study also shows that
entrepreneurs (Dodd and Hynes 2012). For this rea- the association between EE and entrepreneurial skills
son, EE in elective courses might be ill-prepared to can be better understood by unpacking the audience of
satisfy the demand of students who have already EE into participants to elective and compulsory entre-
acquired experience in the family business and want preneurship courses. In particular, it finds that the pos-
to advance their entrepreneurial skills further. itive impact of EE exercised on students in elective
Fourth, to test for the presence of non-linear effect, a courses is rather general and transcends university ef-
Ramsey RESET test was implemented. This test did not fects. In contrast, in compulsory courses, EE augments
permit rejection of the null hypothesis that the model has the entrepreneurial skills only of those students who
no omitted variables (F = 0.38, p > 0.764), thereby perceive higher levels of performance in their parents’
confirming that it was not necessary to introduce qua- entrepreneurial activities. By underscoring the role
dratic terms of moderators. played by students’ PPE, this study also disentangles
Finally, measures have been taken to mitigate the the mechanisms through which the exposure to an en-
concern that the results are biased by the heterogeneity terprising family moderates the EE-entrepreneurial
of time passing between the end of a course (between skills relationship.
2013 and 2016) and students’ responding to the survey
to assess their skills (in 2016): since students took EE 5.1 Contribution to theory and research
during 2013 to 2016 but the dependent variable entre-
preneurial skills16 was assessed in 2016, the students in By conceptually and empirically integrating course type
the sample have different time horizons after the EE and students’ exposure to an enterprising family in the
course in which their skills are measured. It was as- assessment of EE impact, the research sheds light on the
sumed that, holding field and level of study constant pivotal role played by the audience of EE as presenting
(which are controlled for), EE is offered at approximate- key boundary conditions to the effectiveness of educa-
ly the same point of the course of studies across cam- tional interventions aimed at increasing university stu-
puses. Therefore, by controlling for students’ closeness dents’ entrepreneurial skills. In fact, while previous
to the end of their studies, it should be possible to take research proposed these as two relevant dimensions that
into account students’ heterogeneous time horizons after characterize participants to EE and thus affect the EE-
taking EE. In particular, a dummy variable called medi- entrepreneurial skills relationship (e.g., Martin et al.
an was added to the main specification, taking the value 2013; Nabi et al. 2017; Naia et al. 2014), their actual
of 1 for students who answered the 2016 survey that will and joint influence on the entrepreneurial learning pro-
finish their studies later than the median of the sample cess and outcomes remained both theoretically and em-
(i.e., 2017). Results remain substantially unchanged. pirically unveiled. Bearing this in mind, we believe our
work offers several contributions to the current dialog in
EE.
5 Discussion First, the work takes advantage of a quasi-
experimental research design to move towards a recon-
A vibrant research stream in entrepreneurship literature ciliation of the contradictory findings of the vast litera-
has been focusing on education as a means to promote ture on EE impact (Martin et al. 2013; Rideout and Gray
mind-sets, attitudes, and behaviors that increase the rate 2013). The data is longitudinal, the study restricts the
of enterprising individuals, especially across university sample to more homogenous countries, and several
campuses (Martin et al. 2013; Nabi et al. 2017; Rideout factors (e.g., education, intention, skills prior to first
and Gray 2013). The overarching question of this liter- survey) are controlled for. Taken together, these aspects
ature is to what extent and under what circumstances EE permit better exploration of the boundary conditions
has an impact, with a dominant focus on entrepreneurial under which EE is more effective.
learning, including the development of new entrepre- Second, the work shows that the effects of EE in
neurial knowledge and skills (Hahn et al. 2017; Neck elective and compulsory courses are different. By doing
and Greene 2011). In this respect, this paper explores so, it illustrates that disentangling EE into course type is
some boundary conditions of the education-learning important in assessing its outcomes and in understand-
relationship that have been largely overlooked by extant ing the contrasting results produced by extant EE impact
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 277

studies. In particular, the results endorse the suggestion discriminates the participants to EE in compulsory
of being cautious in generalizing the positive effect of courses. At lower levels of PPE, on the other hand,
EE (Nabi et al. 2017; Rauch and Hulsink 2015; Rideout students have more likely been exposed to extrinsic
and Gray 2013). Indeed, while EE in elective courses is and intrinsic downsides of entrepreneurship, such as
beneficial over university effects, EE in compulsory higher level of stress (Uy et al. 2013), financial concerns
courses is beneficial over university effects only for (Binder and Coad 2013), and family conflicts (Shepherd
certain types of students. One possible explanation is and Haynie 2009). All this will diminish the motivation
that students taking EE on a voluntary basis are more to learn. Especially in the context of compulsory
motivated to learn from the courses, while audience of courses, without such motivation, students will not dis-
compulsory EE is composed of a mix of students inter- play significant benefit from EE.
ested and not interested in entrepreneurship (Karimi Finally, our work offers a contribution to the growing
et al. 2016). Additionally, EE in compulsory courses is literature of student entrepreneurship (Kacperczyk
usually aimed at imparting students the basic knowledge 2013; Minola et al. 2016b). While the results confirm
about entrepreneurship (von Graevenitz et al. 2010), the importance of the enterprising family as a learning
whereas elective courses tend to be more specifically environment (Hamilton 2011), the paper also advances
designed for teaching students how to start a business, the integration of university and family embeddedness
combining knowledge from the classroom with hands- as relevant to the students’ learning process. University
on business experience (Rauch and Hulsink 2015). As students are uniquely embedded in both contexts that,
such, students participating to EE in elective courses when considered together, provide a unique contribution
have unique opportunities to acquire knowledge and to explain their entrepreneurial learning.
experience that they could not acquire elsewhere at their
university. In this sense, EE in elective courses does 5.2 Contributions to practice
indeed matter for students’ entrepreneurial learning at
university. At the same time, by showing under which We believe educators and practitioners could derive
circumstances EE in compulsory courses can be actually some valuable suggestions from the research. With EE
effective, this paper offers a more nuanced appraisal of becoming widespread across campuses, there is a
their outcomes. growing and recognized need of more nuanced
In fact, as a third contribution, by taking a family approaches to EE within universities. Building on the
embeddedness perspective on entrepreneurship (Aldrich results, teachers and instructors, as well as course
and Cliff 2003; Sieger and Minola 2017), the modera- designers, need to be aware of different impacts and
tion results offer preliminary insights that could contrib- underlying pedagogical mechanisms that result from
ute to the literature. This work digs deeper into the compulsory versus elective courses. Moreover, as
nuances of family embeddedness, offering evidence on recently reinforced by Fayolle (2013) and Nabi et al.
the mechanisms through which such exposure influ- (2017), there is no one EE initiative like another, since
ences EE outcomes. In particular, this paper comple- the audience composition and students’ individual pro-
ments previous research (e.g., Peterman and Kennedy files contribute to determine the learning environment
2003) showing that the benefits students reap from EE and its effectiveness. Accounting for the audience (e.g.,
do not depend on exposure to self-employed parents as by looking at prior entrepreneurship exposure, such as
such but specifically on the nature of such exposure. In that obtained in family business) becomes a stringent
fact, the results indicate that in compulsory courses, only requirement of professionally developed courses. This
students with better perceptions of their parents’ entre- becomes even more important in different types of EE
preneurial activity learn from EE. On one hand, PPE initiatives (such as compulsory rather than elective
increases the extent to which students are motivated to courses).
learn entrepreneurial skills. For example, PPE is associ- Family business practitioners can also take advantage
ated with positive role models (Criaco et al. 2017), of the results. Ensuring that entrepreneurial skills are
which raise students’ interest towards entrepreneurship. acquired by the next generation is central to the pros-
This can be particularly beneficial to students in com- perity and renewal of family businesses to facilitate
pulsory EE, who are, on average, less attracted to entre- intergenerational succession, to spur corporate entrepre-
preneurship; as such, PPE powerfully characterizes and neurship, and to ensure the transmission of
278 D. Hahn et al.

transgenerational entrepreneurship (Jaskiewicz et al. courses (Hahn et al. 2017). On the other hand, students
2015). While the family business represents a valuable who already have acquired some background knowl-
learning environment as such (Hamilton 2011), our edge of entrepreneurship from their prior experience
findings suggest that it can also prepare young family might not receive additional benefits from the programs
members to take better advantage of and better integrate (Fayolle and Gailly 2015). To shed light on this contro-
learning opportunities outside the family, in particular versy, we recommend future research that more closely
formal education ones. In doing so, these students can looks at the nature of the course, to offer a better under-
represent a valuable resource to the entrepreneurial ac- standing of pedagogies and course design that would
tivities developed by their family. For example, they can allow even students with prior experience to further
act as knowledge brokers and bring into the family increase their entrepreneurial skills. For example, it is
business the more recent developments of entrepreneur- interesting that when students in emerging non-
ial practices learned at university, thereby sustaining European countries are included in the sample, analysis
renewal and entrepreneurial development. shows that experience reduces the benefits of EE in
Finally, policy makers have been deeply involved in elective courses. Since the way entrepreneurship is
funding and fostering EE within and outside universi- taught is known to vary around the world (Dodd and
ties. As the two boundary conditions proffered here add Hynes 2012), our evidence could be explained by the
a nuanced understanding of the outcomes of EE, policy fact that in these countries, EE is less adequate to meet
makers should encourage students’ interest towards EE the expectations of students with prior experience who
initiatives and, especially in Europe, take advantage of take EE on a voluntary basis. Unfortunately, in compar-
the prevalence of family businesses in the region when ing the effect of EE in compulsory versus elective
promoting university-based EE. courses, it was not possible to control for the goals,
content, and specific pedagogy of EE (and how these
5.3 Limitations and future research directions differ between compulsory and elective courses in
particular; Fayolle 2013). This would help in validating
Despite the contributions offered by this study, it is nec- the conceptualization and shedding light on the very
essary to note a few limitations. First, the sample presents specific mechanisms that explain how EE in different
some shortcomings. It is not representative of the whole types of courses actually works.
student population of the selected countries; some of the Third, since in quasi-experimental design partici-
students participating to the 2013 wave of the survey were pants select themselves to groups of those not taking
not reached by the second wave due to different reasons EE and taking EE in elective versus compulsory
(e.g., they left the university, their class was not assigned courses, caution is necessary in inferring causality
the survey in 2016); also, the sample could not be explic- from the results; this is only partially mitigated by
itly tested for non-response bias by identifying early and introducing a rich set of control variables (Rauch
late respondents. Even though the use of GUESSS data and Hulsink 2015) and by addressing self-selection
presents some limitations, its benefits (e.g., the uniqueness into EE through an instrumental variable approach.
of the sample and research design in terms of geographical We encourage future research to consider these as-
scope, number of responses, and longitudinal nature), pects and to integrate methodological rigor with
taken together, have been largely recognized to more than conceptual richness. Research in this direction, for
outweigh its disadvantages in recent research on example, could combine experimental research de-
university-based entrepreneurship (cf. Criaco et al. 2017; sign with the EE teaching model frameworks devel-
Laskovaia et al. 2017; Morris et al. 2017). oped in literature (cf. Béchard and Grégoire 2005;
Second, even though this study explored the effect of Fayolle 2013). Taking advantage of true experiments,
having professional experience in the family business in which students are randomly assigned to different
on the EE-entrepreneurial skills relationship, we obtain- types of entrepreneurship courses with clearly de-
ed non-significant results; this resonates with the am- fined and distinct teaching objectives and peda-
biguous impact of EE programs on participants with gogies, would allow researchers to formulate recom-
different levels of prior capabilities (Lyons and Zhang mendations on which type of EE is most suitable to
2018). On the one hand, experience leads to greater achieve the intended effects (Naia et al. 2014; Rauch
ability to process information and better learn from the and Hulsink 2015; Rideout and Gray 2013).
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 279

Fourth, this study measured EE through binary var- commercialization that are central to economic growth
iables; when combined with the Bboundary conditions^ and innovation. To facilitate the proliferation of these
approach, more nuanced measures of EE (e.g., continu- individuals, universities offer EE aimed at equipping
ous measures of the number of EE courses taken by the students with entrepreneurial skills and preparing them
student or the intensity of exposure to EE measured by to engage in entrepreneurial activities. However, literature
course length) would have the potential to better illus- that assesses the impact of EE on university students
trate how human capital accumulates and becomes use- presents contradicting results, which casts some doubts
ful in entrepreneurship (Hahn et al. 2017). on the effectiveness of these courses and leads scholars to
Fifth, due to the survey-based research design, the explore the boundary conditions under which EE pro-
present work could not identify how much time passed duces the intended effects. This study contributes to this
between the completion of a course and students’ ex- debate by focusing on the type of EE and on students’
post assessment of their skills, which occurred by the exposure to an enterprising family as key contingencies of
time of the survey. Future research that allows the ob- the EE-entrepreneurial skills relationship. It conceptual-
servation of students who attended EE just after the end izes how these factors jointly describe the profile of
of the course and for a larger timespan would be wel- participants to entrepreneurship courses, thereby influenc-
come to draw more accurate conclusions about the ing the heterogeneous outcomes observed by EE impact
effects of EE and their persistence over time. studies. Drawing on a sample of 427 university students
Finally, future EE impact studies could also focus on who participated to two consecutive waves of the Global
other outcomes of EE (such as attitude, career intention, University Entrepreneurial Spirit Students’ Survey
or job market performance over the long run) resulting (GUESSS), the work relies on a quasi-experimental re-
from family-university integrated learning or consider search design to identify the impact of EE in different
the diversity that might result from the students’ level types of courses (elective vs. compulsory) on students’
(bachelor, master, PhD) or field of study. For example, it entrepreneurial skills, contingent to participants’ exposure
would be particularly interesting to explore whether to entrepreneurship in their families. It finds that partici-
self-assessed skills actually correspond to real learning pants to the two types of EE benefit in terms of entrepre-
and allow students to perform entrepreneurial tasks neurial skills to different extents. In particular, the impact
more effectively. Scholars in EE could also take advan- of compulsory courses is positive only for students with
tage of emerging issues in the literature of family busi- self-employed parents who perceive their parents to per-
ness and enterprising families (Jaskiewicz et al. 2016; form well as entrepreneurs; conversely, the positive im-
Jaskiewicz et al. 2015), by looking, for example, at how pact of elective courses holds even when university ef-
the student offspring’s gender or the family composition fects are controlled for. The findings show that both
additionally influence the normative consequences of university and family, two contexts in which students
family embeddedness. In a similar vein, formal and are socially embedded, jointly influence the acquisition
informal institutions in the region or the country where of entrepreneurial skills among young individuals.
students take EE are likely to influence the knowledge
and social norms that shape the dynamics of EE classes Acknowledgments We would like to thank the organizers and
(Dodd and Hynes 2012; Walter and Dohse 2012). Fol- participants of the Workshop Knowledge Frontiers and Knowl-
edge Boundaries in Europe (Bozen, October 2017) where we
lowing this, adding cultural or geographical contextual received insightful comments that helped us to refine this paper.
dimensions to the university and family embeddedness We are grateful also for the suggestions offered at the 28th
of students (both separately and jointly) represents a Riunione Scientifica Annuale Associazione italiana di Ingegneria
promising direction for future research (Leitch et al. Gestionale (Bari, October 2017), the Technology Transfer Society
Annual Conference (Washington, November 2017), and at the
2012; Nabi et al. 2017). International Research Conference on Science and Technology
Entrepreneurship Education (Toulouse, April 2017) where earlier
versions of the manuscript have been presented.
6 Conclusion
Funding information Support for this research was provided
by the BCampus Entrepreneurship^ project, financed by the Uni-
Individuals who are endowed with the entrepreneurial
versity of Bergamo through the BExcellence Initiative^ funding
skills required to identify and act on new business oppor- scheme, and by the Italian Ministry of Education and Research
tunities drive the processes of knowledge spillover and through the BContamination Lab^ funding scheme.
280 D. Hahn et al.

References managers? J Bus Ventur, 13(4), 295–316. https://doi.


org/10.1016/S0883-9026(97)00029-3.
Chlosta, S., Patzelt, H., Klein, S. B., & Dormann, C. (2012).
Aiken, L. S., & West, S. G. (1991). Multiple regression: testing Parental role models and the decision to become self-
and interpreting interactions. Newbury Park: Sage. employed: the moderating effect of personality. Small Bus
Aldrich, H. E., & Cliff, J. E. (2003). The pervasive effects of Econ, 38(1), 121–138. https://doi.org/10.1007/s11187-010-
family on entrepreneurship: toward a family embeddedness 9270-y.
perspective. J Bus Ventur, 18(5), 573–596. https://doi. Cohen, J., & Cohen, P. (1983). Applied multiple regression/
org/10.1016/S0883-9026(03)00011-9. correlation analysis for the behavioral sciences (2nd ed.).
Audretsch, D. B. (2014). From the entrepreneurial university to the Hillsdale: Erlbaum.
university for the entrepreneurial society. J Technol Transf, Cope, J. (2005). Toward a dynamic learning perspective of entre-
39(3), 313–321. https://doi.org/10.1007/s10961-012-9288-1. preneurship. Entrep Theory Pract, 29(4), 373–397.
Audretsch, D. B., & Belitski, M. (2013). The missing pillar: the https://doi.org/10.1111/j.1540-6520.2005.00090.x.
creativity theory of knowledge spillover entrepreneurship. Criaco, G., Sieger, P., Wennberg, K., Chirico, F., & Minola, T.
Small Bus Econ, 41(4), 819–836. https://doi.org/10.1007 (2017). Parents’ performance in entrepreneurship as a
/s11187-013-9508-6. Bdouble-edged sword^ for the intergenerational transmission
Audretsch, D. B., & Lehmann, E. E. (2005). Do university policies of entrepreneurship. Small Bus Econ, 49(4), 841–864.
make a difference? Res Policy, 34(3), 343–347. https://doi. https://doi.org/10.1007/s11187-017-9854-x.
org/10.1016/j.respol.2005.01.006. Cumby, R. E., & Huizinga, J. (1992). Investigating the correlation
Bae, T. J., Qian, S., Miao, C., & Fiet, J. (2014). The relationship of unobserved expectations: expected returns in equity and
between entrepreneurship education and entrepreneurial in- foreign exchange markets and other examples. J Monet Econ,
tentions: a meta-analytic review. Entrep Theory Pract, 38(2), 30(2), 217–253. https://doi.org/10.1016/0304-3932(92
217–254. https://doi.org/10.1111/etap.12095. )90061-6.
Bascle, G. (2008). Controlling for endogeneity with instrumental Dodd, S. D., & Hynes, B. C. (2012). The impact of regional
variables in strategic management research. Strateg Organ, entrepreneurial contexts upon enterprise education. Entrepr
6(3), 285–327. https://doi.org/10.1177/1476127008094339. Reg Dev, 24(9–10), 741–766. https://doi.org/10.1080
/08985626.2011.566376.
Béchard, J. P., & Grégoire, D. (2005). Entrepreneurship education
Eddleston, K. A., Kellermanns, F. W., & Sarathy, R. (2008).
research revisited: the case of higher education. Acad Manag
Resource configuration in family firms: linking resources,
Learn Educ, 4(1), 22–43. https://doi.org/10.5465
strategic planning and technological opportunities to perfor-
/amle.2005.16132536.
mance. J Manag Stud, 45(1), 26–50. https://doi.org/10.1111
Bergmann, H., Hundt, C., & Sternberg, R. (2016). What makes
/j.1467-6486.2007.00717.x.
student entrepreneurs? On the relevance (and irrelevance) of
Fayolle, A. (2013). Personal views on the future of entrepreneur-
the university and the regional context for student start-ups.
ship education. Entrep Reg Dev, 25(7–8), 692–701.
Small Bus Econ, 47(1), 53–76. https://doi.org/10.1007
https://doi.org/10.1080/08985626.2013.821318.
/s11187-016-9700-6.
Fayolle, A., & Gailly, B. (2015). The impact of entrepreneurship
Bergmann, H., Geissler, M., Hundt, C., & Grave, B. (2018). The education on entrepreneurial attitudes and intention: hystere-
climate for entrepreneurship at higher education institutions. sis and persistence. J Small Bus Manag, 53(1), 75–93.
Res Policy, 47(4), 700–716. https://doi.org/10.1016/j. https://doi.org/10.1111/jsbm.12065.
respol.2018.01.018. Grilo, I., & Thurik, R. (2008). Determinants of entrepreneurial
Binder, M., & Coad, A. (2013). Life satisfaction and self-employ- engagement levels in Europe and the US. Ind Corp Chang,
ment: a matching approach. Small Bus Econ, 40(4), 1009– 17(6), 1113–1145. https://doi.org/10.1093/icc/dtn044.
1033. https://doi.org/10.1007/s11187-011-9413-9. Haase, H., & Lautenschläger, A. (2011). The ‘teachability dilem-
Bird, B., Schjoedt, L., & Baum, J. R. (2012). Editor’s introduction. ma’ of entrepreneurship. Int Entrep Manag J, 7(2), 145–162.
Entrepreneurs’ behavior: elucidation and measurement. https://doi.org/10.1007/s11365-010-0150-3.
Entrep Theory Pract, 36(5), 889–913. https://doi. Hahn, D., Minola, T., Van Gils, A., & Huybrechts, J. (2017).
org/10.1111/j.1540-6520.2012.00535.x. Entrepreneurial education and learning at universities: ex-
Block, J., Thurik, R., Van der Zwan, P., & Walter, S. (2013). ploring multilevel contingencies. Entrep Reg Dev, 29(9–
Business takeover or new venture? Individual and environ- 10), 945–974. https://doi.org/10.1080
mental determinants from a cross-country study. Entrep /08985626.2017.1376542.
Theory Pract, 37(5), 1099–1121. https://doi.org/10.1111 Hamilton, E. (2011). Entrepreneurial learning in family business: a
/j.1540-6520.2012.00521.x. situated learning perspective. J Small Bus Enterp Dev, 18(1),
Carter, N. M., Gartner, W. B., & Reynolds, P. D. (1996). Exploring 8–26. https://doi.org/10.1108/14626001111106406.
start-up event sequences. J Bus Ventur, 11(3), 151–166. Hoffmann, A., Junge, M., & Malchow-Møller, N. (2015).
https://doi.org/10.1016/0883-9026(95)00129-8. Running in the family: parental role models in entrepreneur-
Chang, J., & Rieple, A. (2013). Assessing students’ entrepreneur- ship. Small Bus Econ, 44(1), 79–104. https://doi.org/10.1007
ial skills development in live projects. J Small Bus Enterp /s11187-014-9586-0.
D e v, 2 0 ( 1 ) , 2 2 5 – 2 4 1 . h t t p s : / / d o i . o r g / 1 0 . 11 0 8 Honig, B. (2004). Entrepreneurship education: toward a model of
/14626001311298501. contingency-based business planning. Acad Manag Learn
Chen, C. C., Greene, P. G., & Crick, A. (1998). Does entrepre- Educ, 3(3), 258–273. https://doi.org/10.5465
neurial self-efficacy distinguish entrepreneurs from /amle.2004.14242112.
The impact of entrepreneurship education on university students’’ entrepreneurial skills: a family... 281

Hytti, U., Stenholm, P., Heinonen, J., & Seikkula-Leino, J. (2010). Liñán, F., & Fernandez-Serrano, J. (2014). National culture, en-
Perceived learning outcomes in entrepreneurship education: trepreneurship and economic development: different patterns
the impact of student motivation and team behaviour. Educ across the European Union. Small Bus Econ, 42(4), 685–701.
Train, 52(8/9), 587–606. https://doi.org/10.1108 https://doi.org/10.1007/s11187-013-9520-x.
/00400911011088935. Lyons, E., & Zhang, L. (2018). Who does (not) benefit from
Jaskiewicz, P., Combs, J. G., & Rau, S. B. (2015). Entrepreneurial entrepreneurship programs? Strateg Manag J, 39(1), 85–
legacy: toward a theory of how some family firms nurture 112. https://doi.org/10.1002/smj.2704.
transgenerational entrepreneurship. J Bus Ventur, 30(1), 29– Martin, B. C., McNally, J. J., & Kay, M. J. (2013). Examining the
49. https://doi.org/10.1016/j.jbusvent.2014.07.001. formation of human capital in entrepreneurship: a meta-
Jaskiewicz, P., Combs, J. G., Ketchen, D. J., & Ireland, R. D. analysis of entrepreneurship education outcomes. J Bus
(2016). Enduring entrepreneurship: antecedents, triggering Ventur, 28(2), 211–224. https://doi.org/10.1016/j.
mechanisms, and outcomes. Strateg Entrep J, 10(4), 337– jbusvent.2012.03.002.
345. https://doi.org/10.1002/sej.1234. McGee, J. E., Peterson, M., Mueller, S. L., & Sequeira, J. M.
Kacperczyk, A. J. (2013). Social influence and entrepreneurship: (2009). Entrepreneurial self-efficacy: refining the measure.
the effect of university peers on entrepreneurial entry. Organ Entrep Theory Pract, 33(4), 965–988. https://doi.org/10.1111
Sci, 24(3), 664–683. https://doi.org/10.1287/orsc.1120.077. /j.1540-6520.2009.00304.x.
Kacperczyk, A., & Marx, M. (2016). Revisiting the small-firm Memili, E., Fang, H., Chrisman, J. J., & De Massis, A. (2015). The
effect on entrepreneurship: evidence from firm dissolutions. impact of small-and medium-sized family firms on economic
Organ Sci, 27(4), 893–910. https://doi.org/10.1287 growth. Small Bus Econ, 45(4), 771–785. https://doi.
/orsc.2016.1065. org/10.1007/s11187-015-9670-0.
Karimi, S., Biemans, H. J., Lans, T., Chizari, M., & Mulder, M. Minola, T., Criaco, G., & Cassia, L. (2014). Are youth really
(2016). The impact of entrepreneurship education: a study of different? New beliefs for old practices in entrepreneurship.
Iranian students’ entrepreneurial intentions and opportunity Int J Entrep Innov Manag, 18(2-3), 233–259. https://doi.
identification. J Small Bus Manag, 54(1), 187–209. org/10.1504/IJEIM.2014.062881.
https://doi.org/10.1111/jsbm.12137. Minola, T., Brumana, M., Campopiano, G., Garrett, R. P., &
Karlsson, T., & Moberg, K. (2013). Improving perceived entre- Cassia, L. (2016a). Corporate venturing in family business:
preneurial abilities through education: exploratory testing of a developmental approach of the enterprising family. Strateg
an entrepreneurial self efficacy scale in a pre-post setting. Int Entrep J, 10(4), 395–412. https://doi.org/10.1002/sej.1236.
J Manag Educ, 11(1), 1–11. https://doi.org/10.1016/j. Minola, T., Donina, D., & Meoli, M. (2016b). Students climbing
ijme.2012.10.001. the entrepreneurial ladder: does university internationaliza-
Kennedy, P. (2008). A guide to econometrics. Malden: Blackwell tion pay off? Small Bus Econ, 47(3), 565–587. https://doi.
Publishing. org/10.1007/s11187-016-9758-1.
Kickul, J., Gundry, L. K., Barbosa, S. D., & Whitcanack, L. Morris, M. H., Shirokova, G., & Tsukanova, T. (2017). Student
(2009). Intuition versus analysis? Testing differential models entrepreneurship and the university ecosystem: a multi-
of cognitive style on entrepreneurial self-efficacy and the new country empirical exploration. Eur J Int Manag, 11(1), 65–
venture creation process. Entrep Theory Pract, 33(2), 439– 85. https://doi.org/10.1504/EJIM.2017.081251.
453. https://doi.org/10.1111/2Fj.1540-6520.2009.00298.x. Mueller, S., & Anderson, A. R. (2014). Understanding the entre-
Kuratko, D. F. (2005). The emergence of entrepreneurship educa- preneurial learning process and its impact on students’ per-
tion: development, trends, and challenges. Entrep Theory sonal development: a European perspective. Int J Manag
Pract, 29(5), 577–598. https://doi.org/10.1111/j.1540- Educ, 12(3), 500–511. https://doi.org/10.1016/j.
6520.2005.00099.x. ijme.2014.05.003.
Laskovaia, A., Shirokova, G., & Morris, M. H. (2017). National Mungai, E., & Velamuri, S. R. (2011). Parental entrepreneurial
culture, effectuation, and new venture performance: global role model influence on male offspring: is it always positive
evidence from student entrepreneurs. Small Bus Econ, 1–23. and when does it occur? Entrep Theory Pract, 35(2), 337–
https://doi.org/10.1007/s11187-017-9852-z. 357. https://doi.org/10.1111/j.1540-6520.2009.00363.x.
Laspita, S., Breugst, N., Heblich, S., & Patzelt, H. (2012). Mustar, P. (2009). Technology management education: innovation
Intergenerational transmission of entrepreneurial intentions. and entrepreneurship at MINES ParisTech, a leading French
J Bus Ventur, 27(4), 414–435. https://doi.org/10.1016/j. engineering school. Acad Manag Learn Educ, 8(3), 418–
jbusvent.2011.11.006. 425. https://doi.org/10.5465/amle.8.3.zqr418.
Laukkanen, M. (2000). Exploring alternative approaches in high- Nabi, G., Liñán, F., Fayolle, A., Krueger, N., & Walmsley, A.
level entrepreneurship education: creating micromechanisms (2017). The impact of entrepreneurship education in higher
for endogenous regional growth. Entrep Reg Dev, 12(1), 25– education: a systematic review and research agenda. Acad
47. https://doi.org/10.1080/089856200283072. Manag Learn Educ, 16(2), 277–299. https://doi.org/10.5465
Leitch, C., Hazlett, S. A., & Pittaway, L. (2012). Entrepreneurship /amle.2015.0026.
education and context. Entrep Reg Dev, 24(9–10), 733–740. Naia, A., Baptista, R., Januário, C., & Trigo, V. (2014). A system-
https://doi.org/10.1080/08985626.2012.733613. atization of the literature on entrepreneurship education:
Liñán, F., & Chen, Y. W. (2009). Development and cross-cultural challenges and emerging solutions in the entrepreneurial
application of a specific instrument to measure entrepreneur- classroom. Ind High Educ, 28(2), 79–96. https://doi.
ial intentions. Entrep Theory Pract, 33(3), 593–617. org/10.5367/ihe.2014.0196.
https://doi.org/10.1111/j.1540-6520.2009.00318.x.
282 D. Hahn et al.

Neck, H. M., & Greene, P. G. (2011). Entrepreneurship education: 1245–1264. https://doi.org/10.1111/j.15 40-
known worlds and new frontiers. J Small Bus Manag, 49(1), 6520.2009.00344.x.
55–70. https://doi.org/10.1111/j.1540-627X.2010.00314.x. Shinnar, R., Pruett, M., & Toney, B. (2009). Entrepreneurship
Nunnally, J. (1978). Psychometric theory. New York: McGraw education: attitudes across campus. J Educ Bus, 84(3), 151–
Hill. 159. https://doi.org/10.3200/JOEB.84.3.151-159.
Oosterbeek, H., Van Praag, M., & Ijsselstein, A. (2010). The Sieger, P., & Minola, T. (2017). The family’s financial support as a
impact of entrepreneurship education on entrepreneurship Bpoisoned gift^: a family embeddedness perspective on en-
skills and motivation. Eur Econ Rev, 54(3), 442–454. trepreneurial intentions. J Small Bus Manag, 55(S1), 179–
https://doi.org/10.1016/j.euroecorev.2009.08.002. 204. https://doi.org/10.1111/jsbm.12273.
Oster, E. (2017). Unobservable selection and coefficient stability: Souitaris, V., Zerbinati, S., & Al-Laham, A. (2007). Do entrepre-
theory and evidence. J Bus Econ Stat, 1–18. https://doi. neurship programmes raise entrepreneurial intention of sci-
org/10.1080/07350015.2016.1227711. ence and engineering students? The effect of learning, inspi-
Peterman, N. E., & Kennedy, J. (2003). Enterprise education: ration and resources. J Bus Ventur, 22(4), 566–591.
influencing students’ perceptions of entrepreneurship. https://doi.org/10.1016/j.jbusvent.2006.05.002.
Entrep Theory Pract, 28(2), 129–144. https://doi. Toutain, O., Fayolle, A., Pittaway, L., & Politis, D. (2017). Role
org/10.1046/j.1540-6520.2003.00035.x. and impact of the environment on entrepreneurial learning.
Politis, D. (2005). The process of entrepreneurial learning: a Entrep Reg Dev, 29(9–10), 869–888. https://doi.org/10.1080
conceptual framework. Entrep Theory Pract, 29(4), 399– /08985626.2017.1376517.
424. https://doi.org/10.1111/j.1540-6520.2005.00091.x. Uy, M. A., Foo, M. D., & Song, Z. (2013). Joint effects of prior
Rabe-Hesketh, S., & Skrondal, A. (2008). Generalized linear start-up experience and coping strategies on entrepreneurs’
mixed effects models. In G. Fitzmaurice, M. Davidian, G. psychological well-being. J Bus Ventur, 28(5), 583–597.
Verbeke, & G. Molenberghs (Eds.), Longitudinal data anal- https://doi.org/10.1016/j.jbusvent.2012.04.003.
ysis: a handbook of modern statistical methods (pp. 79–106). Volery, T., Müller, S., Oser, F., Naepflin, C., & Rey, N. (2013). The
Boca Raton: Chapman and Hall/CRC. impact of entrepreneurship education on human capital at
Rauch, A., & Hulsink, W. (2015). Putting entrepreneurship edu- upper-secondary level. J Small Bus Manag, 51(3), 429–
cation where the intention to act lies: an investigation into the 446. https://doi.org/10.1111/jsbm.12020.
impact of entrepreneurship education on entrepreneurial be- von Graevenitz, G., Harhoff, D., & Weber, R. (2010). The effects
havior. Acad Manag Learn Educ, 14(2), 187–204. https://doi. of entrepreneurship education. J Econ Behav Organ, 76(1),
org/10.5465/amle.2012.0293. 90–112. https://doi.org/10.1016/j.jebo.2010.02.015.
Rideout, E. C., & Gray, D. O. (2013). Does entrepreneurship
Walter, S. G., & Block, J. H. (2016). Outcomes of entrepreneur-
education really work? A review and methodological critique
ship education: an institutional perspective. J Bus Ventur,
of the empirical literature on the effects of university-based
31(2), 216–233. https://doi.org/10.1016/j.
entrepreneurship education. J Small Bus Manag, 51(3), 329–
jbusvent.2015.10.003.
351. https://doi.org/10.1111/jsbm.12021.
Walter, S. G., & Dohse, D. (2012). Why mode and regional
Rogoff, E. G., & Heck, R. K. Z. (2003). Evolving research in
context m atter for entrepreneurship education .
entrepreneurship and family business: recognizing family as
Entrepreneurship and Regional Development, 24(9–10),
the oxygen that feeds the fire of entrepreneurship. J Bus
807–835 https://doi.org/10.1080/08985626.2012.721009.
Ventur, 18(5), 559–566. https://doi.org/10.1016/S0883-9026
(03)00009-0. Wilson, G., & McCrystal, P. (2007). Motivations and career aspi-
Salvato, C., Sharma, P., & Wright, M. (2015). From the guest rations of MSW students in Northern Ireland. Soc Work
editors: learning patterns and approaches to family business Educ, 26(1), 35–52. https://doi.org/10.1080
education around the world—issues, insights, and research /02615470601036534.
agenda. Acad Manag Learn Educ, 14(3), 307–320. Wilson, F., Kickul, J., & Marlino, D. (2007). Gender, entrepre-
Sánchez, J. C. (2011). University training for entrepreneurial com- neurial self-efficacy, and entrepreneurial career intentions:
petencies: its impact on intention of venture creation. Int implications for entrepreneurship education 1. Entrep
Entrep Manag J, 7(2), 239–254. https://doi.org/10.1007 Theory Pract, 31(3), 387–406. https://doi.org/10.1111
/s11365-010-0156-x. /j.1540-6520.2007.00179.x.
Shah, S. K., & Pahnke, E. C. (2014). Parting the ivory curtain: Zhou, J., & George, J. M. (2001). When job dissatisfaction leads to
understanding how universities support a diverse set of creativity: encouraging the expression of voice. Acad Manag
startups. J Technol Transf, 39(5), 780–792. https://doi. J, 44(4), 682–696. https://doi.org/10.5465/3069410.
org/10.1007/s10961-014-9336-0.
Shepherd, D., & Haynie, J. M. (2009). Family business, identity Publisher’s note Springer Nature remains neutral with regard to
conflict, and an expedited entrepreneurial process: a process jurisdictional claims in published maps and institutional
of resolving identity conflict. Entrep Theory Pract, 33(6), affiliations.

You might also like