1
Job Preferences
Entrepreneurship vs. Job Seeking
University of Groningen Faculty of Economics and Business MSc Marketing – Marketing Intelligence
Name: Robert Lelkes
2
Job Preferences
Entrepreneurship vs. Job Seeking
by
Robert Lelkes
Master Thesis
Marketing Department of University of Groningen
Student number: S2625059
E-mail Address: r.lelkes@student.rug.nl / rob.lelkes@gmail.com Supervisors: Dr. Felix Eggers
Dr. Hans Risselada Faculty of Economics and Business
3 Management summary
This research is conducted in order to determine under what conditions would students prefer entrepreneurship rather than self-employment. During our research we examine how specifically do external factors affect this choice due to the relative ease of adjusting external factors relative to the adjustment of internal factors of students. Due to focus of previously conducted research on perception of actual conditions of contextual factors, we believe that incorporating element of dynamics will extend on existing literature.
Sample of 47 students, consisting of students at Dutch and Slovak universities, completed choice-based conjoint analysis and made their choices in twelve choice sets presented to them.
Hypothesis 1, where we assumed barriers to have negative effect on entrepreneurship intention holds with a negative effect, where increasing barriers will result in lower attractiveness of entrepreneurship. Hypothesis 2, support factors having positive effect on entrepreneurship intention holds aswell, resulting in increased attractiveness of entrepreneurship with increasing support factors. Hypothesis 3 is supported, where job offer has a negative effect on entrepreneurship intention, hence in some cases is more attractive for students to become a job seeker rather than self-employed. Hypothesis 4, attitude towards entrepreneurship having moderating effect on the relation between contextual factors and entrepreneurial intent, could not be supported due to nonsignificance of this effect. Same applies to Hypothesis 4a, where risk taking prosperity does not have significant effect on mentioned relation. Hypothesis 4b does partially hold, where is a significant moderating effect of locus of control on relation between contextual factors and entrepreneurial intent, specifically only on required capital by government.
4 Preface
With this Master Thesis I am a step closer to finishing studies of Marketing and a step closer to starting a new chapter of life.
Thanks to this research I was able to gather new insights, knowledge and most importantly conducted research that I believe will serve as a starting point for future projects that will help students. I would never be able to finish this thesis withouth a significant help of several people. Thus, I would like to show my graditute for their support during the writing of this thesis by naming them:
Dr. Felix Eggers, for his supervision, great constructive feedback and willingness during our meetings
Dr. Olga A. Belousova, for constructive feedback and help with Dutch respondents Ing. Katarína Chomová, PhD., for help with Slovak respondents
5 Table of Contents 1. Introduction 8 1.1. Problem Background 8 2. Literature Review 11 2.1. Internal Factors 11 2.2. External Factors 12 2.3. Entrepreneurship Education 14 2.4. Conceptual Model 15 2.5. Hypotheses 16 3. Methodology 19 3.1. Pre-test 19 3.2. Sampling 19 3.3. Independent variables 20 3.4. Dependent variable 21 3.5. Experimental design 21 3.6. Moderating effects 21 3.7. Data collection 22 4. Results 23 4.1. Sample 23 4.2. Factor Analysis 23 4.3. Model Fit 24 4.3.1. Part-worth Model 24
4.3.2. Numeric Attribute Effects 24
4.4. Preference Estimates 26
4.5. Moderating effects 27
4.6. Importance of attributes 29
4.7. Segmentation of the sample 29
4.7.1. A priori segmentation 29
4.7.2. Preference-based segmentation 30
5. Discussion 32
6. Managerial implications and Limitations 32
6.1. Managerial implications 32
6.2. Limitations 33
6
8. Appendix 37
Table of Figures
Figure 1 – Evolution of unemployment rate of young people in European Union 8
Figure 2 - The Conceptual Model 16
Figure 3 – Example choice set 21
Figure 4 – Gender distribution 23
Figure 5 – Student status distribution 23
Figure 6 – Age distribution 23
Figure 7 – Country of university distribution 23
Figure 8 – Utilities of attribute levels 25
Figure 9 – Utilities of attribute levels - continuous 25
Table 1 – Perceived most important external factors 19
Table 2 – Attributes and attribute levels 21
Table 3 – Reliability Statistics 24
Table 4 – Reliability Statistics 24
Table 5 – Reliability Statistics 24
Table 6 – Likelihood Ratio Test 24
Table 7 – Comparison of Models 25
Table 8 – Parameter Estimates 27
Table 9 – Examples of attribute level combinations exceeding none option 27 Table 10 – Moderating effect of Attitude toeards entrepreneurship 28 Table 11 – Moderating effect of Locus of control on Required capital 28
Table 12 – Importance of Attributes 29
Table 13 – Assessment of Model Fit 30
Table 14 – Comparison of Segments 31
Table 15 – Parameter Estimates 31
7
Appendix 2 – List of support factors used for pre-test 37
Appendix 3 – Personality Traits and Attitude Towards Entrepreneurship questions 38
Appendix 4 – Segmentation by Country of university 38
Appendix 5 – Segmentation by Gender 38
8
Figure 1 – Evolution of unemployment rate of young people in European Union
1. Introduction
1.1. Problem Background Graduate students are every year facing an important decision – whether to seek a job or to become an entrepreneur. While this may be an easy choice for some, others may struggle with it. Recent evolution of unemployment rate among young people (<25 years
old) may also have an effect on students’ consideration of self-employment. Using data from Eurostat, Google created an evolution chart of unemployment rate, which is holding on 10%. Katz (1992) defines “employment status choice” as “the vocational decision process in terms of the individual's decision to enter an occupation as a wage-or-salaried individual or a self-employed one.” and looks into this problem from the psychosocial standing point. For universities and governments it might be interesting to see how will students’ entrepreneurship intention change under different contexts to find an ideal setting that might motivate students toward entrepreneurship. Under the conditions of opened borders in European Union universities and governments might think in the global sense and try to make conditions ideal for young talents or even attract them from neighboring countries to become entrepreneurs. New ventures are considered as an opportunity for growth of regions and country.
9
availability of business incubators, while in Iran there might be barriers like external national conflicts (Fereidouni, 2010), therefore we will focus on students within the European Union. In order to create realistic choices and make the results applicable, we will focus on external factors that can be relatively easily adjusted by universities and governments. For example, according to Iakovleva (2014) capital access is perceived as one of the most important barriers in starting new venture by students, in both developed and developing countries. In this case, relatively convenient solution is offered by crowdfunding platforms. In the recent years are start-ups becoming more available options for small teams thanks to crowdfunding of their ideas. For example, crowdfunding platform Kickstarter helped launching almost 80 000 projects, what represents 39% of all proposed projects on Kickstarter (Kickstarter, 2015). Even though crowdfunding platforms make starting a new venture more available, innovations and knowledge thrive mostly in universities. There is a large number of leading-edge companies in Silicon Valley that have close ties to Stanford, which according to Pfeiffer is a result of classes and programs on entrepreneurship offered by Stanford for a long time (Pfeiffer, 1997).
Although there is extended research on entrepreneurial intention among the population and cultures, it may not be fully applied to the student context. Previously conducted research highly focuses on internal factors (Carey, 2010; Ismail, 2009; Hartog, 2010; Sánchez, 2013), where most researchers consider personality traits like risk aversion, motivation, self-efficacy, autonomy, locus of control and need for achievement. Although these can be changed by entrepreneurship education (Sánchez, 2013), due to limited resources of universities it might by more interesting to choose already motivated students and create ideal environment for them. Results indicate that internal factors have the strongest effect on entrepreneurial intention and these factors are hard to change in short-term, therefore it might be interesting to look into personality traits before accepting students into entrepreneurship programs. For the purpose of our study we will account for moderating effect of internal factors on entrepreneurship intention among university students.
10
other, that can be found in Appendix 1. These factors could be possibly used as attributes in choice-based conjoint analysis and are rather easy to comprehend even by students that do not necessarily have the needed knowledge about current entrepreneurship environment and due to hypothetical nature of this experiment is knowledge of actual entrepreneurship environment not needed.
Universities and governments could be able to motivate students toward entrepreneurship through external factors and creation of ideal conditions for them. We would like to research whether a context of university and government support has a substantial effect on entrepreneurship intention among university students and under what hypothetical conditions students would choose entrepreneurship over job seeking. Subsequently, the following research questions formulated:
1. “Do external factors affect students’ choice of entrepreneurship over job seeking?” 2. “Under what condition are students intending to become entrepreneurs?”
3. “Do personal traits affect entrepreneurial intent?”
11
2. Literature Review
2.1. Internal Factors
Sánchez (2013) focuses on internal factors affecting students’ motivation towards entrepreneurship, where all three categories – self-efficacy, proactiveness and risk taking have positive effect on entrepreneurship intention in both cases – before and after the entrepreneurship education.
Carey (2010) looks also into internal motivations, where they divided new ventures into three types and researched to what extend students intend to start one of those ventures. They ask students about “Intention” to start a venture, “Behavioral Belief” that measures attractiveness of creating new venture and “Perceived Behavioral Control”, where they asked whether students believe they would be able to acquire necessary skills for this venture. Results indicate that students’ entrepreneurship intention differs by type of venture and that “normative values do not appear to be a strong determinant of student intentions to start any type of business.” This is also supported by other research, where for example out of the social network positively affects entrepreneurship intention only being a part of business network (Davidsson & Honig, 2003). On the other hand, Ismail (2009) focuses on “Big Five Personality Traits” – neuroticism, extraversion, conscientiousness, openness, agreeableness among students and their effect on entrepreneurship intention, therefore the research shifts more towards psychology. Results indicate that openness, extraversion and extraversion are significantly and positively related to entrepreneurship intention among students.
Research of Lüthje found a reasonable balance between internal and external factors affecting student entrepreneurial intent and conducted a research on engineering students’ perception of entrepreneurship (Lüthje, 2003). Lüthje chose “risk taking propensity” and “internal locus control” as personal traits affecting students’ “attitude towards entrepreneurship”. Lüthje supports this choice by literature review, where are mentioned personality traits frequently named as important personality traits of new venture creators and support aspiration towards self-employment, which was confirmed by the results, where both effects on attitude are significant – risk taking prosperity (=0.464) and internal locus of control (=0.300).
12
on personality traits – specifically risk taking prosperity and locus of control, adopted from the research of Lüthje (2003).
2.2. External Factors
Naffziger (1994) suggests five categories of factors affecting decision to behave entrepreneurially: “an entrepreneur's personal characteristics”, “the individual's personal environment”, “the relevant business environment”, “the specific business idea” and “the goals of the entrepreneur”. Out of those were two external: “the individual's personal environment” and “the relevant business environment”; and these individual factors were used as one of the sources for Lüthje (2003). Explorative interviews with students were conducted by Lüthje (2003) in order to determine external factors perceived as the most important by students out of 44 items based on literature review in order to determine which factors should be used in their research. Result of mentioned interviews were 6 factors divided into two groups – “perceived barriers” and “perceived support factors”. Due to all interviewees being MIT students, these context factors may not be as important for students in other contexts, as indicate results of Iakovleva (2014). Results indicate slight differences among students of different countries regarding importance of factors. Considering large number of possible external factors is important to determine the most important for specific target group.
13
Research by Lüthje (2002) measures students’ perception of supporting factors of university in Germany and compares them to MIT students, where the latter group perceives all factors – “The creative atmosphere in my university inspires to develop ideas for new businesses”, “The courses foster the social and leadership skills needed by entrepreneurs”, “The courses provide students with the knowledge required to start a new company”, “My university supports building multi-disciplinary student teams”, “The university provides a strong network of new venture investors” and “The university actively promotes the process of founding a new company”, more positive than German students but this research does not include information about how important are individual factors to students. In most of the literature are used both internal and external factors, at least the social environment – support by family, friends or role models (Fretschner, 2013; Katz, 1992; Roudaki, 2009) and the importance of these factors is mostly significantly less important than personal traits, if significant at all. On the other hand, research by Iakovleva (2014) indicates that: “Concerning practical implications, results indicate that business start-up can be stimulated through improving regulative and cognitive institutional structures, but national differences need to be taken into account.”
Naffziger (1994) mentions following factors from business environment: societal attitudes toward starting a business, societal attitudes toward business in general, the economic climate of the market, the availability of accessible funds, and the importance of membership in an entrepreneurship-supportive network and Kouriloff (2000) defines three categories of barriers based on their importance – “Significantly More Important Barriers”, “Mild-level Importance Barriers” and “Significantly Less Important Barrier” and categorized 28 barriers. Multiple articles conclude different barriers to be the most important, for example Samuel (2013) concludes lack of collateral, difficulty of obtaining bank finance and lack of saving to be the most influential barriers. Jain (2013) conducted an extended literature review on possible barriers in starting new venture. Overview of all barriers used in our research can be found in the Appendix 1.
14
intention among Turkish students and concludes that educational and structural support factors affect students’ intention. Support factors used by Turker (2009) will be part of our initial support factors list, which we elaborate on through literature review (Lüthje, 2002; Turker, 2009) and interviews with small number of start-up founders, due to their experience and knowledge in this topic. A list of support factors created through literature review will be presented to interviewees, while they are asked to elaborate on the list if possible. The resulting list was used for pre-test and can be found in the Appendix 2.
2.3. Entrepreneurship Education
The first entrepreneurship course has been taught at Harvard’s Business School by Myles Mace in 1947 (Katz, 2013) and ever since that according to Katz entrepreneurship education grew at universities through three domains – “courses, supplemental infrastructures and publications”. According to paper by Roudaki (2009) most students in the sample believed that entrepreneurship education obtained at university is a key prerequisite for them to become successful entrepreneurs, regardless of having a family member occupied as an entrepreneur.
Students’ entrepreneurship awareness researched by Fretschner (2013) shows interesting insight into how can university shapes students’ perception and intent towards entrepreneurship. After taking awareness course, students’ self-perception shifted and they were more aware of their own abilities. Another insight is students’ affection to social factors, mostly favorable “entrepreneurial climate” in the university setting. Besides that, awareness course supposedly lowered negative effect of external barriers on students’ entrepreneurial intent through stressing the impact students’ personal setting on start-ups’ results despite external factors.
Sánchez (2013) also looked into changes in students’ perception of entrepreneurship after the entrepreneurship education relative to “pretreatment” group. Results show that entrepreneurship education has significantly positive effect on all measured factors – self-efficacy, proactiveness and risk taking, therefore also intention toward self-employment increased.
15 2.4. Conceptual Model
16
Figure 2 - The Conceptual Model
2.5. Hypotheses
The previous chapter stresses the previously conducted research in this area and show a lack of research on this topic considering possible changes in the context and students’ reaction towards those changes. During our research we will focus on changes in contextual factors that can be changed in short-term by universities or governments and are perceived as the most important factors by students. Importance of different barriers is addressed by Kouriloff (2000) and we are expecting significantly negative effects of barriers on entrepreneurship intention of students. Due to increasing barriers for starting a new venture motivated students might consider regular employment as a more attractive choice, relative to entrepreneurship. International sample of students might result in slight differences in perception of importance of barriers (Iakovleva, 2014; Walter, 2012). Although there are expected to be differences between cultures, external barriers are expected to have significantly negative effect on entrepreneurial intent (Chorev, 2006; Jain, 2013; Kouriloff, 2000).
17
Support factors have in all cases positive effect on entrepreneurship intention, although in some cases have the weakest effect on intention among measured variables (Fretschner, 2013; Katz, 1992; Peng, 2013), in other studies were external factors perceived as more important ones (Kristiansen, 2004; Roudaki, 2009). Due to these differences in results we are unable to predict exact effect of support factors on students’ intent, nonetheless, this effect is expected to be positive. As Lüthje (2002) concluded, even slightly improved university context can result in significantly higher students’ entrepreneurial intent, therefore we are expecting higher entrepreneurial intent as a result of increasing support factors.
H2: Support factors will have positive effect on entrepreneurship intention
Job offer will be set as a no-choice option in choice-based conjoint analysis and there might be few issues regarding this setting. Job offer will be a constant, specific for each individual respondent and is expected to have negative effect on entrepreneurship intention due to appearance in each choice set and students might perceive it as the save choice and more fitting option to their personality traits (Hartog, 2010). Based on the global GUESS survey (Sieger, 2014) we expect to majority of students’ to perceive job offer as their main choice. According to Brazell (2006) in order to avoid loss of information about the choices we could use dual response format and obtain data about the most and the least preferred choices. Setting job choice as the no-choice option will give us interesting insights about under what context will students that do not plan start a new venture consider this option and the opposite.
H3: Job offer will have negative effect on entrepreneurship intention
18
H4: Attitude towards entrepreneurship will have moderating effect on the relation between contextual factors and entrepreneurial intent
H4a: Risk taking prosperity will have moderating effect on the relation between contextual factors and entrepreneurial intent
19
3. Methodology
3.1. Pre-test
For the purpose of our research we had to identify barriers and support factors that are perceived as the most important ones by students. In order to present to students variety of possible external factors we conducted interviews with 5 students that have experience with new ventures, where we presented lists of external factors from previously conducted research. The list of barriers was adopted from Jain (2013) and the list of support factors was based on research by Lüthje (2002) and Turker (2009). Both lists were extended by the inputs of the interviewees and are included in Appendix.
Extended lists were used for the pre-test, where we identified external factors perceived as the most important ones by a sample of 47 students and graduates that have experience with entrepreneurship. Results indicated that the following external factors are perceived as the most important ones:
Barriers Number of
respondents Support factors
Number of respondents
Bureaucracy 8 Local start-up community 4
Business risk 5 Available mentorship 3
Required capital to start a
business 4 Available start-up incubators 3
Business contacts 3 Economy provides many
opportunities for entrepreneurs 3 Opportunities to meet investors 3 Tax holidays from government 3 The creative atmosphere inspires to
develop ideas for new businesses 3
Table 1 – Perceived most important external factors
Due to the nature of our research we had to choose factors that can be adjusted by government and/or universities with relative ease and can be quantified and specific. In the case of support factors we decided to choose two factors, while we covered also some of other factors in the attribute levels.
3.2. Sampling
20
focused on the variance in the factors used in our research and designed the experiment according to those differences. We focused on students of business and related fields and on students interested in entrepreneurship, regarding their field of study. Due to the use of online platform for the survey we decided to distribute it online, using e-mail communication and social networks. Assistance from teachers helped with the distribution among students and resulted in higher number of respondents.
3.3. Independent variables
Attribute levels are used as independent variables in our research, where we used 3 attribute levels for each of the attributes determined by the pre-test. By defining the same number of levels for each attribute we avoid number-of-levels effect (Eggers, 2011).
Attributes and attribute levels Description Number of steps to founding a start-up (e.g. sign
documents, create bank account, etc.) – (written as Dutch / Slovak)
1 / 4 4 / 7 7 / 10
As the most important barrier was perceived bureaucracy, which in the case of starting a new venture is the number of necessary steps to its founding. Values are based on the statistics from the World Bank, where we used the actual value and created two more levels by ± 3 steps.
Required capital by government for founding a company
1 € 1000 € 5000 €
Required initial capital by government in order to found a company.
Features of available incubators courses
courses and offices
courses, offices and possible funding
This attribute combines multiple support factors that were perceived as important by our sample. Incubators can be organized by governments and/or by universities.
Taxes
no exemption
exemption of first 100 000 € exemption of first 200 000 €
Taxes exemption for a starting phase of start-ups is a support factor that some governments use in order to motivate people towards starting a new venture.
None option (Job Offer) – (written as Dutch / Slovak)
21 Position: within the field of your studies Company size: 150 employees
Location: within 30 km from University Holidays per year: 20 days
Salary: 2.766 € / 1.244 €
respondents in the salary level. In order to determine the salary level we used average salary of alumni and created semi-attractive job offer.
Table 2 – Attributes and attribute levels
3.4. Dependent variable
Dependent variable is the choice of alternative presented in each choice set, where are three alternatives of starting a new venture under certain contextual conditions or a “none option”, which is accepting the presented job offer.
3.5. Experimental design
The following part contained choice-based conjoint analysis, where we presented 12 choice sets, each containing 3 alternatives of starting a new venture under different hypothetical external conditions together with one “none option”, which was presented as a job offer. This design offers respondents a choice between becoming self-employed and being a job seeker. An example choice set from choice-based conjoint analysis is included:
Figure 3 – Example choice set
3.6. Moderating effects
22
personality traits – risk taking prosperity and locus of control, using a Likert scale from 1 to 5 we may be able to identify moderating effects. Appendix 3 is a list of used questions together with the used scales.
3.7. Data collection
23
4. Results
4.1. Sample
After clearance of data there are 47 respondents that will be used for the analysis of the results. 19 respondents are from the Dutch sample and 28 respondents from the Slovak sample. Gender distribution is balanced with 49% female and 51% male sample. Larger portion of the sample are graduate students – 57%, compared to undergraduate students – 43%.
Figure 4 – Gender distribution
Figure 5 – Student status distribution
Figure 6 – Age distribution
Figure 7 – Country of university distribution
4.2. Factor Analysis
First step in the analysis is to create factors from respondents’ personality traits, namely to create factors for “Risk Taking Prosperity”, “Internal Locus of Control” and “Attitude towards Entrepreneurship”. If possible, these factors will be used as moderating effects on respondents’ choices.
cut-24
off point and still may be considered as sufficient, therefore variables may be combined into one factor. Bartlett’s Test is for all factors significant (p < 0.05). Considering communalities, all variables have a value higher than commonly used 0.4, therefore the variances in all variables explained by all the extracted factors are sufficient.
Risk factor: Cronbach’s
Alpha
N of Items
0.630 3
Table 3 – Reliability Statistics
Locus factor: Cronbach’s
Alpha
N of Items
0.587 2
Table 4 – Reliability Statistics
Attitude factor: Cronbach’s
Alpha
N of Items
0.801 3
Table 5 – Reliability Statistics
For risk and attitude factors is Cronbach’s Alpha above 0.6, while for Locus factor is Cronbach’s Alpha = 0.587, which could be explained by the fact that this factor is created using only two variables (see limitations).
4.3. Model Fit
4.3.1. Part-worth Model
First, a part-worth model was used for estimation, which treats attributes as nominal. The fit of the model is assessed by comparison with the NULL model, therefore we test whether there is a difference between models. Based on the following table we assess the model to be significantly better than the NULL model (p < 0.01). To further assess the fit of the model is used Pseudo Chi-square distribution, where R2 should lie between 0 and 1, while values between 0.2 and 0.4 are desired. Following table shows R2 and R2 Adjusted for the part-worth model, where R2 Adjusted = 0.0815. R2 Adjusted gives a “penalty” for number of
parameters used in the model and unlike
LL(0) LL(B) X2 X2 critical (df=13)
R2 R2 Adjusted
-781.9 -702.1 25446.5 22.362 0.1020 0.0815
Table 6 – Likelihood Ratio Test
4.3.2. Numeric Attribute Effects
25
Figure 8 – Utilities of attribute levels
Figure 9 – Utilities of attribute levels - continuous
Based on the above charts we could assess that number of steps needed before founding a start-up and exemption of taxes for the first “X” € may have linear effects. The following table compares part-worth model with the linear model (number of steps before founding and taxes exemptions are linear).
LL(B) N par Df Hit rate Chi R2 R2 Adjusted
Model – only main effects (part-worth) -722,9778 9 38 40.78% 117.7844 0.0753 0.0549 Model – only main effects (#steps, taxes linear) -723,0309 7 40 40.43% 117.6782 0.0753 0.0548 Model (part-worth) -702.110 33 14 43.44% 159.5200 0.1020 0.0815 Model (#steps, taxes linear) -702.199 31 16 43.26% 159.3420 0.1019 0.0814
26
Based on the above table we can see slight change in the fit of the linear model, which is not significantly better than the part-worth model, based on Likelihood-ratio test, which is not significant with p>0.9. Therefore we will keep the part-worth model, which has slightly better hit rate.
4.4. Preference Estimates
For the purpose of our research is important to assess the effects of attribute levels on the preference of entrepreneurship over job seeking. In order to do so, we look at the table below, where are parameter estimates.
For the attribute number of steps before founding a start-up, where is data shown in following form – Dutch sample / Slovak sample, we can see that actual number of steps needed before starting a start-up by the date of the writing (4 / 7) is perceived slightly negative. Due to different entrepreneurship conditions in Netherlands and Slovakia, we used different values for this attribute, where is used the actual value and +-3 steps. Parameter estimates give us interesting insight and that is that decreasing the number relative to actual number of steps increases entrepreneurship preference and increasing number of steps has opposite effect, regardless the country.
Somehow similar effect, although with a slightly larger variance, is present in attribute required starting capital by government, where with the increasing required capital decreases entrepreneurship intention. Considering barriers, in both cases are negative effects with increasing values, hence the hypothesis H1 holds.
Attributes and levels β
# steps before founding a start-up (Dutch/Slovak) 1 / 4 4 / 7 7 / 10 0.2344 -0.0180 -0.2164 Required starting capital
1 € 1000 € 5000 € 0.4268 0.0085 -0.4352 Features of incubator courses
courses and offices
courses, offices and possible funding
27 no exemption exemption of first 100 000 € exemption of first 200 000 € -0.3352 -0.0265 0.3617
None option (Regular employment) 0.2561
Table 8 – Parameter Estimates
Regarding attribute features of incubator, it is interesting to see that incubator with only courses is perceived negatively, while incubators that offer on top of courses offices or offices and possible funding opportunities are perceived as positive features.
Taxes exemption has also effect increasing with increasing values, while no exemption and exemption of first 100 000 € have negative effect.
Overall, support factors are increasing, hence hypothesis H2 holds.
In case of none option, with the utility of 0.2561, effect of none option – regular employment has a significant effect, hence H3 holds.
Combinations of contextual factors that exceed utility of none option may be more attractive to students than a regular employment. Out of eighty one possible combinations of contextual factor levels are multiple with utility higher than none option. These combinations may be interesting for universities and governments as they represent contexts that may motivate more students towards entrepreneurship. Examples of these combinations are included in the following table.
Attribute Alternative 1 Alternative 2 Alternative 3
# steps before founding a start-up (Dutch/Slovak) 1 / 4 (0.2344) 4 / 7 (-0.0180) 1 / 4 (0.2344)
Required starting capital 1000 €
(0.0085)
1 € (0.4268)
1 € (0.4268) Features of incubator courses and offices
(0.1553)
courses, offices and possible funding
(0.3469)
courses, offices and possible funding (0.3469) Taxes exemption exemption of first 100 000 € (-0.0265) no exemption (-0.3352) exemption of first 200 000 € (0.3617) Utility 0.3717 (>0.2561) 0.4205 (>0.2561) 1.3698 (>0.2561)
Table 9 – Examples of attribute level combinations exceeding none option
4.5. Moderating effects
28
hold. Risk taking prosperity does not have a significant effect on contextual factors (appendix 6), hence H4a does not hold.
Moderating effect β Wald statistic p-value
Att*steps1 0,0261 0,112 0,74 Att*steps2 -0,0656 0,6586 0,42 Att*capit1 0,0702 0,8807 0,35 Att*capit2 0,0393 0,2334 0,63 Att*incub1 -0,1086 1,459 0,23 Att*incub2 -0,092 1,3727 0,24 Att*tax1 -0,0895 1,0814 0,3 Att*tax2 0,0107 0,0172 0,9
Table 10 – Moderating effect of Attitude toeards entrepreneurship
Locus of control has a significant moderating effect on required capital by government. By combining the main effect of required capital by government with the moderating effect “Loc*capit” we achieve following effects:
1 € (capit1) = 0.4268 – 0.2532 * LOC 1000 € (capit2) = 0.0085 + 0.1577 *LOC 5000 € (capit3) = -0.4352 + 0.0955*LOC
Thus, with increasing value of locus of control overall range of capital required by government becomes narrower, hence less relevant. Thus, students with higher locus of control put less importance on the initial capital required by government. Moderating effect of locus of control is significant only for one contextual factor, hence H4b does partially hold.
Moderating effect β Wald statistic p-value
Loc*capit1 -0.2532 12.2860 0.000
Loc*capit2 0.1577 3.9420 0.047
Loc*capit3 0.0955 (reference) (reference)
29 4.6. Importance of attributes
One of many advantages of choice-based conjoint analysis is also ability to assess the importance of individual attributes. Based on the calculation of ranges of each attribute we obtain relative importance of individual attributes, therefore we can conclude which attribute perceived respondents as the most important one and also which they perceived as the least important one. By looking at the following table we conclude that required initial capital by government is perceived as the most important attribute (30.15%), closely followed by featured of available incubator (29.70%), while number of steps needed before founding a new venture is perceived as the least important attribute (15.77%).
These results indicate on what contextual factors universities and governments should focus. By looking at first two attributes – number of steps before founding a start-up and required capital, we could assume that governments would be able to create better conditions for starting a new venture and the benefits from new emerging ventures would possibly be significant.
Attribute Range Importance
# of steps before founding 0.4509 15.77%
Required capital 0.862 30.15%
Features of incubator 0.8491 29.70%
Taxes exemption 0.6969 24.38%
Table 12 – Importance of Attributes
4.7. Segmentation of the sample 4.7.1. A priori segmentation
The initial possible difference between respondents’ preferences could possibly be based on the country of their university, which is shown in previous research that it may have a significant effect on entrepreneurship intention. Segmentation by country of university results in non-significant differences across all attributes except for none option, therefore we may conclude that for our sample is country of university not a significant variable – appendix 4.
30
sample for the differences in preferences by gender we did not find any significant difference in preferences – appendix 5.
4.7.2. Preference-based segmentation
For the purpose of segmenting our sample we decided to drop moderating effects from this model because of the number of included parameters, which results in negative degrees of freedom, therefore some models could not be estimated.
In the comparison of models will be considered mostly three criteria – AIC, AIC3 and classification error. Based on these criteria we will choose number of segments if possible. Model 1 does not have a minimum for AIC nor AIC3 within the possible number of segments for our sample. In this case is important to look at classification error, which indicates two segments as the optimal choice (0.0001), while also four segments may be considered (0.0046). In this case would be more reasonable to choose two segments due to the small sample size.
Covariates are added into the Model 2 in order to obtain better results. Demographics age and gender are added to the model and results in the following table indicate that segmentation model 1 and segmentation model 2 are both significantly improved models relative to aggregate model. On the other hand, improvement in the model after including covariates is not significant, therefore it is reasonable to choose segmentation model 1.
LL(B) N par Hit Rate R2 R2 Adjusted
Aggregate model -722,9778 9 40.78% 0,075 0,055
Segmentation model 1 -557,8072 19 59.75% 0,287 0,266
Segmentation model 2
(covariates: age, gender) -557,034 21 59.75% 0,288 0,267
Table 13 – Assessment of Model Fit
Based on our decision to choose 2 segments we have to describe them and conclude managerial implications. In the case of Segment 1, none option has a negative value, which by definition, means it is not relevant. This segment puts most weight on features of incubator (32%) and the least on the number of steps before founding attribute (13%), while putting almost the same importance to the rest of attributes. This segment may be students that their main focus is entrepreneurship, therefore we see regular employment as not relevant.
31
importance of attributes, we could assume that this segment consists of students that may consider entrepreneurship but perceive the initial investment required to be of a significant barrier. For governments and universities in order to make entrepreneurship more attractive for students from this group, they should look into possible contextual factors combinations that would attract these students – contextual factors combinations with utility exceeding none option. If we look at possible combinations, we find that even the contextual factors combinations with the highest utility (1.8295) does not exceed utility of none option (2.4415). Thus, students in segment 2 are probably not interested in entrepreneurship under any of the proposed external conditions.
Segment 1 Segment 2
# steps before founding 13% 9%
Required capital 27% 32%
Featured of incubator 32% 13%
Taxes exemption 29% 4%
None option not relevant 41%
Segment size 66% 34%
Table 14 – Comparison of Segments
Attributes and levels β (segment 1) β (segment 2) Wald(=) p-value
# steps before founding a start-up (Dutch/Slovak) 1 / 4 4 / 7 7 / 10 0.1657 0.0258 -0.1915 0.3503 -0.1514 -0.1988 0.9436 0.62
Required starting capital 1 € 1000 € 5000 € 0.3838 -0.0104 -0.3734 0.8687 0.1350 -1.0038 5.1250 0.077 Features of incubator courses
courses and offices
courses, offices and possible funding
-0.5433 0.1963 0.3470 -0.1145 -0.3371 0.4516 5.3031 0.071 Taxes exemption no exemption exemption of first 100 000 € exemption of first 200 000 € -0.4034 -0.0093 0.4128 0.1589 -0.0710 -0.0879 7.5765 0.023
None option (Regular employment) -3.0065 2.4415 122.2774 0.000
32
5. Discussion
The discoveries in our research indicate that changes in the contextual factors can result in higher interest of students in entrepreneurship. Unlike previously conducted research, we took into account possible hypothetical scenarios that include changes in external factors and we elaborated on existing findings by including dynamics in contextual factors. This research setting allowed us to predict what contexts would most likely result in more motivated students towards starting new ventures and on the other hand, under what conditions they would rather choose regular employment. Barriers and support factors both have significant effect on students’ choice and these results indicate similar pattern like results of Lüthje (2003). By incorporating the option of choosing regular employment we covered more realistic scenario, where students may consider also a job seeking as their main goal and we were able to create combinations of contextual factors that result in entrepreneurship being more attractive than a job offer presented to students. Personality traits in our research have different effects on students’ entrepreneurship intention compared to results of Lüthje (2003) and only one moderating effect is significant.
6. Managerial implications and Limitations
6.1. Managerial implicationsOur findings have managerial implications for governments and universities. Regarding factors manageable by universities, features of available incubator indicate insight into available features that may increase entrepreneurial intent. Universities may operate incubators for students on different scales, whether offering them only entrepreneurship courses or offer spaces and equipment. Based on our results universities should focus on offering courses, office space and possibilities of being funded. The best university incubator according to Ubi Global – Rice Alliance for Technology and Entrepreneurship, has in twelve years help with founding 169 start-ups. These new ventures raised together over 1.2 billion dollars, which may be of interest for universities.
33
should local governments consult them regarding this issue and follow their model in order to lower barriers for students.
Governments may also consider tax exemption for a certain amount of company’s income or for a starting period of new ventures. Regarding this factor, governments need to look into working examples that are applied and serve as a supporting factor for new ventures.
By defining “start-ups” as a form of new ventures may be governments able to set individual conditions needed by start-up that may be different from needs of other forms of new ventures.
6.2. Limitations
The main limitation of our study is the sample size, which is with 47 respondents not a relevant sample of the target group. This sample is insufficient for achieving adequate variance within the sample, hence we are not able to generalize our results to the target group. With a completion rate of 18.3% for Dutch sample and 8.6% completion rate for Slovak sample is a possible place for improvement in completion rate. Respondents were presented with twelve choice set without additional motivation. Due to complexity of each alternative (considering contextual factors) the choice between alternatives may be perceived as more complex. Therefore we would recommend use of additional motivation if possible. Average time to completing the survey for Slovak sample is 13.3 minutes, while for Dutch sample is time to completing relatively smaller, with 7.1 minutes. This significant variance in time may be due to limited use of English language in Slovak education institutions, which may have resulted in difficulty of understanding presented choices. In order to avoid possibility of this problem we recommend use of local language if possible.
7. Literature
Brazell J, Diener C, Karniouchina E, Moore W, Séverin V, Uldry P. (2006), “The no-choice option and dual response choice designs,” Marketing Letters, 17(4), 255-268.
Carey T., Flanagan D., Palmer T. (2010), “An Examination of University Student Entrepreneurial Intentions by Type of Venture,” Journal Of Developmental Entrepreneurship, 15(4), 503-517.
Chorev S, Anderson A. (2006), “Success in Israeli high-tech start-ups; Critical factors and process,”
Technovation, 26(2), 162-174.
Davidsson, P., & Honig, B. (2003), “The role of social and human capital among nascent entrepreneurs,” Journal of business venturing, 18(3), 301-331.
Eggers F, Sattler H. (2011), “Preference Measurement with Conjoint Analysis OVERVIEW OF STATE-OF-THE-ART APPROACHES AND RECENT DEVELOPMENTS,” Gfk-Marketing
Intelligence Review, 3(1), 36-47.
Fereidouni H, Masron T, Nikbin D, Amiri R. (2010), “Consequences of External Environment on Entrepreneurial Motivation in Iran,” Asian Academy Of Management Journal, 15(2), 175-196.
Fretschner M, Weber S. (2003), “Measuring and Understanding the Effects of Entrepreneurial Awareness Education,” Journal Of Small Business Management, 51(3), 410-428.
Hartog J, Van Praag M, Van Der Sluis J. (2010), “If You Are So Smart, Why Aren't You an Entrepreneur? Returns to Cognitive and Social Ability: Entrepreneurs versus Employees,”
Journal Of Economics & Management Strategy, 19(4), 947-989.
Iakovleva T, Kolvereid L, Gorgievski M, Sørhaug Ø. (2014), “Comparison of perceived barriers to entrepreneurship in Eastern and Western European countries,” International Journal Of
Entrepreneurship & Innovation Management, 18(2/3), 115-133.
Ismail, M., Khalid, S. A., Othman, M., Jusoff, H. K., Rahman, N. A., Kassim, K. M., & Zain, R. S. (2009), “Entrepreneurial intention among Malaysian undergraduates,” International Journal of
Business and Management, 4(10), p54.
Jain R, Ali S. (2013), “A Review of Facilitators, Barriers and Gateways to Entrepreneurship: Directions for Future Research,” South Asian Journal Of Management, 20(3), 122-163.
Katz J. A (1992), “Psychosocial Cognitive Model of Employment Status Choice,” Entrepreneurship:
Theory & Practice, 17(1), 29-37.
35
Kolvereid L. (1996), “Prediction of employment status choice intentions,” Entrepreneurship: Theory
& Practice, 21(1), 47-57.
Kourilo M. (2000), “Exploring Perceptions of A Priori Barriers to Entrepreneurship: A Multidisciplinary Approach,” Entrepreneurship: Theory & Practice, 25(2), 59-79.
Kristiansen S, Indarti N. (2004), “Entrepreneurial Intention among Indonesian and Norwegian Students,” Journal Of Enterprising Culture, 12(1), 55-78.
Lüthje C, Franke N. (2002), “Fostering entrepreneurship through university education and training: Lessons from Massachusetts Institute of Technology,” In European Academy of Management 2nd
Annual Conference on Innovative Research in Management, Stockholm (pp. 9-11).
Lüthje C, Franke N. (2003), “The ‘making’ of an entrepreneur: testing a model of entrepreneurial intent among engineering students at MIT,” R&D Management, 33(2), 135.
Mohamad N, Lim H, Yusof N, Kassim M, Abdullah H. (2014), “Estimating the Choice of Entrepreneurship as a Career: The Case of Universiti Utara Malaysia,” International Journal Of
Business & Society, 15(1), 65-80.
Naffziger D, Hornsby J, Kuratko D. (1994), “A Proposed Research Model of Entrepreneurial Motivation,” Entrepreneurship: Theory & Practice, 18(3), 29-42.
Peng, Z, Lu, G, & Kang, H. (2013), “Entrepreneurial intentions and its influencing factors: A survey of the university students in Xi’an China,” Creative Education, 3(08), 95.
Pfeiffer E. (1997), “What MIT Learned from STANFORD,” Forbes, 59.
Raguž I, Matić M. (2011), “Student's Perceptions and Intentions towards Entrepreneurship: The Empirical Findings from the University of Dubrovnik-Croatia,” International Journal Of
Management Cases, 13(3), 38-49.
Roudaki J. (2009), “University Students Perceptions on Entrepreneurship: Commerce Students Attitudes at Lincoln University,” Journal Of Accounting, Business & Management, 16(2), 36-53.
Røed K, Skogstrøm J. (2014), “Job Loss and Entrepreneurship,” Oxford Bulletin Of Economics &
Statistics, 76(5), 727-744.
Samuel, Y. A, Ernest K, Awuah, J. B. (2013), “An assessment of entrepreneurship intention among Sunyani Polytechnic Marketing students,” International Review of Management and Marketing, 3(1), 37-49.
Sánchez J. (2013) “The Impact of an Entrepreneurship Education Program on Entrepreneurial Competencies and Intention,” Journal Of Small Business Management, 51(3), 447-465.
Sieger P., Fueglistaller, U. & Zellweger, T. (2014), “Student Entrepreneurship Across the Globe: A Look at Intentions and Activities,” St.Gallen: Swiss Research Institute of Small Business and
36
Turker D, Sonmez Selcuk S. (2009), “Which factors affect entrepreneurial intention of university students?,” Journal Of European Industrial Training, 33(2), 142-159.
Walter S, Dohse D. (2012), “Why mode and regional context matter for entrepreneurship education,”
Entrepreneurship & Regional Development, 24(9/10), 807-835.
37
8. Appendix
External Barriers
Age/health issues Lack of access to capital and information Awareness of regulations Lack of capital
Business contacts Lack of skills
Business risk Lack of socialization to entrepreneurship in the home, school and society
Discrimination Low business income
Economic conditions Risk of capital losses
Excessive regulations Socialized ambivalence about competition and profit
Exclusion from traditional business networks Start-up logistics
Find location Taxes
Find marketing method time constraints
Find opportunity Time for family
Firms dishonest Unsupportive family/friends Government bias to large companies Variable business income Government policies discourage Work stress
Industrial relations
Appendix 1 – List of external barriers used for pre-test
Support Factors
Economy provides many opportunities for entrepreneurs
The creative atmosphere inspires to develop ideas for new businesses
Entrepreneurs are encouraged by a structural system including private, public, and non-governmental organizations
The courses foster the social and leadership skills needed by entrepreneurs
If I decided to be an entrepreneur, my family members support me
The courses provide students with knowledge required to start a new company
If I decided to be an entrepreneur, my friends support me
The education in university encourages me to develop creative ideas for being an entrepreneur
My university develops my entrepreneurial skills and abilities
The university actively promotes the process of founding a new company
My university provides the necessary knowledge about entrepreneurship
The university provides a strong network of new venture investors
My university supports building multi-disciplinary student teams
38
Risk Taking Prosperity Q1: "When I travel I tend to use new routes"
Q2: "I like to try new things" (e.g. exotic food or going to new places) Q3: "I have taken a risk in the last six months"
1=fully disagree 2=partially disagree 3=neutral 4=partially agree 5=fully agree Locus of Control
Q4: "I often feel - That’s just what the things are and there’s nothing I can do about it" (scale inversed)
Q5: "When everything goes right, I think that it’s mostly luck" (scale inversed)
1=fully agree 2=partially agree 3=neutral 4=partially disagree 5=fully disagree Attitude Towards Entrepreneurship
Q6: "I’d rather be my own boss than have a secure job" Q7: "You can only make big money if you are self-employed"
Q8: "I’d rather found a new company than be the manager of an existing one"
1=fully disagree 2=partially disagree 3=neutral 4=partially agree 5=fully agree Appendix 3 – Personality Traits and Attitude Towards Entrepreneurship questions
Attribute Wald (=) p-value
# steps before starting a start-up 2.5483 0.28
Required capital by government 1.6872 0.43
Features of incubator 0.6565 0.72
Taxes exemption 0.3005 0.86
None option 25.5523 0.00
Appendix 4 – Segmentation by Country of university
Attribute Wald (=) p-value
# steps before starting a start-up 0.1812 0.91
Required capital by government 1.5195 0.47
Features of incubator 2.3779 0.30
Taxes exemption 2.9995 0.22
None option 0.0084 0.93
Appendix 5 – Segmentation by Gender
Attributes and levels Class 1 Wald(=) p-value
# steps before founding a start-up (Dutch/Slovak) 1 / 4 4 / 7 7 / 10 0.2344 -0.018 -0.2164 11.8061 0.0027
39 1 € 1000 € 5000 € 0.4268 0.0085 -0.4352 Features of incubator courses
courses and offices
courses, offices and possible funding
-0.5022 0.1553 0.3469 36.3828 0.000 Taxes exemption no exemption exemption of first 100 000 € exemption of first 200 000 € -0.3352 -0.0265 0.3617 27.4241 0.000
None option (Regular employment) 0.2561 6.3302 0.012
Risk*steps1 0.0488 0.395 0.53 Risk*steps2 0,0691 0,7212 0,4 Risk*capit1 -0,103 1,9204 0,17 Risk*capit2 -0,0375 0,2085 0,65 Risk*incub1 -0,0107 0,0146 0,9 Risk*incub2 0,0068 0,0075 0,93 Risk*tax1 -0,1491 2,9987 0,083 Risk*tax2 -0,0153 0,0366 0,85 Loc*steps1 -0,0817 1,1741 0,28 Loc*steps2 0,0867 1,2398 0,27 Loc*capit1 -0,2532 12,286 0,00046 Loc*capit2 0,1577 3,942 0,047 Loc*incub1 -0,0538 0,3865 0,53 Loc*incub2 0,0333 0,1925 0,66 Loc*tax1 -0,1525 3,28 0,07 Loc*tax2 0,1492 3,6275 0,057 Att*steps1 0,0261 0,112 0,74 Att*steps2 -0,0656 0,6586 0,42 Att*capit1 0,0702 0,8807 0,35 Att*capit2 0,0393 0,2334 0,63 Att*incub1 -0,1086 1,459 0,23 Att*incub2 -0,092 1,3727 0,24 Att*tax1 -0,0895 1,0814 0,3 Att*tax2 0,0107 0,0172 0,9
