The effect of extraversion on the relation between gender and risk taking behaviour

(1)

The effect of extraversion on the relation

between gender and risk taking behaviour

Bachelor Thesis by Erik de Greef

Student Number: 11011890

Study track: Business Administration

Faculty of Economics and Business

Supervised by: M. Stienstra

(2)

Statement of Originality

This document is written by Student Erik de Greef who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Abstract

This research investigates how gender and the degree of extraversion, both

separately and jointly, influence the degree of risk taking an individual. To do this, 49 surveys were conducted under UvA third years’ business administration students representing a pool of potential future managers. The separate effect of extraversion has been proven. The separate effect of gender showed that for this sample, females are more risk taking than male. A model with both extraversion and gender was proven to be significant and adding more explanatory power than the previous mentioned effects separately. Also, the moderation effect of extraversion on the relation between gender and risk taking was proven to be significant. This expresses itself in such a way that the degree of extraversion of a male, has more effect on the degree of risk taking a male has, than the degree of extraversion does for a female. The effect of extraversion can thereby, lower the effect of gender, and makes extravert males roughly as risk taking as females are (regardless of their level of extraversion).

(3)

1. Introduction

Managing risk is essential in a business context. This goes beyond financial risk, managers have to manage employee risk, data risk and financial risk too (Lipman, 2017). That this can go fatally wrong, can be seen in the fall of the Lehman Brothers (Williams, 2010). Lehman Brothers was an investment bank, based in New York. Because of decreasing trust in financial institutions in 2008, and a low financial reserve available (high leverage), Lehman went bankrupt. This fall in 2008 marked the beginning of the financial crisis worldwide (Rutter, 2012). The degree of leverage the Lehman Brothers used enhanced the profits the company made. However, increasing the leverage increases the risk too (Baker, 1973). It could be possible that the company was able to avoid bankruptcy by choosing a manager that was less risk taking.

The example of the Lehman brother shows how important it is for companies to select executives who can manage risks well. The organization is a reflection of the behaviour of its top managers. Managing and taking risk is behaviour that can have an impact on the success of a company. It is therefore important to choose managers who fit well with the culture of the company and the expectation of stakeholders. This culture is created by the employees working for the company, and not by the company itself (Hambrick & Mason, 1984). Especially top managers play a big role in creating an organizational culture. This makes choosing a manager fitting well with the (desired amount of risk taking of the) organization even more vital. This research focusses on the degree of risk taking in a professional context.

Certain psychographic and demographic factors have an effect on the degree of risk aversion a person has. Of those, gender is the most researched (researched by for example Byrnes, Miller, & Schafer, 1999; Charness & Gneezy, 2012; Finucane, Slovic, Mertz, Flynn, & Satterfield, 2000). The generally used theory hypothesizes that men are more risk taking than women. However, there are more factors influencing the degree of risk taking. Factors researched include race, world view, age and experience in taking risks (Finucane et al., 2000). However, adding personality factors to the model could add more explanatory power (Byrnes, et al., 1999). It is indicated that besides demographic factors such as gender, age, and cultural factors, personal factors could have an effect on risk taking behaviour.

A commonly used tool to assess personality are the ‘Big five personality characteristics’ (Barrick & Mount, 1991). This scale consists of five factors assessing

(5)

personality: Openness, Neuroticism, Agreeableness, Extraversion and Conscien-tiousness. For this research, there was chosen to study the effect of extraversion on risk taking. This was done because, according to the PEN-theory of Eysenck (1967), extravert people are more likely to take part in arousal-seeking behaviour. However, what this arousal-seeking behaviour entails, was never researched in the light of risk taking. It is plausible that risk taking is such arousal-seeking behaviour. This would imply that extravert people take more risks.

Since both extraversion and gender are likely to influence the amount of risk taking behaviour an individual shows, an interaction effect between those two factors is likely. This research therefore also tries to prove a moderation of extraversion on the relation between gender and risk taking behaviour. This would indicate that males react stronger to being extravert in terms of risk taking behaviour than females do. This stems from the hypothesis that males are more risk taking than female, and extraversion strengthening this effect (Charness & Gneezy, 2012).

Research into the effect of extraversion on risk taking behaviour, nor the possible interaction with gender is not done yet according to the author. Researching this could add to the understanding of risk taking behaviour, which is beneficial for the scientific knowledge about risk taking, as well as for companies trying to select the manager with the appropiate amount of risk taking a company needs.

To research the effect of extraversion on risk taking behaviour, the following research question was formed: How do gender and extraversion, both seperatly and jointly, influence the degree of risk taking behaviour? To answer this question, first the theory known about risk aversion and extraversion will be discussed in detail. After this, the methodology of this research will be discussed. Thirdly, the results will be presented and lastly a discussion on these results, including a conclusion and directions for further research.

(6)

2. Theory

2.1 Risk Taking

Different theories about risk taking.

There are several definitions for risk taking available. In this research, the (broad) definition of Byrnes (et al., 1999) will be used. It is stated that risk taking is the implementation of options that could possibly lead to negative outcomes. Three different categories of theories arround risk taking are mentioned: (1) describing differences between people who are risk taking and who don’t regurarly take risk. From this theory stems the commonly used hypotheses that men are usually more risk taking than women. (2) The second group of theories states that certain situations promote risk taking, while others demote this. (3) The third category of theories explains why not everyone takes the same risk in a certain situation. According to these theories, the reason people take risks is that they believe they will have succes, and that value that certain succes (Byrnes et al., 1999). This research mainly focusses on the first theory, explaining that there are differences between people in risk taking.

Risk taking and its impact on the return of an organisation.

The amount of risk an organisation (or its managers) takes, can have an impact on the profit an organisation makes. It is a commonly used investment rule in finance: taking more risk leads to a greater variance in profits, but enhances the average profits. This leads to a risk-return trade-off (Berk & deMarzo, 2014). The existence of this trade-off has already been proved in a stock market context (eg. Bali, 2006). However, since companies make investments too, their risk taking behaviour has also an effect on the return a company gets (Berk & deMarzo, 2014).

The optimal point in this trade-off is dependent on the owners (/stockholders) of the company. However, choosing a manager who acts in line with the prefered choices of the owners is vital for a well functioning company, since the choices of a company are dependent on the choices of its top managers (Hambrick & Mason, 1984). To ensure a good fit with the managers, a look into the amount of risk a manager is willing to take is important.

Some research has already been done into risk taking in a professional context (eg. Weber, Blais, & Betz, 2002). Hereby, it’s possible to translate above

(7)

theories to a business context. One possible limitation of this approach is that in a business context, managers often don’t make decisions about their own money, and that other incentives play a role in decision making than only making profit (Eisenhardt, 1989). This is referred to as the principal-agent theory. This can result in alternative behaviour concerning risks. This misalignment of incentives can partly be fixed by providing bonuses to managers dependent on the success of the company (Eisenhardt, 1989). This research, however, won’t cover this phenomenon. The focus of this research is to discover personal causes of taking risks for managers.

Risk taking and the decision making process.

As mentioned, risk is about taking choices. This choices can lead to negative outcomes. However, before a choice is made, managers (unconsciously) follow a decision making process. How they go through this process affects the choices they ultimately make (Harrison, 1995). In the following section, each part of the decisionmaking process will be shortly discussed.

The first step is identifying a problem (1). This is the stage in which a manager acknowledges there is a need for a choice. Next, the manager collects information (2) about the problem, from which alternatives are identified (3). After that, the manager weighs the evidence (4), and based on that information a choice between

alternatives (5) is made.Thereafter the action is implemented (6) and lastly the result is evaluated (7) (Robbins & Judge, 2009).

While this process seems to be rational in nature, it rarely is. One of Byrnes (et al., 1999) researches indicated that individuals are prone to take more risk, even when it is obviously a bad decision. This shows that not all choices made are rational, which indicates that personality can affect it. This is mostly done in the third, fourth and fifth stage. In this stages, individuals make different assesments of risk and will make different choices because of that (Robbins & Judge, 2009). Also, weighing evidence is a relative process. This supports why it is important to find a manager who fits the organisation culture (Hambrick & Mason, 1984).

(8)

2.2 Gender in relation to risk taking

In the past, research was conducted into the link between gender and risk taking. This was mainly based in the field of psychology. Most of this research indicates that men are more risk taking than women (Byrnes et al., 1999; Finucane et al., 2000).

From a socio-biological perspective, especially young men have a ‘reproductive competition’, which entails that they have to proof that they are the strongest of the herd. To proof this, males are more of a show-off type, and taking risks more often. The basis of this lies in the evolution theory (Wilson & Daly, 1985). There also is a difference in genotype between men and women, which could have an effect on risk taking (Eaves & Eysenck, 1975).

Another theory is that males are more likely to be sensation seeking than females are. This makes males more risk taking than female, since this is a form of sensation seeking behaviour. It is indicated that males sometimes take risks, even when its clear that it is a bad idea to do so (Byrnes et al., 1999). This can be explained that the behaviour of taking risks in itself is prefered by male (just for sensation seeking purposes), and not so much the outcome the taken risk. This effect is strengthened by the belief that risk taking behaviour masculine tendency that in most cultures is highly valued (Kelling, Zirkes, & Myerowitz, 1976).

The commonly used theory states that males are more prone to risk taking behaviour than females. This can be applied to a business context too. Given that decisions in a business context follow the personality and choices in the private life of a manager, this risk taking would happen in a professional context too. This will lead to male managers who take more risk than female managers. This effect is shown in figure 1.

𝐻1: 𝑀𝑎𝑙𝑒 𝑚𝑎𝑛𝑎𝑔𝑒𝑟𝑠 𝑎𝑟𝑒 𝑚𝑜𝑟𝑒 𝑟𝑖𝑠𝑘 𝑡𝑎𝑘𝑖𝑛𝑔 𝑡ℎ𝑎𝑛 𝑓𝑒𝑚𝑎𝑙𝑒 𝑚𝑎𝑛𝑎𝑔𝑒𝑟𝑠.

Figure 1 – Conceptual specification of H1.

2.3 Extraversion in relation to risk taking.

Extraversion

One of the ‘Big-Five’ dimensions of personality is extraversion (eg. Barrick & Mount, 1991). This dimension measures whether people are sociable and dominant. This

(9)

has an impact on how people communicate with each other, but also in the behaviour an individual shows (Costa & McCrae, 1992).

Extraversion in a business context.

Being extravert is helpful for people having managerial jobs (Barrick & Mount, 1991). This is because they are more likely to seek out interaction opportunities with others, and because they are often more incentive and reward minded than introvert managers. Besides that, extraverts are more likely to take a challenge and want to preform complex and varied tasks in their work, whereas introverts are more likely to take the already discovered path (Costa & McCrae, 1992).

On a team level, extraversion can impact the quality of the interaction between team members (O'Boyle, Humphrey, M., Hawver, & Story, 2011). The generally accepted theory is that teams with more extravert members have a high team performance (Barry & Stewart, 1997). However, since extravert people tend to be more dominant, conflicts in those teams also happen more often.

Extraversion in relation to risk taking.

According to the Eysencks PEN-theory, people who score high on extraversion have a low arousal level. This leads to them searching for outside incentives, to heighten their arousal (Eysenck, 1967). One of these incentives could be risk taking (Byrnes et al., 1999). That extraversion leads to sensation seeking behaviour was supported by a research conducted later into the relation between personality and risk taking behaviour (Nicholson, Soane, Fenton-O'Creevy, & Willman, 2005). In this research, sensation seeking was one of the proposed reasons of extravert people taking risks.

The second reason people take risk is that extravert people are more keen on achieving goals and reducing loss (Nicholson et al., 2005). The (irrational) consequence of this is that more risk is borne, to avoid a likely loss. This effect is supported by the prospect theory (Kahneman & Tversky, 2013). This theory states that because of a ‘certainty effect’, things likely to happen are weighed heavier than things unlikely to happen. To illustrate this, consider the following example: According to the theory, people are more likely to prefer to have 10% chance of loosing 100 euro, rather than to loose 10 euro with 90% certainty. Statistically, the

(10)

first option looses more money, but is preferred because of the chance of a loss incurring being lower (Kahneman & Tversky, 2013).

The third reason extravert people are willing to take more risk, is because they are more likely to be risk adaptors (Nicholson et al., 2005). This implies that they are more used to taking risks in a particular domain. This is caused by the need for sensation, as mentioned in the first reason. However, since more risks are being taken already, extraverts get more used to taking risks, resulting in more comfort taking risks, and ultimately in taking even more risks (Nicholson et al., 2005).

The knowledge on what causes individuals to take risks is not complete yet (Byrnes et al., 1999). This is partly explainable by the fact that personality shows variance between people which is not explainable. Expanding the model predicting risk taking behaviour could add explanatory power. It is probable that extraversion has an effect on risk taking. The attempt of extraverts to heighten their arousal level hereby results in taking more risks (Eysenck, 1967). This research adds extraversion as variable in predicting risk taking behaviour.

Extraversion and risk taking in a business context.

Making the model predicting risk taking behaviour explain more of the variance could proof useful for companies, especially during their selection procedure. To ensure a good fit with the company, selecting individuals with a personality fitting the culture of the company is important (Hambrick & Mason, 1984). This research proves that personality, in this case extraversion, has an effect on the risk a manager is willing to take.

Having extraversion can have an effect on the behaviour a person has. According to the PEN-theory, extravert people tend to heighten their arousal level by searching outside incentives (Eysenck, 1967). Risk taking has the characteristics of being such an incentive (Byrnes et al., 1999). From this follows that managers that are extravert, also tend to take more risk. This effect is shown in figure 2.

𝐻2: 𝑇ℎ𝑒 𝑑𝑒𝑔𝑟𝑒𝑒 𝑜𝑓 𝑒𝑥𝑡𝑟𝑎𝑣𝑒𝑟𝑠𝑖𝑜𝑛 𝑖𝑠 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒𝑙𝑦 𝑟𝑒𝑙𝑎𝑡𝑒𝑑 𝑡𝑜 𝑡ℎ𝑒 𝑑𝑒𝑔𝑟𝑒𝑒 𝑜𝑓 𝑟𝑖𝑠𝑘 𝑡𝑎𝑘𝑖𝑛𝑔.

(11)

2.4 Gender and extraversion in relation to risk taking.

Because both gender and extraversion are both separately hypothesized to explain the risk taking behaviour a person shows, it would be likely that the both are

complementary in explaining risk taking behaviour. This means that a model that contain the effects of both gender and extraversion has more explanatory power than the models of only gender or extraversion on risk taking have (The models explained under H1 and H2). This is shown in figure 3, explaining H3.

𝐻3: 𝐺𝑒𝑛𝑑𝑒𝑟 𝑎𝑛𝑑 𝐸𝑥𝑡𝑟𝑎𝑣𝑒𝑟𝑠𝑖𝑜𝑛 𝑎𝑟𝑒 𝑐𝑜𝑚𝑝𝑙𝑒𝑚𝑒𝑛𝑡𝑎𝑟𝑦 𝑖𝑛 𝑒𝑥𝑝𝑙𝑎𝑖𝑛𝑖𝑛𝑔 𝑟𝑖𝑠𝑘 𝑡𝑎𝑘𝑖𝑛𝑔 𝑏𝑒ℎ𝑎𝑣𝑖𝑜𝑢𝑟.

Figure 3 – Conceptual specification of H3.

Lastly, both extraversion and gender are hypothesized to have an effect on the risk taking behaviour of an individual. The hypothesis hereby is that male individuals react more strongly to extraversion than females do in terms of risk taking behaviour. This implies that the difference in risk taking behaviour between low and high extravert males is higher than it is between low and high extravert females. An explanation for this could be that risk is partly dependent on the urge of sensation seeking behaviour (Byrnes et al., 1999).

In most cultures, males are more likely to have the urge of sensation seeking behaviour, and thus are more likely to take risks. The degree of extraversion is hypothesized to have a positive effect on sensation seeking behaviour (and thus risk taking) too. Since both males and extravert people are more likely to show risk taking behaviour, being both extravert and male could strengthen this effect. This would indicate that males, because of their naturally higher sensation seeking level, would react more strongly to being extravert, than females do. This effect is shown in figure 4.

(12)

𝐻4: 𝐸𝑥𝑡𝑟𝑎𝑣𝑒𝑟𝑠𝑖𝑜𝑛 𝑚𝑜𝑑𝑒𝑟𝑎𝑡𝑒𝑠 𝑡ℎ𝑒 𝑒𝑓𝑓𝑒𝑐𝑡 𝑜𝑓 𝑔𝑒𝑛𝑑𝑒𝑟 𝑜𝑛 𝑟𝑖𝑠𝑘 𝑡𝑎𝑘𝑖𝑛𝑔 𝑏𝑒ℎ𝑎𝑣𝑖𝑜𝑢𝑟 𝑖𝑛 𝑠𝑢𝑐ℎ 𝑎 𝑤𝑎𝑦 𝑡ℎ𝑎𝑡 𝑚𝑎𝑙𝑒𝑠 𝑟𝑒𝑎𝑐𝑡 𝑠𝑡𝑟𝑜𝑛𝑔𝑒𝑟 𝑡𝑜 𝑒𝑥𝑡𝑟𝑎𝑣𝑒𝑟𝑠𝑖𝑜𝑛 𝑖𝑛 𝑡𝑒𝑟𝑚𝑠 𝑜𝑓 𝑟𝑖𝑠𝑘 𝑡𝑎𝑘𝑖𝑛𝑔 𝑏𝑒ℎ𝑎𝑣𝑖𝑜𝑢𝑟

𝑡ℎ𝑎𝑛 𝑓𝑒𝑚𝑎𝑙𝑒𝑠.

(13)

3. Methodology

3.1 Design and sample

To test the aforementioned hypotheses, a survey was conducted under third years University of Amsterdam (UvA) business administration bachelor students. This sample is selected because they have a high potential to be a future manager (in line with the method of Zeff, Fremgen, & Martinez (1994)). The external validity will be low, due to the population selection. This is done however, to ensure the research is do-able in the below given timespan.

To ensure that the sample is representative for the population (third years UVA business administration bachelor students), a good amount of surveys should be conducted. Since no accurate information could be retrieved about the number of third years’ bachelor students at the UvA, this number will be estimated being 801. Based on this, and a confidence level of 95%, at least 67 survey should be conducted. However, after conducting the surveys, only 49 were collected, since no more students responded. This has some implications for the external validity of this study, which will be discussed below.

3.2 Data collection procedure

The survey was built and distributed via Qualtrics. Access to this online online survey tool was provided by the UvA, making it possible to design and distribute online surveys. There is chosen for an online data collection method in the hope to easily reach students who don’t go to the UvA campus regularly. Before the survey was distributed, it was pretested by the supervisor of this thesis and a few potential targets of the survey. After this, the survey was adjusted to the provided feedback, and participants were given the option to leave feedback about the survey. Despite this option for respondents, no critical feedback was given, making it necessary to make adjustments to the survey. Participants were granted anonymity when doing the survey.

The survey was distributed via social media, and was put on Blackboard too. However, when after a week not enough responses were collected, there was

1_{This number is based on a obligatory second years course in which only business administration}

(14)

decided to also conduct the survey at the UvA using a tablet. In total, the stage of data collection took two weeks, after which the data analysis was started.

3.3 Variable measurement

Risk taking

To measure the amount of risk a participant is willing to take, the financial measures for risk aversion mentioned in Weber, Blais, & Betz (2002) are used. Those measurements were asked on a five-point Likert scale based on likelihood. An example of a question used is: ‘Lending a friend an amount of money equivalent to one month’s income’, rated on the scale from ‘not at all likely (1)’ to ‘extremely likely (5)’. This questions are included in appendix A of this proposal. The Cronbachs alpha for the questions in the study of Weber (et al., 2002), was 0.69, which is acceptable. The degree of risk taking was calculated using the calculation given below.

𝑅𝑖𝑠𝑘 𝑇𝑎𝑘𝑖𝑛𝑔 = 𝐴𝑣𝑔(𝑄1 + 𝑄2 + 𝑄3 + 𝑄4 + 𝑄5 + 𝑄6 + 𝑄7 + 𝑄8)

Extraversion

For the measurement of the variable extraversion some of the questions used in Eaves & Eysenck (1975) were used. Besides that, the question for extraversion of Woods & Hampson (2005) was used. In total, 12 questions asking how extravert a respondent is were included in the survey. Those questions, asking for example whether people feel shy when meeting people, were measured on a five-point Likert scale ranging from ‘strongly disagree (1)’ to ‘strongly agree (2)’. The Cronbachs alpha for the questions in the study of Eaves & Eysenck (1975) was 0.76, which is acceptable. After inversing back inversed questions, the answers on the questions led to a degree of extraversion via the following calculation:

𝐸𝑥𝑡𝑟𝑎𝑣𝑒𝑟𝑠𝑖𝑜𝑛 = 𝐴𝑣𝑔(𝑄1 + 𝑄2 + 𝑄3 + ⋯ + 𝑄10 + 𝑄11 + 𝑄12)

Gender

One question is added to the survey asking for the gender of the participant. To make an analysis using SPSS of the results possible, female participants were coded as 0, and male participants as 1.

(15)

Control variables

Three control variables were added in the analysis. The first control variable is whether the participant is Dutch or international. The study Business Administration has an English section too, which makes it possible for international students to study at the UvA. The second control variable added was how long the students have been studying till this point. Because of study delays it is possible that students are studying for more than three years while they are in the third year of the study. It could for be possible that Dutch people behave different than international students. The same goes for the years the student has been studying. This could say something about the lifestyle a student has, which in turn could have an effect on risk taking behaviour. The last used control variable is the age of the respondents. While this age only varied between 20 and 26, it could be possible that the behaviour of older respondents is different than that of younger respondents. That age has an effect risk taking behaviour, was proven by Vroom & Pahl (1971). They indicated that older individuals are less likely to take risks.

3.4 Reliability and validity

To check for reliability, the questions for both risk taking and extraversion were tested using a Cronbachs alpha analysis. The reliability of the questions for extraversion proofed good (α=0,853). On top of that, deleting questions wouldn’t increase the α by much (>0,05), so the choice was made not to delete any questions in the analysis. For the risk taking questions, in this study the reliability was acceptable (α=0,742). On top of that, deleting questions wouldn’t increase the α by much (>0,05), so the choice was made not to delete any questions in the analysis (See appendix B for reliability analysis).

To ensure internal validity, the above mentioned control variables age, years of studying and nationality were used. This ensures that the effect of extraversion (IV) and gender (IV) on risk taking (DV) are tested, taking into account differences in other characteristics tested by the control variables.

The external validity of the sample was improved by giving every member of the population the chance to participate in the survey. By posting the survey to the blackboard page of the course of bachelor thesis, every third years’ bachelor student Business administration (the population) had access to the survey. This is a form of

(16)

random sampling. In the second phase of the research, in which the researcher approached possible participants directly on campus, self selection sampling was used. This was done to make the sample size bigger, reducing the margin of error. However, since only 49 individuals participated in the survey, the external validity of this study is relatively low.

3.5 Data analysis

The results of the survey were analysed via the statistical software SPSS. From the theory follow a few hypotheses. The first three hypotheses were tested using a ordinary least square regression analysis. The fourth hypothesis was tested using the software written by Hayes (2018 version). This is a software tool within SPSS, written for testing moderation.

(17)

4. Results

4.1 Descriptive statistics

Sample characteristics

Before a regression was done, several descriptive statistics (mean, standard deviations and maxima) were tested for the used variables was tested. Those are shown in table 1.

The survey was completed by 49 participants of which 55% were female (as shown in figure 6). The participants are all third years’ bachelor business administration students at the UvA (After deleting all invalid data, and incomplete responses). However, some students are already studying more than 3 years: 11 respondents 4th year (22,4%), 2 respondents 5th year (4,1%). The age of the participants ranges from 20 to 26 (Mean = 22,24, SD = 1,234, as shown in table 1). Of the participants, 8 (16,3%) have a non-Dutch nationality (as shown in figure 5).

The degree of extraversion, calculated by the formula given above, varied between 1.58 and 4.25 (on the scale of 1-5). The mean degree of extraversion was 3.0 (SD=0,60). The degree of risk taking varied between 1.25 and 3.00 with a mean of 2,1 (SD=0,51). 16% 84%

Figure 5: Distribution of

nationality

International Dutch 45% 55%

Figure 6: Distribution of

gender

Male Female

(18)

Both standard deviations are relatively low, indicating that the spread between the data is relatively small (ie. The amount of extreme scores is low). This is supported by looking at the questionnaires, where the mode in answers to questions mostly was 2 and 4.

Testing for Normality

To preform a regression analysis, the variable risk taking should be distributed normally. This can be tested by assessing the skewness of the data. For a distribution to be normally distributed, skewness has to be as close to zero as possible. For the variable risk taking, the skewness didn’t present normality issues (Skew=0.093, SE=0.340). Also, the kurtosis, which also should be as close to zero as possible didn’t indicate for serious non-normality (Kurtosis=-1.038, SE=0.668). This result is supported by the Shapiro-Wilk test, which indicates that there is no proof for the distribution to be non-normal (p=0.09). The distribution of the variable risk taking is shown below in figure 7.

(19)

4.2 Correlation analysis

To assess whether the variables are associated with each other, a bivariate correlation test was performed. The correlation in this research is measured using Pearson’s correlation. This value varies between -1 and 1, with extreme values indicating a correlation.

As table 2 shows, only two correlations are significant. The first significant correlation is between extraversion and risk taking. This correlation is 0.336. This implies that participants that are more extravert people are more risk taking. This is in line with the expectation and the second hypothesis stating that the degree of extraversion is positively related to the degree of risk taking. The second positive correlation is between years of studying and age. This result can be expected too, since every student starts studying at roughly the same age. This implies that when students are studying for more years, usually their age is higher than students studying for a shorter amount of time.

4.3 Regression analysis

A hierarchical regression analysis in SPSS was done using the variables risk taking, extraversion, gender and the control variables. In the table below, the results of the analysis are stated, after which they will be discussed per hypothesis.

(20)

H1:Male managers are more risk taking than female managers.

After a first insight into the data was discovered, a linear regression analysis for testing H1 was done. This was done using the variables risk taking (DV) and gender (IV) and the before mentioned control variables. The regression analysis however, indicated that, for this sample, females are more risk taking than males are. The regression analysis showed that, for this sample, females score on average 0.250 higher on the scale of risk taking. This result is not in line with the stated theory that male managers are more risk taking than female managers.

H2: The degree of extraversion is positively related to the degree of risk taking.

To test the second hypothesis whether extraversion has a positive effect on the risk taking, a second regression analysis was done using the variables extraversion (IV) and risk taking (DV). This led to a significant outcome supporting H2 (p=0.01; B=0.309). This indicates that with an increase of 1 point on the scale of extraversion, the risk taking increases by 0.309 points on the scale of risk taking. The R-squared of the model is 0.143, which indicates that 14.3% of the variability in risk taking is explained by the model.

(21)

H3: Gender and Extraversion are complementary in explaining risk taking behaviour.

To test whether a model with both gender and extraversion adds explanatory power to previous models, another regression analysis was done. This regression analysis used the control variables, as well as the variables extraversion (IV), gender (IV) and risk taking (DV). This led to a significant effect for both extraversion (B=0.314,

P=0.01) and gender (B=-0.250, P=0.09). This indicates that with an increase of 1 on the scale of extraversion, the degree of risk taking increases by 0.314 on the scale of risk taking. This also indicates that male are less risk taking than female by 0.250 on the scale of risk taking in this model. Lastly, the model as a whole proved significant (p=0.08), and the R-squared of the model is 0.199 which indicates that a proportion of 19.9% of the variance in risk taking was explained by the model.

4.4 Testing for moderation

H4:Extraversion moderates the effect of gender on risk taking behaviour in such a way that males react stronger to being extravert in terms of risk taking behaviour than females.

To test whether there is an interaction between gender and extraversion, the tool for testing moderation written by Hayes (2018 version) was used. From this test, the model, as shown in the theory, proofed to be significant (p=0.01). The explanatory power of the model predicting risk taking behaviour, R-squared, was 22%. The effect of gender was proven to be significant (p=0.04; B=-1.4930). This indicates that females score 1.4930 higher on the scale of risk taking (This is against H1, which will be discussed in the discussion). To test the direction of the effect and test whether H4 is supported, a graph of the interaction was constructed, which is shown in figure 8. As the figure shows, extravert males are more risk taking than introvert males, while such an effect of extraversion is not as strong for females.

(22)

Figure 8 – The effect of gender and extraversion on risk taking

4.4 Testing hypotheses

In table 4 the tested hypotheses are stated again, showing that H1 is not supported by the data, and H2-H4 are supported by the data.

Hypothesis Support

H1:Male managers are more risk taking than female managers. Not supported H2:The degree of extraversion is positively related to the degree

of risk taking.

Supported H3:Gender and Extraversion are complementary in explaining

risk taking behaviour.

Supported H4:Extraversion moderates the effect of gender on risk taking

behaviour in such a way that males react stronger to

extraversion in terms of risk taking behaviour than females.

Supported

Table 4 – Results of the analysis. 0 0,5 1 1,5 2 2,5

Low extraversion Average extraversion High extraversion

De gr ee o f r is k ta ki ng Degree of extraversion

The effect of gender and extraversion on risk taking.

Males Females

(23)

5. Discussion and conclusion

5.1 Discussion

This research was done to discover whether extraversion has an influence on the relation between gender and the degree of risk taking a manager has. The definition of risk taking hereby used was that of Byrnes (et al., 1999). Byrnes states that taking risks is implementing options that could lead to negative outcomes.

To study this, four hypotheses were formed. The first hypothesis which indicated that male managers are more risk taking than female managers was not supported by the data. This is not in line with the commonly used theory that states that male are more risk taking than female. This can be partly explained by a ‘catch up behaviour’ of female, as indicated by Byrnes (et al., 1999). This theory states that males are confronted with certain risks at a younger age (before their 20’s), making them more precocious in their twenties. Females, who aren’t confronted with risks as much before their twenties, will react to risks differently in their twenties, because of a lack of experience in dealing with risks. This leads to females making less precocious decisions, which makes them more risk taking than males at this specific age.

The second hypothesis, indicating that extraversion is positively related with taking risks was supported by the data. This expected result can be explained by the PEN theory (Eysenck, 1967). This theory states that extravert people take part in arousal seeking activities, of which risk taking is one. Another possible explanation could be that extravert people usually think less before they act (Eaves & Eysenck, 1975). This makes the chance of a rational weighting of the perceived risk lower, resulting in a higher taken risk.

The third hypothesis, indicating that gender and extraversion are

complementary in explaining risk taking behaviour was supported as well. The model including control variables, gender and extraversion was proven to be significant in predicting risk taking behaviour (p=0.08). However, because the low R-squared (0.2), indicating that only 20% of the variance is explained by this model, proves that the model is not complete with only these variables.

The fourth hypothesis, indicating that male react more in terms of taking risks to being extravert than female was supported by the data too. After data analysis, it became visible that the degree of risk taking for female hardly differs between low and high extraversion, while for males, high extravert male are more risk taking (and

(24)

indicates that at this age, the difference in the degree of risk taking behaviour between males and females is smaller, as the degree of extraversion of the male heightens.

5.2 Scholarly implications

By adding extraversion to the model of gender explaining risk taking behaviour, this research shows that explanatory power of the model increases. Besides, by also investigating the possibility of a moderation of extraversion on the relation between gender and risk taking behaviour, this research has proven that the effect for

extraversion is only significant for males. This research added one of the commonly used personality factors to the model, including an explanation why extraversion could have an effect on risk taking behaviour, a field not researched before.

5.3 HRM implications

It is important to make sure that a individual fits with the culture of the company (Hambrick & Mason, 1984). A company can use the knowledge obtained in this study to select the individuals with the appropriate amount of risk taking. This means that if a risk taking (young) individual is needed, the company is better of hiring an extravert male or a female. On the contrary, if a risk averse individual is needed, the company could hire an introvert young male, who is on average less risk taking than extravert males and females.

5.3 Limitations and further research

The most apparent limitation of this study is that the sample size is lower than necessary to achieve a low margin of error. Because of the sample size being only 49, the margin of error for the results is 8.77%. Another limitation of this study is that the population (third years UVA business administration students) is rather small for explaining the behaviour of future managers. This comes with the risk of results not being generalizable to a broader population (being future managers). This could be an explanation for the lack of support for the first hypothesis. Also, the small range of age of the participants could skew the results. Future research could investigate a wider range of age, to account for this.

(25)

This study proves useful in exploring more antecedents for risk taking than the ones already known by adding extraversion to the model. Future research could add more relevant variables to get a more complete knowledge about what causes risk taking, and make more reliable predictions about the risks a manager takes. It would, in line with this research, make sense to pick one of the other ‘Big-Five personality dimensions’. This way, a more complete view on personality its effect on risk taking behaviour can be constructed.

5.4 Conclusion

The research question for this thesis was as follows: How do gender and extraversion, both separately and jointly, influence the degree of risk taking

behaviour? Through a quantitative analysis, the positive association of the degree of extraversion and the degree of risk taking, and a complementary effect of gender and extraversion were proven. Also a moderating effect of extraversion on the relation between gender and risk taking behaviour was proven to be significant.

This implies that both gender and extraversion have an effect on risk taking. From this follows that extravert males are more risk taking than introvert males, and that females are more risk taking than males (which was not in line with the

hypothesis). The difference between males and females hereby was minimal for males with high extraversion.

(26)

References

Baker, S. H. (1973). Risk, Leverage and Profitability: An Industry Analysis. The Review of Economics and Statistics, 55(4), 503-507.

Bali, T. G. (2006). Is there a risk–return trade-off? Evidence from high-frequency data. Journal of Applied Econometrics, 21(8), 1169-1198.

Barrick, M. R. & Mount, M. K. (1991). The big five personality dimensions and job performance: a meta-analysis. Personnel psychology, 44(1), 1-26.

Barry, B. & Stewart, G. L. (1997). Composition, process, and performance in self-managed groups: The role of personality. Journal of Applied psychology, 82(1), 62-72.

Berk, J. & deMarzo, P. (2014). Corporate finance. London: Pearson Education. Byrnes, J. P., Miller, D. C. & Schafer, W. D. (1999). Gender differences in risk taking:

A meta-analysis. Psychological bulletin, 125(3), 367-383.

Charness, G. & Gneezy, U. (2012). Strong evidence for gender differences in risk taking. Journal of Economic Behavior & Organization, 83(1), 50-58.

Costa, P. T. & McCrae, R. R. (1992). Normal personality assessment in clinical practice: The NEO Personality Inventory. Psychological assessment, 4(1), 5-13. Eaves, L. & Eysenck, H. (1975). The nature of extraversion: a genetical analysis.

Journal of personality and social psychology, 32(1), 102-112.

Eisenhardt, K. M. (1989). Agency theory: An assessment and review. Academy of

management review, 14(1), 57-74.

Eysenck, H. (1967). The biological basis of personality. Springfield, IL: Thomas Publishing.

Finucane, M. L., Slovic, P., Mertz, C. K., Flynn, J. & Satterfield, T. A. (2000). Gender, race, and perceived risk: The 'white male' effect. Health, risk & society, 2(2), 159-172.

Hambrick, D. C. & Mason, P. A. (1984). Upper echelons: The organization as a reflection of its top managers. Academy of management review, 9(2), 193-206. Harrison, E. F. (1995). The managerial decision-making process. Boston: Houghton

Mifflin.

Kahneman, D. & Tversky, A. (2013). Prospect theory: An analysis of decision under risk. Handbook of the fundamentals of financial decision making, 1(1), 99-127.

(27)

Lipman, V. (2017, October 17). All Good Managers Are Risk Managers. Retrieved May 13, 2017, from Forbes:

https://www.forbes.com/sites/victorlipman/2017/10/17/all-good-managers-are-risk-managers/#8e8108242764

Nicholson, N., Soane, E., Fenton-O'Creevy, M. & Willman, P. (2005). Personality and domain-specific risk taking. Journal of Risk Research, 8(2), 157-176.

O'Boyle, E. H., Humphrey, R. H., M., Pollack, J., Hawver, T. H. & Story, P. A. (2011). The relation between emotional intelligence and job performance: A meta-analysis. Journal of Organizational Behavior, 32(5), 788-818.

Robbins, S. P. & Judge, T. A. (2009). Organiational behaviour. London: Pearson Education.

Rutter, J. (2012). Kicking the Can: Understanding the finacial crisis of 2008. Doctoral

dissertation. University of Denver.

Weber, E. U., Blais, A. R. & Betz, N. E. (2002). A domain-specific risk-attitude scale: Measuring risk perceptions and risk behaviors. Journal of behavioral decision

making, 15(4), 263-290.

Williams, M. T. (2010). Uncontrolled risk: the lessons of Lehman Brothers and how

systemic risk can still bring down the world financial system. New York:

McGraw-Hill.

Wilson, M. & Daly, M. (1985). Competitiveness, risk taking, and violence: The young male syndrome. Ethology and sociobiology, 6(1), 59-73.

Woods, S. A. & Hampson, S. E. (2005). Measuring the Big Five with single items using a bipolar response scale. European Journal of Personality, 19(5), 373-390.

Zeff, L. E., Fremgen, B. & Martinez, J. C. (1994). Implications of gender differences for managers. Psychological reports, 74(3), 755-763.

(28)

Appendix

A - Questions used in Survey Eaves & Eysenck (1975)

Questions in italics are inversed, answers on a 1-5 likert scale.

o Do you suddenly feel shy when you want to talk to an attractive stranger? o Generally, do you prefer reading to meeting people?

o Do you prefer to have few but special friends? o Do other people think of you as being very lively? o Are you mostly quiet when you are with other people?

o If there is something you want to know about, would you rather look it up in a

book than talk to someone about it?

o Would you be unhappy if you could not see lots of people most of the time? o Are you usually carefree?

o Do you stop and think things over before doing anything? o When people shout at you, do you shout back?

o Do you like doing things in which you have to act quickly? Woods & Hampson (2005)

Generally, I come across as:

1: Someone who is talkative, outgoing is comfortable, but could be noisy and attention seeking.

5: Someone who is a reserved, private person, doesn’t like to draw attention to themselves and can be shy around strangers.

Weber, Blais, & Betz (2002)

Likelyhood on the scale from 1-5

o Betting a day’s income at the horse races. o Co-signing a new car loan for a friend.

o Investing 10% of your annual income in a very speculative stock.

o Investing 10% of your annual income in government bonds (treasury bills). o Investing in a business that has a good chance of failing.

o Lending a friend an amount of money equivalent to one month’s income. o Spending money impulsively without thinking about the consequences. o Taking a job where you get paid exclusively on a commission basis.

(29)

B - Cronbachs Alpha analysis

Extraversion – Original alpha in this research: 0.853

Question Cronbach’s alpha if deleted

People think of me as being very lively. _0.833 I would be unhappy if i couldn’t see lot’s of people

most of the time.

0.855

I am usually carefree. _0.857

When people shout at me, i shout back. _0.853 I like to do things in which i have to act quickly. _0.861

I’m Talkative and outgoing. _0.812

I usually don't stop and think things over before I do anything.

0.837

I don't suddenly feel shy when I want to talk to an attractive stranger.

0.834

I prefer meeting people to reading _0.836

I don't prefer having few, but special friends as opposed to many friends.

0.844

I am mostly not quiet when I'm with other people _0.829 If there is something I want to know about, I prefer

talking to other people about it, rather than looking it up for myself.

0.835

Risk taking – Original alpha in this research: 0.742

Question Cronbach’s alpha if deleted

Betting a day’s income on sports. _0.704

Co-signing a new car loan for a friend. _0.765 Investing 10% of your annual income in a very

speculative stock.

0.670

Investing 10% of your annual income in government bonds (treasury bills).

0.697

Investing in a business that has a good chance of failing, but when it succeeds grants higher returns.

0.662

Lending a friend an amount of money equivalent to one month’s income.

0.732

Spending money impulsively without thinking about the consequences.

0.765

(30)

The effect of extraversion on the relation between gender and risk taking behaviour