Small Business & Entrepreneurship
The Entrepreneurship-Prone Personality Traits of
Extraversion & Openness to Experience
to Predict Self-Employment: What Set the Stage?
Author: Sander Postma Student number: S3855295 Supervisor: Dr. M. Wyrwich Co-assessor: Dr. E.M. Meerstra-de Haan
Date: 12-01-2021 Word count: 12,808
Faculty of Economics and Business University of Groningen
Semester 1.2
2
Abstract
The aim of this study is to address empirically the two highly relevant entrepreneurship-prone personality traits of extraversion and openness to experience to predict self-employment. These traits are obtained from the Big Five Framework. The decision to become self-employed is rooted in the social cognitive career theory (SCCT). Given that SCCT highlights that external influences need to be examined, the social learning theory is of importance in this paper. Therefore, self-employed parents acting as observable role models for their children serve as a contextual factor influencing these two relationships. The novelty of this study in explaining self-employment, compared to the existing literature, is combining these two traits with self-employed parental role models. Based on household panel data from the German Socio-Economic Panel Study (SOEP), these relationships have been tested empirically. Empirical results reveal support for the positive effect of openness to experience to predict self-employment. However, no support is found for extraversion in the main analysis. Furthermore, there is no evidence for parental (specifically: paternal) role models moderating these two relationships positively. Additional analyses revealed some differences between men and women in predicting self-employment. In general, these findings contribute to an increasing understanding of personality and its interplay with contextual factors explaining subsequent self-employment. Finally, explanations, managerial and theoretical implications, limitations, and avenues for future research are proposed.
Keywords: Extraversion; Openness to Experience; Self-Employment; Self-Employed Parental Role Models; Big Five Framework
1. Introduction
3
managerial talents” (Barot, 2015). Specifically, new firm formation is increasingly being related towards economic growth (Knoben, Ponds & Van Oort, 2011). Therefore, in this paper, I stick to the first definition proposed by Shane & Venkataraman (2000) to define entrepreneurship.
There are many ways in which entrepreneurship may affect economic growth positively. First, entrepreneurs may introduce innovations to enter markets by means of new products or production processes (Acs & Audretsch, 1990). Second, entrepreneurs are often at the forefront in the early evolution of industries, as did Michael Dell, Thomas Edison, Henry Ford, and Bill Gates, to name just a few (Van Stel, Carree & Thurik, 2005). Third, they may increase productivity by increasing competition (Nickell, Nicolitsas & Dryden, 1997). Lastly, entrepreneurs might be inclined to work more efficiently since their income is related to their effort (Van Stel et al., 2005). Summarized in the words of Schumpeter (1934), entrepreneurship can be seen as “an engine for economic development”. Therefore, the firms created by these entrepreneurs are essential to accelerate economic growth. These facts show and justify the growing interest in the domain of entrepreneurship among politicians and researchers.
Given the importance of new firm formation, people who are creating them need to be examined. Therefore, studying determinants of these entrepreneurs is of utmost importance. Previous research identified personality characteristics (traits) that are associated with a higher chance of becoming an entrepreneur (Viinikainen, Heineck, Böckerman, Hintsanen, Raitakari & Pehkonen, 2017). Historically, personality traits have received extensive attention in the entrepreneurial context. Cantillon (1759) was one of the firsts looking at the attitude towards coping with risk. Uncertainty (Knight, 1921), creativeness (Schumpeter, 1942), imagination (Shackle, 1970) and alertness (Kirzner, 1983) were also taken into account. Specific traits, such as risk-taking propensity, need for achievement and self-efficacy are associated with entrepreneurship (Obschonka, Schmitt-Rodermund, Silbereisen, Gosling & Potter, 2013). The notion “the choice of a vocation is an expression of personality” (Holland, 1997, p. 7) shows the importance that has been given towards the role of personality in the entrepreneurship decision. Broader personality dimensions, such as the Big Five (extraversion, conscientiousness, agreeableness, openness to experience and neuroticism) have also been shown to be linked to entrepreneurial activity (Obschonka et al. 2013). Moreover, this approach is regarded as the best established and most valid across different cultures of personality structure (Obschonka et al. 2013) and will therefore be used in this paper as framework.
4
in the careers literature, it could facilitate the ability to explain why individuals choose a career as entrepreneur. This is particularly important, since being an entrepreneur is often considered as one of the most risky and unstructured choices an individual can make (Campbell, 1992). Since becoming self-employed is in essence a career decision, the theory deserves attention in the entrepreneurial context (Segal et al., 2002). Lent, Brown & Hackett (2000) highlight in their paper about SCCT that external influences need to be examined by career development theorists. This is because these influences complement internal motivations (Meoli, Fini, Sobrero & Wiklund, 2019).
Even though personality plays a role, according to SCCT its role should not be overestimated. It does not fully explain why some people become entrepreneurs and others do not. Literature shows that a strong predictor of entrepreneurship is parental entrepreneurship (Lindquist, Sol & Van Praag, 2015). Furthermore, they highlight some studies in their paper that mention having a self-employed parent increases the likelihood that a child will enter self-employment by a factor of 1.3 to 3.0. This can be clarified by means of the theory of social learning (Bandura, 1986). Moreover, this theory also assists to incorporate external influences demanded by the SCCT. Besides the personality of (potential) entrepreneurs, the determinant of role models in the family context is of utmost importance. This theory suggests that an individual learns by observing the actions of its parents transferring cues into “internal codes” (Chlosta, Patzelt, Klein & Dormann, 2012). These internal codes form a part of the individuals’ mental models and determine its decision policies (Bandura, 1986), which also includes the later occupational choice and the decision to become self-employed (Schröder & Schmitt-Rodermund, 2006). Parents serve as substantial role models in their children’s career choices since they influence their children’s aspirations and work values in adolescence and early adulthood (Halaby, 2003). Growing up in an entrepreneurial family offers an opportunity to learn from the self-employed parent as a role model and thereby getting a realistic job image about being self-employed (Chlosta et al., 2012). Why an individual becomes self-employed can be seen through two different lenses: a personality-driven and behavioral explanation (Chlosta et al., 2012). The first relates to a higher likelihood of becoming an entrepreneur because of certain traits, as highlighted earlier (i.e., Zhao & Seibert, 2006). However, the latter stresses that individuals tend to learn from others who are role models for them. Dyer, Gibb & Handler (1994) and more recently Carr & Sequeira (2007) showed that early exposure to parental role models in the family business affects the children’s attitude towards becoming self-employed. Furthermore, these children can take advantage of mentoring by their parents and could access their business networks (Chlosta et al., 2012). That is why I combine both lenses in this paper.
5
understanding of the entrepreneurial decision process. By complementing SCCT with the social learning theory, I enlarge knowledge about the decision to become self-employed. Social learning theory (parental influence) serves as an external factor (moderator) to this decision process. As research showed that entrepreneurship is mainly influenced by the traits of openness to experience and extraversion of the Big Five Framework (Caliendo, Fossen & Kritikos, 2011; Judge & Higgins, 1999), I take these two highest ranked traits into consideration in this study.Namely, in the study of Caliendo et al. (2011), the correlation coefficients between the Big Five personality variables in their sample were 0.60 for openness, 0.53 for conscientiousness, 0.66 for extraversion, 0.55 for agreeableness, and 0.59 for neuroticism (all were significant at the 1% level).
Personality traits have been researched widely (i.e., Zhao et al., 2010) in the relationship with becoming an entrepreneur, also based on data on household level (Caliendo et al., 2011). However, to the best of my knowledge there has not been published a study that focuses on these two highly relevant personality traits of the Big Five, combined with SCCT and social learning theory (especially parents as role models) in the entrepreneurial context. Therefore, the novelty of this study compared to the existing literature examining self-employment is combining these traits with parental role models.
To close this literature gap and showing highly relevant determinants of self-employment, I want to answer the following research question: What are the roles of the entrepreneurship-prone
personality traits extraversion and openness to experience to predict employment and how do self-employed parental role models influence these two relationships?
The study helps scholars to get additional insights in the personality traits – entrepreneurial decision relationship. In addition, the highly relevant role of self-employed parents as role models is examined. For policy makers, it helps to set policies for addressing entrepreneurship effectively and ultimately decreasing the level of failing firms.
The remainder of the paper is structured as follows. Firstly, a literature review of the extant research is provided, resulting in several hypotheses and a comprehensive conceptual model (section 2). In the third section I discuss the methodology consisting of the data collection and measurements of the constructs. After that, the results will be presented (section 4). The paper ends in section 5 with a discussion of the results leading to a conclusion. Moreover, managerial and theoretical implications, limitations and suggestions for further research will be given.
2. Literature review
2.1 Personality traits
6
the adult life cycle. The Big Five, originally proposed in Goldberg (1971), personality traits consist of extraversion, conscientiousness, agreeableness, openness to experience, neuroticism, and have been shown to be linked to entrepreneurial activity (Obschonka et al., 2013). However, in entrepreneurship research, there has been given insufficient attention to the Big Five personality characteristics (Antoncic, Bratkovic Kregar, Singh, DeNoble, 2015). They mention that its potential to assess several (rather than one) personality traits makes it worthwhile to discover differences between entrepreneurs and non-entrepreneurs. This is especially important because there is still a gap in entrepreneurship literature about the unique contributions of entrepreneurs as persons to the entrepreneurial process (Mitchell, Busenitz, Lant, McDougall, Morse & Smith, 2002).
7
they must work in stressful and uncertain environments in which the family-work balance is often fuzzy (Zhao & Seibert, 2006).
Taken all together, entrepreneurs should score high on the personality traits extraversion, conscientiousness, and openness to experience. It is assumed they score low on agreeableness and neuroticism. Research has shown that entrepreneurship is mainly influenced by the traits of openness to experience and extraversion (Caliendo et al. 2011; Judge & Higgins, 1999). Caliendo et al. (2011) showed in their sample of 22,000 individuals in 2005 and 2009 that the correlation coefficients were highest for those two of the Big Five personality traits (0.66 for extraversion and 0.60 for openness to experience). That is why it is chosen to incorporate these two personality traits in this study to see their influence in the actual entrepreneurial decision.
2.2 Social cognitive career theory (SCCT)
SCCT is derived from the field of vocational psychology, wanting to explain career-related decision- making behavior (Aure, Dui, Jimenez, Daradar, Gutierrez, Blasa & Sy-Chango, 2019). SCCT proposes that there are three social cognitive mechanisms relevant to career development: (1) self-efficacy beliefs, (2) outcome expectations and (3) goal representations (Lent et al., 1994). In their paper, the first refers to “people’s judgement of their capabilities to organize and execute courses of action required to attain designated types of performances”. In other words, it is their own assessment of own abilities, which must be sufficient to reach a desirable performance. The second refers to “the imagined consequences of performing particular behaviours” (Lent et al., 1994). Thirdly, by setting goals, people are able to organize and guide their behaviour and sustain it over a long period of time, and this increases the probability that desired outcomes will be reached (Lent et al., 1994). This paper focuses on the first two social cognitive mechanisms. The decision between being self-employed or working for a boss can, therefore, be viewed as a cognitive process in which individuals compare the relative desirability of both career options (Segal et al., 2002). Thus, regarding outcome expectations, if an individual believes self-employment leads to a more valuable outcome than working for a boss, then he or she is more likely to become self-employed. Examples of outcome expectations are related to beliefs about extrinsic reinforcement (receiving tangible rewards for successful performance), self-directed consequences (satisfaction about mastering a challenging task) or the process of performing an activity (absorption in the task itself) (Lent et al., 1994). However, the first social cognitive mechanism (self-efficacy beliefs) must be considered as well. This will be linked with the two personality traits below.
8
consistent and substantive correlations between some of the Big Five personality traits and Holland’s (1997) Big Six vocational interest themes (realistic, investigative, artistic, social, enterprising, and conventional). Openness to experience was correlated with artistic (r=.48) and investigative (r=.28), whereas extraversion was correlated with enterprising (r=.41) and social (r=.31). This shows again the importance of these two traits, as only conscientiousness and agreeableness of the Big Five personality traits were only once and weaker correlated with respectively conventional (r=.25) and social (r=.19) of Holland’s Big Six vocational interest themes.
Now the role of self-efficacy will be examined in more detail. Hereby is meant that an individual makes an assessment about its own capabilities regarding self-employment. This assessment of skills must be sufficient to become self-employed. As discussed, the role of personality plays a major role here. Regarding the two personality traits of the Big Five, extraversion manifests itself in socialization. Extraverts get energy by working and talking with other people, so once becoming self-employed this “social skills baggage” is a great initial advantage and incentive (Remeikienė, Startienė & Vasauskaitė, 2011). This social aspect helps to develop social networks more easily, which might result in stronger partnerships with clients and suppliers (Caliendo et al., 2011). Summarized, linking it to SCCT, if an individual believes it has this set of skills at its disposal and expects a positive outcome (i.e., higher earnings, higher satisfaction, or absorption in the task), it increases the likelihood that he or she will become self-employed.
The second personality trait, openness to experience, is crucial for the decision to become self-employed. This is because its plays an important role in the process of recognizing an entrepreneurial opportunity (Antoncic et al., 2015). People who are more alert to opportunities than others have cognitive frameworks that help them to recognize these opportunities (Kirzner, 1983). According to Caliendo et al. (2011), exploring new ideas, being creative and taking novel approaches to the entrepreneurial process are essential to start a new venture. Thus, linking it to SCCT, if an individual believes it possesses these capabilities and expects a positive outcome, it increases the probability he or she will become self-employed.
Based on the findings above, I propose a positive relationship between the entrepreneurship-prone personality trait extraversion and becoming self-employed:
H1a: The entrepreneurship-prone personality trait extraversion positively influences becoming self-employed.
I also propose a positive relationship between the entrepreneurship-prone personality trait openness to experience and becoming self-employed:
9
2.3 Social learning theory (self-employed parents as role models)
It must be reminded that SCCT mentioned that external influences need to be examined as well. Namely, SCCT focuses on the interplay among person, environment, and behavioral influences in career development (Lent et al, 2014). This gives room for the social learning theory to incorporate these external influences. In the social sciences, it is acknowledged that an individual choice is highly influenced by the process of socialization (Maccoby, 1992). According to Bandura’s (1986) social learning theory, the actions of others mainly influence an individual’s behaviour. Learning from role models can initiate similar behaviours, strengthen or weaken their restraints against a given behaviour or develop even new patterns of behaviour (Chlosta et al., 2012).
10
(1986) states that the stronger the effects of role models on the observing individual, the more relevance and credibility they entail for this individual. Therefore, given the importance of parents, it can be assumed that the Big Five personality traits are related to parenting behaviours. For example, regarding the two personality traits, Goldberg (1990) stated that extraverts correspond with “social, outgoing and high levels of positive effect”. Thus, self-employed parents scoring high on extraversion are more likely to show positivity during interactions with their children (De Haan, Deković & Prinzie, 2012), which strengthens the children’s behaviour regarding self-employment. Further, it reinforces the (self-efficacy) belief of children that extraversion is an essential trait to become successful as a self-employed individual, especially once the parents are successful (outcome expectation). Self-employed parents scoring high on openness to experience, are likely to provide new and stimulating experiences to their children (Prinzie, Stams, Deković, Reijntjes & Belsky, 2009). This also strengthens the children’s behaviour regarding self-employment, who even more enjoy new experiences. Furthermore, it reinforces the (efficacy) belief that openness to experience is needed to become a successful self-employed individual. Again, this is especially the case once parents are successful (outcome expectation). In this study, I stress the positive (rather than a potential negative) role displayed by parents, since self-employed parents should score high on these traits and pass these positive characteristics on to their children. Indeed, otherwise they will not be self-employed themselves in the first place.
Based on the findings above, I propose a positive relationship for self-employed parents as role models to moderate the relationship between (a) extraversion and becoming self-employed and (b) openness to experience and becoming self-employed:
H2: The higher the level of exposure to self-employed parental role models, the stronger the relationships between (a) extraversion and becoming self-employed and (b) openness to experience and becoming self-employed.
2.4 Conceptual model
11 Figure 1: Conceptual model
3. Methodology
In this paper I use the theory testing approach. The starting point for this research approach to test existing theories that have not been tested yet in the empirical world. I use the SCCT and the social learning theory (parental role models). These theories rarely have been used together in the entrepreneurial context based on (1) household data and (2) two highly important traits of the Big Five Framework, rather than all five. The aim is to explain the hypotheses with the help of these existing theories.
3.1 Data collection
Data was obtained from the German Socio-Economic Panel Study (SOEP). SOEP is a multidisciplinary household panel study covering a wide range of social and behavioural sciences: economics, sociology, psychology, survey methodology, econometrics and applied statistics, educational science, political science, public health, behavioural genetics, demography, geography, and sport science (Caliendo et al., 2011). The SOEP is an annual representative panel survey covering detailed information about the socio-economic situation of about 22,000 individuals living in 12,000 households across Germany (Caliendo et al., 2011). Thus, the database was already available and is categorized as secondary data. I used a single-nation sample to control for potential international and cultural variations (such as gender roles). This SOEP database consists of 60,168 observations in total. This covers one-, two- and three-time respondents for 2005, 2009 and/or 2013. The sample will be adjusted later.
3.2 Measurements
12
Extraversion. For the Big Five personality trait extraversion I use the three personality items in the SOEP questionnaires (Caliendo et al., 2011). It used a 7-point Likert scale ranging from 1 (does not apply to me at all) to 7 (applied to me perfectly). The three questions that were asked were: I see myself
as someone who (1) is communicative, talkative, (2) is outgoing, sociable, (3) is reserved. The last
question was reversed first.
Openness to experience. For the Big Five personality trait openness to experience I use the three personality items in the SOEP questionnaires (Caliendo et al., 2011). It used a 7-point Likert scale ranging from 1 (does not apply to me at all) to 7 (applied to me perfectly). The three questions that were asked were: I see myself as someone who (1) is original, comes up with new ideas, (2) values artistic
experiences, (3) has an active, lively imagination.
Becoming self-employed. According to Caliendo et al. (2011), there is a difference between entrepreneurship and self-employment. The former usually includes the risk bearing of innovation, whereas the latter goes along with income risk but not necessarily with innovation. However, in the dataset, self-employment serves as a proxy for entrepreneurship. Working individuals are self-employed when they report self-employment as their main activity. In the dataset, measurements for the Big Five were examined in 2005, 2009 and 2013. Transitions (into or out of) self-employment can be seen in the dataset when an individual has occupied different employment states in two different years: t and t+4. The year 2005 guides as baseline and is going to be compared with 2009. In this way, the role of the financial crisis of 2008 (still relevant in 2009) could be examined regarding the decision to become self-employed. It is chosen to exclude the year 2013 since I want to compare the outcomes with the study of Caliendo et al. (2011), who also used the waves of 2005 and 2009 of the SOEP questionnaire.
Self-employed parents as role models. More specifically, I look at the father as role model.
Paternal role model was coded as 1 if the father of the participating individual was self-employed and
coded 0 otherwise. At this point, the individual was 15 years of age. Since the database misses too many values for maternal role model (mother as role model), I decided to focus on paternal role model only. This is also based on the argument of Chlosta et al. (2012) that fathers play a more significant role in their children’s occupational decision than mothers do.
13
most self-employed persons are married or engage in a partnership. Therefore, a score of 1 means married, whereas a score of 0 includes all other categories in the database (i.e., single, divorced, or widowed). In addition, I controlled for amount of education or training (measured in rounded years). This is because more years of education/training results in better prospects as an employee (such as a higher salary). Therefore, the higher the amount of education/training, the less likely an individual becomes self-employed. I also controlled for industry sector. This is important since, for example, the service sector accounts for an increasing share of the economy (Schettkat & Yocarini, 2006) making it worthwhile for entrepreneurs to enter this industry. I controlled for manufacturing [M], services [S] and construction [C], leading to three control variables. A score of 1 indicated the relevant industry, whereas a 0 consisted of other industries or was filled in as does not apply. A list of industries in the SOEP questionnaire in 2005 and the relevance selection can be seen in table A in the appendix. The respondents that filled in does not apply consist of unemployed/ not employed people (i.e., students). Unemployment was coded 1 when the respondent answered does not apply on the industry sector, whereas a 0 meant being employed in one of the industries. Being non-employed and then starting a firm can refer to many situations, such as a student that starts a venture directly after studies or a housewife starting-up a firm after children are older. Therefore, unemployment is also a control variable. For all the variables where the respondent did not provide an answer, it is seen by SPSS as a missing value and is not used in the analyses.
3.3 Analysis
As there are not only observations per individual (living in households), but also for different years (2005, 2009 and 2013) when the SOEP questionnaire contained questions regarding the Big Five traits, I take into account the panel characteristics of the data. Even though STATA is a very convenient software package, I chose to use SPSS (version 26) because of familiarity with this software.
14
items that survived the factor analysis, again for the base year 2005. The Cronbach’s alpha (α) was calculated for every construct to make sure the scales are reliable, with normally a threshold of .7. Since I only use three items for each construct, a threshold of .6 is set. When this is below .6, items will be deleted to increase the internal consistency. It turns out to be that the α for extraversion is .673 and for openness to experience is .609, so there is no real need to delete items to increase the Cronbach’s alpha. It must be noted these values are rather low. For example, regarding the first, deleting the item reserved would increase the alpha towards .724. In case of the latter, deleting items would always lower the alpha. Since I use a threshold of .6, I continue with the same three items. After that, I calculated the scores of the constructs extraversion and openness to experience as a mean of the scores of items. To obtain first insights in the existence of relationships between constructs, a correlation matrix and descriptives will be presented in the next section. To test the hypotheses, a (binary) logistic regression analysis will be carried out by first incorporating control variables, then the main effect variables and ultimately adding the interaction variables. Of course, issues of multicollinearity will be checked first.
3.4 Sample becoming self-employed
15
4. Results
4.1 Checking assumptions for regression analysis
To make valid inferences from regression, the residuals of the regression should follow a normal distribution. The residuals are the error terms or, put differently, the differences between the observed value of the dependent variable and the predicted value. By means of a normal Predicted Probability plot it can be checked if this is indeed the case. It turns out to be that the data is not normally distributed. This might be the case since the data is sorted and just a subset of the total database. A non-normal distribution is not a severe issue since regression is very robust, meaning it can still be performed. Nevertheless, logistic regression is more suitable in this case than linear regression or general linear models that are based on ordinary least squares (OLS) algorithms. This is because the error terms (residuals) do not need to be normally distributed and homoscedasticity (meaning error terms do not vary much as the value of the predictor variable changes) is not a requirement. However, multicollinearity among the independent variables should be just a little, or even better non-existent. This will be shown in the next sections. Since the dependent variable is binary (0/1), I use binary logistic regression and build the model by a stepwise inclusion of variables.
4.2 Correlation matrix and descriptives
In order to obtain first insights in the existence of relationships between variables, a correlation matrix is made, as can be seen in table 1 below. The matrix shows that there are no issues regarding multicollinearity since there are no correlations higher than .7 or very high in general.
16
I used listwise exclusion of missing values, since only cases without missing values on all variables will be used in the analyses. This resulted into a total sample of N=7,209. For example, there is a significant positive relationship between extraversion and openness to experience, r=.352, p<.01. Furthermore, the correlation matrix shows that an increase in openness to experience is significantly related with an increase in start-up (r=.056, p<.05). A significant relationship is not found for extraversion and start-up at the .01 and .05 level but found at the .1 level. However, due to space constraints, I just report the first two levels here. The descriptives show that the mean (x̄ ) of the approximate age of the total sample of N=7,209 is 38.61 years. Also, the sample consists of more females than males (value of .46 where male was coded 1). Further descriptives of all variables such as the standard deviations and range (maximum – minimum), can be found in table C in the appendix.
4.3 Main binary logistic regression analysis
To test the hypotheses derived from the literature, I perform a binary logistic regression. First, all control variables are added in model 1 of table 2. Model 2 incorporates the main variables extraversion and openness to experience. Paternal self-employment is also included here, since this variable should be observed before it is incorporated with the interaction effects. Note that paternal self-employment is not interpretable in all interaction models. This is because it measures the effect of paternal self-employment only for the (hypothetical) case that the other interacted variable is zero. It is not interpretable as the traits range from a score of 1 to 7. The interaction between the separate personality traits and paternal self-employment can be seen in model 3.
17
Also, the Nagelkerke R2 is mentioned, which is an adapted version of the R2 that is used in linear
regression. It estimates the proportion of the total variance in the data that the model accounts for. Last, the -2-log-likelihood shows how much unexplained variation is present in the model: the higher the value the less accurate the model. Among the three models, the Nagelkerke R2 and -2-log-likelihood
show that, overall, the last model is best in predicting the dependent variable. However, the difference with the second model is marginal. Even though there are no issues with multicollinearity, the standard errors (S.E) are mentioned. This serves as a kind of robustness check since the threshold of .7 is often set lower. As all standard errors are below 2, there are no obvious issues regarding multicollinearity. At model 1, only the industry control variables of manufacturing, construction and services are insignificant. When looking at model 2, also marital status becomes insignificant. Note that all control variables except marital status and industry (M, C and S) are highly significant in all models. The control variable indicating whether respondents were not in employment (does not apply, i.e., unemployed/ non-employed) is to a lesser extend significant in all models (p<.05). Once an individual was not in employment, the likelihood of becoming self-employed was greater as compared to individuals in employment. Special attention is for the role of approximate age and approximate age2. Since the first
is significant and positive and the latter significant and negative, it can be said the effect of age is lessoned. In other words, there is an inverted U-shaped relationship between age and becoming self-employed. Furthermore, the significant and positive effect of gender (1=male) means the propensity to become self-employed is greater for men. Of the main variables in model 2, only openness to experience and paternal self-employment are significant (p<.01). This means that once openness to experience increases with 1 unit, becoming self-employed increases with .273. Based on this regression analysis, hypothesis 1a cannot be confirmed, but hypothesis 1b can be confirmed. The same effect counts for having a father who is self-employed, where becoming self-employed increases by .759. However, when incorporating the interaction effects (model 3), only openness to experience remains significant. The interaction terms are both highly significant (not reported here but they are .595 and .930 respectively). This means that paternal self-employment does not significantly moderate the relationships between (a) extraversion and becoming self-employed and (b) openness to experience and becoming self-employed based on this logistic regression.
As an alternative, I first entered the main effects, then adding only the significant control variables as in the main binary logistic regression and finally adding the interaction effects (total sample became N=7,229 and respondents becoming self-employed remained N=193). This is done to improve the regression, since insignificant variables might make the overall regression less accurate. However, it turned out to be that this did not led to improvement in terms of the Nagelkerke R2, -2-log-likelihood
18
this item was excluded in the construct extraversion. However, there was no major difference in significance when constructing this new variable and the new interaction effect. This was also the case once replacing the whole construct of extraversion for one of the three respective items in the main model (not reported). This procedure is not carried out for openness to experience, since (a) the Cronbach’s alpha could not become higher by excluding items and (b) the construct itself was already highly significant in the main logistic regression.
Once replacing both constructs of extraversion and openness to experience by their respective items, only original (p=.016 & p=.018) and values artistic experiences (p=.007 & p=.024) were significant and positive in model 2 and 3. Both items belong to the construct openness to experience. This regression can be seen in table 3 below. The control variables show the same (in)significance levels as in table 2. Items of extraversion remain insignificant (no support found for hypothesis 1a). Once
original increases with 1 unit, becoming self-employed increases with .166 (main effects) and .177
(interaction effects). For values artistic experiences this effect is .127 and .115, respectively. This shows partly support for hypothesis 1b. Having a self-employed father increases becoming self-employed by .774. Given that the interactions between paternal self-employment and all items are insignificant in model 3, there is no evidence to accept hypothesis 2.
19
Note that the Nagelkerke R2 (higher) and -2-log-likelihood (lower) are better in predicting
self-employment in model 2 and 3 than the regression did with only the constructs itself (table 2). This means that the respective items are preferred above mean-centred scores of the constructs to predict becoming employed. To act as a robustness check for the main effect variables only, I excluded paternal self-employment from this analysis (table not reported due to space constraints). The number of respondents became 8,224 wherefrom 215 respondents became self-employed. However, the (in)significance of the constructs (and its items) remained unchanged. Differences will be covered later in the discussion and conclusion section.
4.4 Additional separate binary logistic regression analyses
4.4.1 Analyses extraversion and openness to experience in two different regressions
In addition to the main logistic regression, there will be conducted sensitivity checks (additional analyses). These will be carried out to assess whether altering the analysis set-up might lead to different outcomes. Put differently, the data will be approached differently and this effect on the results will be examined. Furthermore, especially the last part of these additional analyses serves as an assessment of effect heterogeneity. Namely, there might exist effects for particular groups and constellations. As a side-analysis, I will look at the role of gender to incorporate effect heterogeneity.
The first additional analysis will exclude openness to experience and its interaction effect with paternal self-employment in order to try to obtain significant results for extraversion. It will be constructed in the same way as the main binary logistic regression in section 4.3. Since missing values for openness to experience do not need to be considered anymore, the total sample becomes larger (N=7,249, wherefrom 194 respondents became self-employed in 2009). The outcomes of this three-stage logistic regression model can be seen below in table 4.
20
As can be read, the control variables show the same (in)significance levels as the main logistic regression analysis for all models. However, by excluding openness to experience and its interaction with the moderator paternal self-employment, extraversion becomes significant at the .05 level in the second model. This level changes to .1 when incorporating the interaction effect (model 3). This regression shows support for hypothesis 1a. The interaction remains highly insignificant (.663 compared to .595 in the main analysis). There is no support for the first part of hypothesis 2. The second analysis is conducted for openness to experience, which can be found in table 5 below. Note that the total sample here became N=7,215, wherefrom 195 became self-employed.
Table 5: Additional logistic regression openness to experience (excluding extraversion)
Focussing on the second model of table 5, openness to experience is highly significant, but again, the interaction effect with paternal self-employment remains highly insignificant (.971 compared to .930 in the main analysis). In addition, this model shows that extraversion explains some variance, since the Nagelkerke R2 is slightly lower compared to the main regression analysis. The -2-log-likelihood is also
greater here, meaning a less accurate model. All in all, in this regression, hypothesis 1b is confirmed, whereas the second part of hypothesis 2 cannot be confirmed.
21
no real effect once the regression incorporates openness to experience as well. Both Cronbach’s alphas are below the commonly used threshold of .7. This might indicate that, in this case, items of extraversion measure different elements of openness to experience. This might result into an insignificant effect for extraversion to predict self-employment. Thus, some overlap in measurement might occur, which disappears when analysed both separately. That is why both table 4 and 5 have been computed as a robustness check.
4.4.2 Analyses items of extraversion and openness to experience in two different regressions At this point, is it valuable to mention that both constructs (and items) of extraversion and openness to experience should be incorporated in one regression. If analysed separately, both models become incomplete, as table 4 and 5 showed. The additional analyses so far obtained different results compared to the main analysis for extraversion – which is remarkable. Namely, the same pattern was expected. Therefore, despite the preference for all items in one regression (table 3), two separate regressions for the items of both constructs are computed. Nevertheless, these outcomes are only presented in tables E and F in the appendix. Based on these analyses for extraversion, the significant and positive item
communicative shows partly support for hypothesis 1a. The first part of hypothesis 2 cannot be
confirmed. For openness to experience, the positive and significant items original and values artistic
experiences are in both models significant at the .01 and .05 level, respectively. This shows partly
support for hypothesis 1b. The interactions are highly insignificant, meaning that paternal self-employment does not moderate the relationship between any of the items of openness to experience and becoming self-employed. Again, the second part of hypothesis 2 cannot be accepted based on this regression. Nonetheless, these results should be interpreted with caution since the main regression is preferred above two separate regressions. The outcomes are still reported (in the appendix) given the deviation with the main regression for extraversion in table 4. In the discussion section the results will be compared with the main regression.
4.4.3 Gender analyses
22
table 6 below. Once extraversion increases with 1 unit, becoming self-employed increases by .201 (model 2) and .187 (model 3). Surprisingly, the outcomes for men are completely the opposite compared to the main regression where extraversion was insignificant. Thus, based on this analysis, extraversion seems to be more important for men compared to the whole population (also incorporating women). This will be discussed later.
Table 6: Main logistic regression for gender (men only)
In table G in the appendix can be seen a side-analysis for men, but now focusing on the respective items of both extraversion and openness to experience. Again, examining the items rather than the constructs in table 6 increases the Nagelkerke R2 and lowering the -2-log-likelihood in model 2 and 3. This means
table G is better in predicting self-employment. Surprisingly, contrary to the previous men analysis, paternal self-employment is significant at the .1 level. Even more remarkable is that there are no items of extraversion found to be significant (whole construct was significant in table 6). Rather, the item
values artistic experiences of openness to experience is significant at the .05 level (model 2) and .1 level
(model 3). Further, the interaction between imagination and paternal self-employment is significant (r=-.579, p=.033). Contrary to my expectation, this effect is negative. This will be discussed later. However, given the higher accuracy of the analysis in table G, this analysis is preferred over the construct analysis. This again shows that (an item of) openness to experience is more important than (items of) extraversion. Moreover, paternal self-employment is now found to be significant. Summarized, hypothesis 1a cannot be confirmed, whereas hypothesis 1b partly can. Paternal self-employment is significant as main variable and significant in the interaction with the item imagination. Since this effect is negative, hypothesis 2 cannot be accepted.
23
sample of N=3,866, wherefrom 93 women became self-employed. Paternal self-employment was highly significant and positive in the second model. Nevertheless, contrary to the main regression of men, openness to experience is in both models highly significant and positive (p=.00 twice). Extraversion was always insignificant. The main regression analysis for women can be seen in table 7. Once openness to experience increases with 1 unit, becoming self-employed increases by .431 (model 2) and .434 (model 3). These outcomes are in line with the main regression (table 2).
Table 7: Main logistic regression for gender (women only)
In table H in the appendix can be seen a side-analysis for women, focusing on the respective items of both constructs. Focusing on the items rather than the constructs in table 7 increases the Nagelkerke R2
and lowering the -2-log-likelihood in model 2 and 3. This means that table H is better in predicting self-employment. Paternal self-employment is highly significant at the .01 level. Also, for women, there is not found any significant item of extraversion. Instead, the item original of openness to experience is highly significant in model 2 and 3. There is not found any significant interaction between paternal self-employment and one of the items. The item analysis is preferred, given the higher accuracy compared to the construct analysis. However, both analyses for women show that (an item of) openness to experience is more important than (items of) extraversion. Summarized, hypothesis 1a cannot be confirmed, whereas hypothesis 1b can. There is no evidence to accept hypothesis 2.
5. Discussion and conclusion
5.1 Explanations and interpretation
24
extraversion and openness to experience were chosen as main personality variables influencing the decision the become self-employed. This selection was strengthened by the findings of Larson et al. (2002), who linked in their meta-analysis mainly these two traits to Holland’s (1997) Big Six vocational interest themes. Holland (1997) mentioned that “the vocational interests are a direct expression of personality” – which was empirically confirmed in their meta-analysis mainly for extraversion and openness to experience. The decision to become self-employed is a distinctive career choice and therefore rooted in the SCCT. This paper followed the first two social mechanisms presented by SCCT that self-efficacy beliefs and outcome expectations determine if an individual decides to be an employee or self-employed. Individuals scoring high on extraversion tend to have “social skills baggage”. If an individual believes it has this set of skills and expects a positive outcome (i.e., higher earnings), it was hypothesized that extraversion increases the likelihood he or she will become self-employed. Individuals scoring high on openness to experience are more alert to opportunities and have cognitive frameworks to recognize them. If an individual believes it possesses these capabilities and expects a positive outcome, it was hypothesized that openness to experience increases the chance of becoming self-employed. To test my hypotheses, I made use of household panel data provided by the SOEP. This assisted in answering the research question: What are the roles of the entrepreneurship-prone
personality traits extraversion and openness to experience to predict employment and how do self-employed parental role models influence these two relationships? The second part will be discussed
later in this chapter.
25
openness to experience, where the items original and values artistic experiences are found to be significantly positive. This shows some support for hypothesis 1b, given that not all items are significant. It could be the case that imagination shows some overlap in measurement with (one of) the other items. Being original/coming up with new ideas and valuing artistic experiences are better predictors for becoming self-employed than being original or items of extraversion. This is an important finding, since the item analysis is a more accurate analysis than the mean-centered construct analysis shown by the Nagelkerke R2 and -2-log-likelihood.
Table 4 and 5 were computed as robustness checks and showed that separating constructs in two regressions is not preferred, making both regressions incomplete and therefore less accurate. However, it was chosen to conduct an additional analysis in the same manner as these tables. This was carried out for the respective items (rather than the constructs), in tables E and F. Because extraversion in table 4 deviated from the main analysis – which was unexpected – these regressions have been computed as additional check. Once separated, the item communicative was significant of extraversion. The items
original and values artistic experiences were, again, significant for openness to experience. As expected,
both table E and F were more accurate in predicting self-employment than table 4 and 5 (given Nagelkerke R2 and -2-log-likelihood). This confirms again that an item analysis is preferred above a
(mean-centered) construct analysis. However, table E and F were less accurate than the main analysis in table 3, which confirms that one regression is preferred above two. Summarized, the main regression is leading (table 3 and 2 in that order). The additional separate item analyses and construct analyses follow in that order – but are less accurate and should be interpreted with caution. Therefore, the significant item communicative of extraversion is not really reliable, given the insignificance once combining all items of both constructs. The significant items of openness to experience support the outcomes in the main regression.
The second part of the research question introduces self-employed parental role models. SCCT mentioned that external influences (besides personality) should be considered in the decision to become self-employed. This gave room to incorporate the social learning theory in this paper. According this theory, (observational) learning from role models influences behaviour. Since children are mainly exposed to parental behaviours, self-employed parents as role models are a major source of shaping their behaviour. This study focused on the positive role of self-employed parents, who score high on these traits and pass the positive characteristics on to their children. A negative role (i.e., stressing negative aspects of self-employment) would not be optimal, since then it is questionable why parents became employed in the first place. It was hypothesized that the higher the level of exposure to self-employed parents, the stronger the relationships between (a) extraversion and becoming self-self-employed and (b) openness to extraversion and becoming self-employed.
26
hypothesis 2. Once analyzing the items of extraversion and openness (table 3), there is still not found a significant result for the interaction effects. The same holds in the (less accurate) additional analyses. Given the significance as main effect in all analyses, the role of a father as self-employed remains important in explaining becoming self-employed. However, his role does not strengthen the relationships between (a) extraversion and becoming self-employed and (b) openness to experience and becoming self-employed. A potential reason might be that self-employed fathers are not able to influence these two relationships once their children are 15 years of age. For example, their influence might be greater once their children are younger. Another possibility is that children of self-employed fathers do not observe or interact that much with their fathers as the literature suggests. Even though paternal self-employment is present, it might be less visible for children. Moreover, divorced parents might make it more difficult for children who live at their mother to observe or interact with their father more frequently.
As a side-analysis, I investigated the role of gender in the decision to become self-employed. For two reasons, I was mainly interested in a sample consisting of only males. First, because my earlier findings showed that men have a greater propensity to become self-employed. Second, given the moderating effect of paternal self-employment in the literature, my interest derived from the finding of Dunn & Holtz-Eakin (2000). They showed that sons of employed fathers more often enter self-employment than sons of self-employed mothers.
The main construct regression for men showed that paternal self-employment was not significant (against the literature) and only extraversion was significant in the main effect and interaction model. Openness to experience was always insignificant. Surprisingly, this was completely the opposite compared to the main regression (incorporating both genders). Since previous analyses showed that (a) an item analysis is preferred above a construct analysis and (b) one regression is more accurate than two, table G was computed. This item analysis is contradicting the construct analysis for men but is in line with the main regression analysis. Namely, paternal self-employment became significant. No single item of extraversion was found to be significant, therefore fully contradicting the construct analysis for men. Rather, the item values artistic experiences of openness to experience is significant. Given the preference for an item analysis of both constructs in one regression (higher Nagelkerke R2 and lower
self-27
employed might weaken the (self-efficacy) belief of children that imagination is essential for becoming self-employed. Given that this effect is negative, there is no support for hypothesis 2.
Since the regression for men showed surprising results, it was decided to compare this group with women. The main construct regression for women showed that paternal self-employment was highly significant. Extraversion was not significant, whereas openness to experience was in both models found to be highly significant. This is in line with the main regression (incorporating both genders), but the opposite of the construct analysis for men. This difference between both genders will be examined and explained only by means of the item analyses of both constructs (table G and H). This is because these analyses are better/more accurate in predicting self-employment than the construct analyses do (table 7 for women). The item analysis for women shows that paternal self-employment is highly significant. No items of extraversion were significant. The item original of openness to experience is highly significant. All in all, these analyses for women show that hypothesis 1a cannot be accepted, whereas the construct analysis shows support for hypothesis 1b. The item analysis shows partly support for hypothesis 1b. Since there is not found any significant interaction effect between paternal self-employment and one of the items, there is no evidence to accept hypothesis 2.
28
My findings are not fully in line with the study of Caliendo et al. (2011). They reported significant results for both traits, and the effect should even be stronger for extraversion. What causes these different outcomes? Given that extraversion is significant without incorporating openness to experience but is insignificant once incorporating both traits, there seems to be some overlap in measurement. Both Cronbach’s alphas are below the commonly used threshold of .7. This might indicate that, in this case, items of extraversion measure different elements of openness to experience. This might result into an insignificant effect for extraversion to predict self-employment in the full regression analysis.
All in all, my study extends the study of Caliendo et al. (2011) by incorporating paternal self-employment as a moderator on the relationship between (a) extraversion and becoming self-employed and (b) openness to experience and self-employment. Further, as a side analysis, the role of gender was examined for these traits. Moreover, the moderating role of the father as self-employed was investigated for gender as well.
5.2 Managerial and theoretical implications
29
5.3 Limitations and future research
30
subjective to overestimation bias. Seventh, since I used logistic regression, it would be preferable to have more cases (observations). Even though the sample of N=7,209 for the main analysis is quite large, the likelihood of becoming self-employed is rather low (N=193). A larger database would be more representative as well. Furthermore, to increase the internal consistency, more items to measure a construct would be preferred in future studies. The SOEP database only used three items for both constructs, which is a low amount. More items would also allow to delete low or wrong loading items. Eighth, my study only examined two of the Big Five traits. Even though the choice is justified, the Caliendo et al. (2011) study showed that the correlation coefficients did not differ that much in their sample (i.e., openness to experience 0.60, neuroticism 0.59). It would be of added value to investigate the remaining three traits as well. This could lead to unexpected results, as did my study for extraversion. Nineth, my findings showed that gender explains differences in the decision to become self-employed. This revealed that gender is an important antecedent that deserves more attention than just a side analysis as it was in this paper. Future research should take this into account. Last, my control variables are straightforward, but other control variables might have influence as well. For example, migration background could influence the decision to become self-employed. As an example, Clark & Drinkwater (1998) suggested that self-employment is not always a freely choice of migrants, but a way to escape discrimination in paid employment. My control variables show interesting results that might be interesting to study as a dependent variable, such as the amount of education/training. For example, entrepreneurship education (courses) might have a strong potential to assist people to gain skills and generate their own skilled jobs (Premand, Brodmann, Almeida, Grun & Barouni, 2016). This paper serves as a motivation to encourage scholars to investigate antecedents of the decision to become self-employed.
References
Acs, Z.J., Audretsch, D.B. 1990. Innovation and Small Firms (Cambridge, MA: MIT Press).
Antoncic, B., Bratkovic Kregar, T., Singh, G., DeNoble, A.F. 2015. The Big Five Personality-Entrepreneurship Relationship: Evidence from Slovenia. Journal of Small Business Management, 53(3): 819-841.
Aure, P.A.H., Dui, R.P., Jimenez, S.V., Daradar, D.D., Gutierrez, A.N.A., Blasa, A.C., Sy-Changco, J. 2019. Understanding social entrepreneurial intention through social cognitive career theory: A partial least squares structural equation modelling approach. Organizations and markets in emerging
economies, 10(1): 92-110.
31
Barot, H. 2015. Entrepreneurship - A Key to Success. The International Journal of Business and
Management, 3(1): 163-165.
Bosma, N., Hessels, J., Schutjens, V., Van Praag, M., Verheul, I. 2012. Entrepreneurship and role models. Journal of Economic Psychology, 33(2): 410-424.
Caliendo, M., Fossen, F., Kritikos, A. 2011. Personality traits and the decisions to become and stay self-employed. Small Business Economics, 42(4): 787–814.
Campbell, C.A. 1992. A decision theory model for entrepreneurial acts. Entrepreneurship Theory and
Practice, 17(1): 21–27.
Cantillon, R. 1759. Essai sur la Nature du Commerce in Général (London: Reissued in 1959 for the Royal Economic Society).
Carr, J.C., Sequeira, J.M. 2007. Prior family business exposure as intergenerational influence and entrepreneurial intent: A theory of planned behavior approach. Journal of Business Research, 60: 1090– 1098.
Chlosta, S., Patzelt, H., Klein, S.B., Dormann, C. 2012. Parental role models and the decision to become self-employed: The moderating effect of personality. Small Business Economics, 38: 121-138.
Clark, K., Drinkwater, S. 1998. Ethnicity and self-employment in Britain. Oxford Bulletin of Economics
& Statistics, 60(3): 383-407.
Costa, P.T., Jr., McCrae, R.R. 1992. Revised NEO Personality Inventory (NEO-PI–R) and NEO Five
Factor Inventory (NEO-FFI) professional manual (Odessa, FL: Psychological Assessment Resources).
De Haan, A.D., Deković, M., Prinzie, P. 2012. Longitudinal Impact of Parental and Adolescent Personality on Parenting. Journal of Personality & Social Psychology, 102(1): 189-199.
Dunn, T., Holtz-Eakin, D. 2000. Financial capital, human capital and the transition to self-employment: Evidence from intergenerational links. Journal of Labor Economics, 18(2): 282–305.
32
Fairlie, R.W., Robb, A. 2007. Families, human capital, and small business: Evidence from the characteristics of business owners survey. Industrial and Labor Relations Survey, 60: 225–245.
Fellnhofer, K., Mueller, S. 2018. “I Want to Be Like You!”: The Influence of Role Models on Entrepreneurial Intention. Journal of Enterprising Culture, 26(2): 113-153.
Gartner, W.B. 1988. Who is an entrepreneur? Is the wrong question. American Journal of Small
Business, 12(4): 11-33.
Goldberg, L. R. 1971. Five models of clinical judgment: An empirical comparison between linear and nonlinear representations of the human inference process. Organizational Behavior and Human
Performance, 6(4): 458–479.
Goldberg, L.R. 1990. An alternative “description of personality”: The Big-Five factor structure. Journal
of Personality and Social Psychology, 59: 1216–1229.
Greene, F.J., Han, L., Marlow, S. 2013. Like Mother, Like Daughter? Analyzing Maternal Influences Upon Women's Entrepreneurial Propensity. Entrepreneurship: Theory & Practice, 37(4): 687-711.
Halaby, C.N. 2003. Where job values come from: Family and schooling background, cognitive ability, and gender. American Sociological Review, 68: 251-278.
Hamilton, E. 2011. Entrepreneurial Learning in Family Business: A Situated Learning Perspective.
Journal of Small Business and Enterprise Development, 18(1): 8–26.
Hamilton, B.H., Papageorge, N.W. Pande, N. 2019. The right stuff? Personality and entrepreneurship.
Quantitative Economics, 10(2): 643-691.
Hoffmann, A., Junge, M., Malchow-Møller, N. 2015. Running in the family: Parental role models in entrepreneurship. Small Business Economics, 44: 79–104.
Holland, J.L. 1997. Making vocational choices: A theory of vocational personalities and work
environments (Odessa, FL: Psychological Assessment Resources).
33
Hundley, G. 2006. Family Background and the Propensity for Self-Employment. Industrial Relations, 45(3): 377-392.
Judge, T.A., Higgins, C.A. 1999. The Big Five personality traits, general mental ability, and career success across the life span. Personnel Psychology, 52(3): 621-652.
Kets de Vries, M.F.R. 1977. The entrepreneurial personality: A person at the crossroads. The Journal of
Management Studies, 14(1): 34–57.
Kim, P., Aldrich, H.E., Keister, L.A. 2006. Access (not)denied: The impact of financial, human, and cultural capital on entrepreneurial entry in the United States. Small Business Economics, 27(1): 5–22.
Kirzner, I.M. 1983. Entrepreneurship and the future of capitalism, in: J. Bachman (Ed),
Entrepreneurship and the Outlook for America (New York, NY: Free Press).
Knight, F.H. 1921. Risk, Uncertainty and Profit (New York: Harper and Row).
Knoben, J., Ponds, R., Van Oort, F. Employment from new firm formation in the Netherlands: Agglomeration economies and the Knowledge Spillover Theory of Entrepreneurship. Entrepreneurship
& Regional Development, 23(3/4): 135-157.
Larson, L.M., Rottinghaus, P.J., Borgen, F.H. 2002. Meta-analyses of Big Six interests and Big Five personality factors. Journal of Vocational Behavior, 61: 217–239.
Laspita, S., Breugst, N., Heblich, S., Patzelt, H. 2012. Intergenerational transmission of entrepreneurial intentions. Journal of Business Venturing, 27: 414–435.
Lent, R.W., Brown, S.D., Hackett, G. 1994. Toward a unifying social cognitive theory of career and academic interest, choice, and performance. Journal of Vocational Behavior, 45: 79–122.
Lent, R.W., Brown, S.D., Hackett, G. 2000. Contextual supports and barriers to career choice: A social cognitive analysis. Journal of Counseling Psychology, 47(1): 36–49.
34
Lindquist, M.J., Sol, J., Van Praag, M. 2015. Why Do Entrepreneurial Parents Have Entrepreneurial Children? Journal of Labor Economics, 33(2): 269-296.
Maccoby, E.E. 1992. The role of parents in the socialization of children: An historical overview.
Developmental Psychology, 28: 1006–1017.
Maitlo, A.A., Memon, S.B., Kumar, M. 2020. The mediating role of networking orientation between entrepreneurial personality characteristics and entrepreneurial intentions. Advances in Business-Related
Scientific Research Journal, 11(1): 48-66.
Meoli, A., Fini, R., Sobrero, M., Wiklund, J. 2019. How entrepreneurial intentions influence entrepreneurial career choices: The moderating influence of social context. Journal of Business
Venturing, [105982]. https://doi.org/10.1016/j.jbusvent.2019.105982
Minniti, M., Arenius, P., Langowitz, N. 2005. GEM report on women and entrepreneurship, Resource
document. Babson College. Retrieved from http://www3.babson.edu/
CWL/research/GEM-Report-on-Women-and-Entrepreneurship.cfm
Mitchell, R.K., Busenitz, L., Lant, T., McDougall, P.P, Morse, E.A., Smith, B.J. 2002. Entrepreneurial Cognition Theory: Rethinking the People Side of Entrepreneurial Research. Entrepreneurship Theory
and Practice, 27(2): 93–104.
Mungai, E., Velamuri, S.R. 2011. Parental Entrepreneurial Role Model Influence on Male Offspring: Is It Always Positive and When Does It Occur? Entrepreneurship: Theory & Practice, 35(2): 337-357.
Nickell, S., Nicolitsas, P., Dryden, N. 1997. What makes the firm perform well? European Economic
Review, 41: 783-796.
Obschonka, M., Schmitt-Rodermund, E., Silbereisen, R.K., Gosling, S.D., Potter, J. 2013. The regional distribution and correlates of an entrepreneurship-prone personality profile in the United States, Germany, and the United Kingdom: A socioecological perspective. Journal of Personality & Social
Psychology, 105(1): 104-122.
35
Pfeifer, S., Šarlija, N., Zekić Sušac, M. 2016. Shaping the entrepreneurial mindset: entrepreneurial intentions of business students in Croatia. Journal of Small Business Management, 54(1): 102-117.
Premand, P., Brodmann, S., Almeida, R., Grun, R., Barouni, M. 2016. Entrepreneurship Education and Entry into Self-Employment Among University Graduates. World Development, 77: 311-327.
Prinzie, P., Stams, G.J.J.M., Deković, M., Reijntjes, A.H.A., Belsky, J. 2009. The relations between parents’ Big Five personality factors and parenting: A meta-analytic review. Journal of Personality and
Social Psychology, 97: 351–362.
Remeikienė, R., Startienė, G., Vasauskaitė, J. 2011. The influence of psychological-sociological factors on self-employment. Economics & Management, 16: 287-294.
Reynolds, P.D. 2016. Start-up Actions and Outcomes: What Entrepreneurs Do to Reach Profitability.
Foundations and Trends in Entrepreneurship, 12(6): 443–559.
Schröder, E., Schmitt-Rodermund, E. 2006. Crystallizing enterprising interests among adolescents through a career development program: The role of personality and family background. Journal of
Vocational Behavior, 69(3): 494–509.
Schumpeter, J.A. 1934. The Theory of Economic Development (Harvard University Press, Cambridge, MA).
Schumpeter, J.A. 1942. Capitalism, Socialism and Democracy (London: Unwin).
Schettkat, R., Yocarini, L. 2006. The shift to services employment: a review of the literature. Structural
Change and Economic Dynamics, 17(2): 127-147.
Segal, G., Borgia, D., Schoenfeld, J. 2002. Using Social Cognitive Career Theory to Predict Self-Employment Goals. New England Journal of Entrepreneurship, 5(2): 47-56.
Shackle, G.L.S. 1970. Expectations, Enterprise and Profit (London: George Allen & Unwin).
