The influence of ADHD on entrepreneurial intentions and workplace difficulties

T

HE

I

NFLUENCE OF

ADHD

ON

E

NTREPRENEURIAL

I

NTENTIONS AND

WORKPLACE DIFFICULTIES

MASTER THESIS BY MARISE VAN STEIN

F

S

:

M

K

S

S

:

J

V

THIRD SUPERVISOR: CHI-MAO HSHIEH

Little research has been done to study if ADHD has an (in)direct effect on entrepreneurial intentions. Using data of 153 Dutch labor market participants, I test whether people with an ADHD diagnosis have higher entrepreneurial intentions and are more likely to become an entrepreneur. I also check if entrepreneurs with ADHD experience less difficulties at work now than when they were working for a boss, using a shortened ASRS scale and data of 24 Dutch entrepreneurs. I find suggestive evidence that ADHD has a positive influence on entrepreneurial intentions and that the difficulties at work increase.

TABLE OF CONTENTS

Verklaring eigen werk ... 1

1. Introduction ... 1

2. Literature ... 2

2.1 Studies on Students with ADHD ... 2

2.2 Studies about Adults with ADHD-Like Behavior ... 4

2.3 Hypotheses ... 6

3. Research Methodology ... 7

3.1 Methodology ... 7

3.2 Sampling ... 9

4. Descriptive Statistics ... 10

4.1 Descriptive Statistics Entrepreneurial Intentions... 12

4.2 Descriptive Statistics Change in Difficulties for Entrepreneurs ... 15

5. Results ... 16

5.1 Testing Hypothesis 1 ... 16

5.2 Testing Hypothesis 2 ... 21

6. Conclusion ... 24

Recommendations for future research ... 24

Limitations of the study ... 24

References ... 26

Appendices ... 29

Appendix 1: Theory on ADHD ... 29

Childhood disease or prevalence in adulthood? ... 30

Working impairments with ADHD ... 32

Appendix 2: The survey ... 33

Appendix 3: Adult ADHD Self-Reporting Scale (ASRS) ... 44

Appendix 4: Diagnosing ADHD ... 45

V

ERKLARING EIGEN WERK

Hierbij verklaar ik, Marise van Stein, dat ik deze scriptie zelf geschreven heb en dat ik de

volledige verantwoordelijkheid op me neem voor de inhoud ervan.

Ik bevestig dat de tekst en het werk dat in deze scriptie gepresenteerd wordt origineel is en dat

ik geen gebruik heb gemaakt van andere bronnen dan die welke in de tekst en in de referenties

worden genoemd.

De Faculteit Economie en Bedrijfskunde is alleen verantwoordelijk voor de begeleiding tot

het inleveren van de scriptie, niet voor de inhoud.

1. INTRODUCTION

The decision to become an entrepreneur instead of working for a boss has been subject for lots of studies (Shapero, 1982; Brockhaus, 1986; van Gelderen et al., 2008; Krueger, Reilly, & Carsrud, 2000). Governments want to understand the mechanism behind it because they see entrepreneurship as the drive behind the economy (European Commission, 2003; Rijksoverheid, 2014) and want to stimulate entrepreneurship (Amway Global Entrepreneurship Report, 2014). On the other hand, failing in entrepreneurship is a big problem too. Almost 90 percent of all start-ups fail (Forbes, 2015). When the start-ups do survive, most of them would have earned more if they would have worked for a boss (Hamilton, 2000). Boosting entrepreneurship still is important for governments (Rijksoverheid, 2014) and therefore it should be determined why people become entrepreneurs. Is it because of self-efficacy (Chen, Greene, & Crick, 1998), is it because of risk tolerance (Segal, Borgia, & Schoenfeld, 2005), or are there other personal characteristics like individualism (Tiessen, 1997) that play a crucial role? It is understandable that the combination of all the separate reasons leads to the answer to the question, but answering the question becomes easier when more of these separate reasons are known. This study tries to find if entrepreneurial intentions can be influenced by ADHD as well.

The question that might rise is: Why ADHD? Psychology and psychological disorder distinguishes people. ADHD is known impulsive- and risk-taking behavior but has many other impairments. The combination between ADHD and entrepreneurship has been subject for a few researches (Verheul, Block, Burmeister-Lamp, Thurik, Tiemeier, & Turturea, 2013; de Graaf, et al., 2008; Nicolaou, Scott, Cherkas, & Spector, 2008), but this study will replace the ADHD-like behavior for a ADHD diagnosis. In my opinion, people who go to a doctor with a problem experience more trouble in their daily lives than people who experience the difficulties only sometimes. Therefore, people who are getting themselves in to the medical system to know what is “wrong” with them, experience more difficulties and are a better representation for people with ADHD. Another important difference is that this study focusses on people who are participating in the labor market, instead of students. The focus of this study is on two subjects: it will not just give an answer to the question if entrepreneurial intentions are caused by ADHD, but it will also give a hint if and why people with ADHD experience less difficulties at work since they have become entrepreneurs. This result could be used to boost the work performance of employees with ADHD.

To test the hypotheses, I have conducted a survey among 153 Dutch labor market participants. The survey did not only ask for the diagnosis of ADHD, but also questioned the work related impairments that follow from having ADHD. After assessing the work related

impairments, participants who are entrepreneurs were asked to estimate the change in these impairments since they are working as an entrepreneur.

I find suggestive evidence that ADHD has a positive influence on entrepreneurial intentions, using two different scales to measure the entrepreneurial intentions. The first measure is the suggestive chance that people will become entrepreneurs within five years from now and the second measure is a question whether they want to become an entrepreneur or not, both with 129 participants. The second result of the study is that the difficulties at work increased since the participants started working as an entrepreneur, with the note that the group of entrepreneurs in the study is small: 24 participants.

First literature about ADHD and entrepreneurship will be discussed and hypothesis will be formed in Section 2. In Section 3 the survey is presented to check whether people with ADHD have some preference of becoming an entrepreneur. Section 4 discusses the descriptive statistics and Section 5 presents the results of the study. The conclusion of this study and recommendations for future research can be found in Section 6.

2. LITERATURE

In this part of the study the existing literature about ADHD related impairments and their effect on work performance in both self-employment or as an employee will be discussed. At the end of this paragraph, hypotheses will be formed. ADHD is described as “a persistent pattern of inattention and/or hyperactivity-impulsivity that interferes with development, has symptoms presenting in two or more settings (e.g. at home, school, or work), and negatively impacts directly on social, academic or occupational functioning”. The symptoms must be present before age 12 (ADHD Institute). In this study ADHD will be used as an abbreviation for both Attention Deficit Hyperactivity Disorder and Attention Deficit Disorder. More evidence of the existence of ADHD and prevalence into adulthood can be found in Appendix 1 of this study.

2.1

S

S

ADHD

Since adult ADHD is a recent development in the official DSM IV scale, there is more literature on ADHD considering children and students. One of the studies focusing on students is by Sifrin, Proctor and Prevatt (2010). The study examined the differences between students with and without ADHD. The study is conducted among students who work during their study. It has become more usual to be employed when in college; 50% of all full-time students enrolled in October 2004 were employed (U.S. Census Bureau, 2006 in Shifrin, Proctor, & Prevatt, 2010). Working during college has some positive sides, like the work experience and the money that can be used to pay for increased study costs, but also a negative side. The negative side might

include less time to study and an increased workload compared to non-working students. Both positive and negative effects of working during college have been subjects of research, there is no conclusion whether the negative effect is bigger or smaller than the positive effect. The working and non-working groups were compared, as well as within the ADHD group to check for gender related effects. The ADHD group was divided in the three sub-types of ADHD as well. It is expected in the light of this research that the group of students with ADHD experienced more difficulties with both their part-time jobs as their study, since combining two activities that both consume time and effort can be hard for people with ADHD.

The criteria used to diagnose the students was the same as described in the Appendix, originally founded by Barkley (2006). Learning disabilities were excluded to be the cause of the difficulties in studying by letting the participants do tests about their skills. This is a critical point, since some of the other studies in this paragraph consider ADHD to be a direct influence on the learning abilities of a person. The results of this study show that there is a significant difference between the work performance of college students with ADHD and without ADHD. The students with ADHD experience more difficulties at work, as they score themselves higher on the problems at work scale. The gender effect was not significant, so neither man or woman experienced more difficulties with ADHD than the other gender. No significant difference was found for gender on the severity of the symptoms. The students with ADHD were more frequently fired from their jobs and again this did not differ for gender. The last question of the work scale consisted of a self-assessment of the work performance. The students with ADHD consider their performances poorer than the students without ADHD (Biederman, Faraone, Spencer, Mick, Monuteaux, & Aleardi, 2006).

Symptom severity was not related to the number of times the participant was fired from a job. However, symptom severity was related to the self-assessment of the performance on the job. A negative correlation was reported between these two variables. There was a difference for the endorsed symptoms on the scale for the different types of ADHD, when looking at the combined ADHD type and the inattentive type. However, the researchers ask themselves a question: “Are the difficulties associated with adult ADHD related to the work difficulties within this population or is the poor self-rating a self-fulfilling prophecy leading to problems on the job?” The idea behind the lower self-assessment is reasonable: when you have just filled in a survey which underwrites the symptoms and the difficulties that you have to live with each day, you are in a more negative brainwave than the person who just thought: these symptoms do not apply to me at all. The researchers state that they think that the answer to their question is somewhere in the middle (Biederman, Faraone, Spencer, Mick, Monuteaux, & Aleardi, 2006). To be sure about the performance of the workers, there should be a combination of self-assessment and assessment by the boss of the worker. The authors also state that despite the difference in

performance and with more limitations than the average student, the students with ADHD were still on average less than one time fired. However, it was not noted how many jobs the students held, just the ones they are fired from. The percentage of fired would be a better estimation of the performance than the absolute number when there is nothing to relate the number to. This study relates to college students and their work related problems, however, when in college your most important task is graduating. Therefore, it might be different when work is the number one priority in your life.

2.2

S

A

ADHD-L

B

The former study discussed concludes that there is a difference between work performance of students with and without ADHD, now it is important to check if these differences are prevalent after college too. The World Mental Health Survey Initiative has studied the work performance of people with and without ADHD. The World Mental Health Survey can be conducted in 28 countries which “represent all the regions of the world with a total eventual sample size of over 154,000 people” (The World Mental Health Survey Initiative, 2005). For this study, 10 countries (Belgium, Colombia, France, Germany, Italy, Lebanon, Mexico, the Netherlands, Spain and the USA) were used to represent both developed and less developed countries (de Graaf, et al., 2008). In total there were 11,422 respondents surveyed across the 10 countries. The survey consisted of two parts, one general part about the work performance of the person and one part where it would be determined whether the person would qualify as having ADHD. However, the respondents from the USA, have previously been diagnosed. A strong association was found between the questions about ADHD in their survey and the clinical diagnosis (de Graaf, et al., 2008). The result of the study of WMO confirms that there is a significant difference in productivity of workers with and without ADHD. ADHD was associated with 22.1 annual days of excess lost role performance compared to otherwise similar respondents without ADHD (de Graaf, et al., 2008). The lost role performance measure is a combination of absenteeism and reduced quality/quantity. The researchers note at the end of the study that more than half the days out of role associated with ADHD are due to reduced quality/quantity of role performance rather than to days out of role. This is important from employer perspective, since they expect their employees to be working when they are on the job. The result of the study of WMO offers a different view, since more than half of the ADHD related lost role performance occurs on days in role (de Graaf, et al., 2008). It might be in the best interest of employers who suffer from this loss of productivity to screen their employees for ADHD and provide treatment. Another result of the study is that people are not treated for their adult ADHD in most countries. The authors state that: “Especially in the Netherlands and the USA, a much larger group of people would be diagnosed if professionals treating patients with

other emotional problems screened for comorbid ADHD, as sizable minorities of ADHD cases in both countries receive treatment for other emotional problems” (de Graaf, et al., 2008).

In 2013 a group of Dutch scientists investigated the link between ADHD and entrepreneurship. The biggest difference with this research is that their research concerned ADHD-like behavior, which was assessed by a scale of symptoms. The research of 2013 was conducted by Verheul et al.. The data that was used was provided by 10,000 students and looked at entrepreneurial intentions. The result of the 2013 study is that students who report higher levels of ADHD-like behavior are more likely than their peers to become a student-entrepreneur (Verheul, Block, Burmeister-Lamp, Thurik, Tiemeier, & Turturea, 2013). The linkage between ADHD and entrepreneurship is explained by the fact that people with ADHD have more need for independence and a higher risk tolerance than people without ADHD. The results are in line with the previous studies: they find that students who score high on the ADHD symptom scale are more likely than their peers to display entrepreneurial intentions and become student entrepreneurs (Verheul, Block, Burmeister-Lamp, Thurik, Tiemeier, & Turturea, 2013).

There is no absolute test for ADHD; however, studies focusing on genes have tried to find it. There are some genes that are associated with ADHD (Nicolaou, Shane, Adi, Mangino, & Harris, 2011). In their paper, Nicolaou et al. test if the genes that are associated with ADHD are associated with dopamine receptor genes and with the tendency to become an entrepreneur. The dopamine receptor genes are proved to be related to novelty seeking/sensation seeking (Benjamin et al. Nat Genet 12:81–84, 1996; Ebstein et al. Nat Genet 12:78–80, 1996; Noblett and Coccaro Curr Psychiatry Rep 7:73–80, 2005 in Nicolaou et al., 2011). Even though there might not be a gene that will determine whether someone will become an entrepreneur, it is proven that some characteristic factors enlarge the chance of becoming an entrepreneur (Zuckerman, 1979). These characteristics are however determined by genes. Therefore, becoming an entrepreneur might be inherited, as is ADHD. One of these characteristics is novelty seeking or sensation seeking, which is defined as “the need for varied, novel, and complex sensations and experiences and the willingness to take physical and social risks for the sake of such experience” (Zuckerman, 1979, p. 10). Novelty seeking/sensation seeking is more prevalent under entrepreneurs. People with a genetic predisposition to sensation seeking are more likely to be entrepreneurs, and shared genes influence the co-variation in the tendency to be an entrepreneur and sensation-seeking (Nicolaou, Scott, Cherkas, & Spector, 2008). People with a tendency to seek sensation are more likely to have dopamine receptor gene variations that increase the arousal necessary to achieve a given level of dopamine (Van Tol et al. 1991). This higher arousal threshold leads to greater odds of sensation-seeking activity, which might lead to entrepreneurship (Nicolaou, Shane, Adi, Mangino, & Harris, 2011). The link with ADHD here is that people with ADHD have lower levels of certain brain chemicals called neurotransmitters in

their brain. Dopamine is one of those. Risky behaviors can increase the dopamine levels (Van Tol et al. 1991).

In their study of 2011, Nicolaou et al. found that polymorphism in dopamine receptor DRD3 is positively related to becoming an entrepreneur. However, genetic variations in DRD3 are also related to susceptibility to gambling (Lobo et al. 2010), impulse control in Parkinson’s disease (Lee et al. 2009), and obsessive compulsive personality disorder (Light et al. 2006) (Nicolaou, Shane, Adi, Mangino, & Harris, 2011). No significance influence on ADHD associated genes and entrepreneurship was found in this study.

2.3

H

To answer the first research question “Is someone with ADHD more likely to become an entrepreneur?”, can be answered testing the hypotheses based on the literature. The first hypothesis states, based upon the studies of Verheul et al. (2013) and Nicolaou et al. (2011) that a person with an ADHD diagnosis is more likely to become an entrepreneur.

Hypothesis 1: A person with ADHD will be more likely to become an entrepreneur.

It is expected that people with ADHD are more likely to become entrepreneurs or that their entrepreneurial intentions are higher than their peers. Entrepreneurial intentions will be measured in two separate ways. The separation of people with and without ADHD will be done using the official diagnosis which people could fill out at the end of the survey. The difference between this study and the study of Verheul et al. (2013) is interesting since I expect that people with an official ADHD diagnosis are the ones with more difficulties, but a majority of adults with ADHD are untreated (Maucieri, 2014).

Hypothesis 2: For entrepreneurs with ADHD, the difficulties at work have a smaller influence on their work since they’ve become an entrepreneur.

I expect that entrepreneurs with ADHD experience less difficulties since they are working as an entrepreneur for various reasons. First, they can take breaks whenever they want to, switch between different tasks to keep their focus and change their environment in such a way that they are less distracted. When working for a boss, there is less freedom for the employee with ADHD to implement these changes to their working environment and therefore it is expected that the difficulties still arise, but are a lower influence on the job of the entrepreneur. These reasons are based upon the research of Nadeau (2005), which is described in Appendix 1. It discusses the working impairments of people with ADHD. We will also take a look at the reasons that were given for the change in difficulties.

3. RESEARCH METHODOLOGY

3.1 METHODOLOGY

To test the hypotheses formulated in the previous chapter, a group of people with and without ADHD need to be closer examined. The methodology that was chosen to obtain this data was by means of a survey. The survey consisted of four parts, see Appendix 3. Part one focused on the previous and current working history of the participant. The questionnaire was designed in such a way that entrepreneurs would see other questions than people who worked for a boss. The participants were asked to answer the questions keeping in mind the job that they value the most. These words were chosen to give room to the future expectations of the participant: even though he or she might earn more money from being employed, their entrepreneurial encounters might be more important to them since they expect it to become their main income in a few years from now.

In part two, the participants were asked to complete a scale representing difficulties which some people experience at work. These difficulties at work were taken from the official scale, the Adult ADHD Self-Reporting Scale (ASRS) which is used to diagnose adults. There are several scales used to diagnose ADHD in adults but since the official diagnosis is only valid if the symptoms were present before the 12th _{birthday, the ASRS scale was the best scale to use. The} ASRS scale has been studied by multiple researchers, (Kessler et al., 2005; Garland, 2006), to validate the scale. The authors note that the clinical interview used as the diagnostic standard has not been validated, even though it is commonly used (Kessler, et al., 2005). This is however no comment on the scale itself, but on the clinical instrument as a way to diagnose. Since it is widely used, it can give a good estimation of the patient’s behavior. Kessler et al. (2005) studied the usefulness of the scale with 154 participants in the USA, who already participated in the US National Comorbidity Survey. The conclusion of the study of Garland was: “The 6-item short form ASRS is effective for screening adults for ADHD, and is be a useful tool for identifying people with adult ADHD in community and clinical settings” (Garland, 2006, p. 38). The study of Kessler et al. (2005) concluded that the 6 –item ASRS screener is outperformed by the 18-question ASRS. In this study, seven statements were used to determine how the limitations of having ADHD affect people when they are working. The complete ASRS is added in Appendix 3 of this paper.

After the scoring of difficulties, participants were asked how they would think the impact of these difficulties would change if they would change from their current state of employment (entrepreneurship or employed), to entrepreneurship or being employed. If the participants answered that they would anticipate a significant change in impact, they could motivate why they would expect such a change. I further explore these questions to verify if entrepreneurship is a better option for people with ADHD and why. This might give new insights to keep improve

job performance of people with ADHD. Employers might be able to make minor adjustments in the jobs so that people with ADHD, and other people with difficulties in focusing, will be more productive.

In the third part of the questionnaire, the participants had to complete an entrepreneurial intentions scale. People had to estimate the chance of success for their own business venture. Entrepreneurs tend to overestimate themselves, their qualities and the chance that their business would be a success. This is underlined by statistics in the Netherlands in 2012, only 54% of all startups are still in business after five years (Spijkerman, 2012). However, there continue to be more new businesses each year: in 2013 there were 172,000 new businesses in the Netherlands and over 129 thousand businesses, a record number, stopped (Centraal Bureau van de Statistiek, 2014). These numbers prove that entrepreneurs are overoptimistic about their own capabilities to make their business a success; you will not start a company when you don’t think it will be a success.

Intentions are the best predictor for behavior (Ajzen, 1991) and entrepreneurial intentions are a deciding factor for performing entrepreneurial behavior (Kolvereid & Isaksen, 2006 in Moriano, Gorgievski, Laguna, Stephan, & Zarafshani, 2012). Therefore, participants were questioned if they had ever participated in schooling on entrepreneurship, non-entrepreneurs were asked if they wanted to become an entrepreneur, to estimate the chance of becoming one themselves within 5 years and after 5 years. These answers were used to test hypothesis 1. The entrepreneurs were asked to estimate the chance that they would switch towards becoming employed. The measures that were used to determine entrepreneurial intentions are based upon the Entrepreneurial Intentions Questionnaire (EIQ). The EIQ was found to be satisfying for investigating entrepreneurial intentions (Linan and Chen, 2006).

In the final part of the survey people were asked about background characteristics. The final question considered whether people were ever diagnosed by a certified physician as having ADD or ADHD. This way, the ADHD group and control group could be constructed according to having a diagnosis instead of by severity of the shortened symptoms scale. So that only participants who went through the medical system would qualify as having ADHD.

The language of the survey was kept as neutral as possible. The words ADD and ADHD where no sooner mentioned than in the very last question. This was done to reduce the experimenter demand effects. When people are aware from the start that the survey is about ADD or ADHD, they might answer the questions different than when they would just think that it is about experiencing difficulties at work, as was used in this survey. The importance of neutral language in surveys and questionnaires is underlined by Susanne Gerritsen in Schrijfgids voor Economen (2006) (Writing guide for economists), by Erik Feijen and Pepijn Trietsch in their book about graduating (2010) and in Arlene Fink’s book titled “How to conduct a survey”

published by 2013. Gender neutral language was used in this survey. The complete survey can be found in Appendix 2 of this study, the questions used to test the hypotheses can be found in Appendix 5.

The survey was published in Dutch, since the study was conducted in the Netherlands, or at least among Dutch people who are part of the Dutch labor market. The conditions to be able to participate in the study were to be above 18 years of age and to be a part of the labor market in the Netherlands. The Netherlands is a representative country for the western countries, because of good health care systems and schooling (The World Mental Health Survey Initiative, 2005). A good health system is important for the validation of the ADD/ADHD diagnosis.

3.2

S

To reach out to as many people as possible, several ways were used to promote the questionnaire. First, I send out an email to family and friends, about 30 persons, and asked them to participate in my study regarding functioning at work. Second, I posted the link to the survey on several newsgroups. Another way of distributing the survey was by handing out flyers at the hospital. Finally, I contacted the person behind ADHD Nederland if she would post the link on their Facebook (www.ADHDer.nl) and they did. Their Facebook has 1,825 likes. The text I used in the emails and the text that I posted online to get people to participate is added in Appendix 3. Some of the newsgroups deleted the link I posted since it was against the rules of their newsgroup. I took the biggest newsgroups in the Netherlands: one is focused on students, everyday life and gaming with 441,000 members, one is focused on woman in particular, no number of active members available, one is focused on ADHD and has 651 members and the last newsgroup that was used focuses on the equestrian lifestyle, with 45,000 active members. I cannot describe the non-respondents in my survey because of multiple reasons. First, the newsgroups have a large reach and there is no way of checking how many people saw the invitation to participate in the study. The second reason why I cannot describe the non-responders is that some of my friends and family forwarded the invitation to their friends and family as well, using Facebook, the electronic environment at their work or by forwarding the email. The third and final reason concerns the 50 paper invitations I handed out and placed at the hospital; I am not sure how many people saw the invitation and took the time and effort to read it, let alone participate in the study. Not being able to describe the non-responders to the questionnaire can be a limitation to the study, however, if the researcher is not able to select certain groups and distributes the survey as broad as possible, the sample can still be representative for the community. The participants spent on average 9.473 minutes on the survey, which means that the participants took enough time to answer the questions. There

were no surveys completed quicker than 5 minutes and the longest it took to complete the survey was 30 minutes.

The survey was the same for ADD/ADHD participants and the control group. Participants who participated in the survey when they found the link on websites for ADD/ADHD people might have had suspicions about the goal of the survey, but otherwise it would have been very difficult to find enough participants with ADHD for the study. In the announcement of the study, the word ADHD was not mentioned, except for the part that the student who made the survey has ADHD herself. This way it was thought that people would think that the reason to post the survey on the ADHD website was because of her own ADHD. The text that was used to announce the survey was as neutral as on the other websites, stating that the researcher needed as much participants in the study as possible and that the study would be about difficulties that some people experience at work and their goals in careers.

4. DESCRIPTIVE STATISTICS

Before testing of the two hypotheses, this section further describes the sample. In total 153 participants completed the survey. An important characteristic that is needed to test the hypotheses is whether participants were ever diagnosed as having ADHD by someone who is legally approved to do so. In my sample I find that there are 89 participants with a diagnosis and 64 without a diagnosis. The percentages are shown in Figure 1. Of course, this is not a representative proportion for the Dutch labor market , but the survey was distributed through some channels to specifically target those who have ADHD. When there are more participants who have ADHD, the sample will better reflect the ‘average’ person with ADHD. The next important division of the participants is on their current state of employment. There are more employees, 119, in the study than entrepreneurs, 24. The entrepreneurs represent in total 15% of the participants whereas the employees account for 71%. The students were directed to the end of the study after this question, since they have no or too little experience on the labor market. The Central Bureau of Statistics in the Netherlands estimates that there are a record-breaking 1.5 million businesses in the Netherlands (April 2015). The most of these businesses, 1.1 million are sole proprietorships, where often the owner is the only employee. The Central Bureau of Statistics also estimates that there are 12.7 million people of age between 15 and 75 years old and that two third of these people have a paid job. This estimation leads to roughly 8.5 million people who have a job. Then we take in to account the estimation of 1.1 million sole proprietorships, which leads roughly 1.5 million entrepreneurs in the country. The portion of entrepreneurs in the labor market will then be 17.6 percent, which differs only 2% from the

distribution in the sample. This difference might be explained by the fact that some people are both employed and owner of, for example, a small web shop as well.

The division between people who have an ADHD diagnosis and those who do not is made so we can check if there are any important differences in current state of employment. Table 1 shows the differences between the current state of employment for both groups. Here we can see that there are slightly more entrepreneurs in the sample who have ADHD than those who don’t: 20% vs 12%.

Diagnosis ADHD

Employed Entrepreneur In Between Jobs Total

Yes 43 67.19% 13 20.31% 8 12.5% 64 41.83%

No 76 85.39% 11 12.36% 2 2.25% 89 58.17%

Total 119 77.78% 24 15.69% 10 6.54% 153 100%

Table 1: Current state of employment tabulated with ADHD diagnoses.

Entrepreneur No entrepreneur Total

ADHD No ADHD ADHD No ADHD

Age 36.2311 _{45.727 35.169 38.179 37.562}2

Schooling 2.077 2.455 2.431 2.564 2.470

Male 0.6923 _{0.545 0.333 0.295 0.359}

N 13 11 51 78 153

Table 2: General characteristics of participants in the study divided into the subgroups

T-test were conducted to test if the subgroups within the study differed significantly from each other. It is important to test for these differences since you want to compare the groups on the characteristics that you test for, while all the other factors remain constant. This way, you get the result of having an ADHD diagnosis, not influenced by an difference in age. Three significant differences were found, for a significance level of 5%. These differences are marked with footnotes, which can be found at the bottom of this page. The first two significant differences can be found for Age: when we compare entrepreneurs with ADHD to entrepreneurs without ADHD, entrepreneurs without ADHD have a higher age. Overall, age is significantly lower for those with ADHD compared to those without ADHD. The last significant difference in this Table can be found for gender: there are significantly more male entrepreneurs than male employees in the sample. These differences can have an influence on the applicability of the results, if the differences in the sample are different from the differences that occur in the community. 11.5

1 _{Age is significantly higher for entrepreneurs without ADHD compared to entrepreneurs with ADHD.}

(P=0.024)

2 _{Age is significantly lower for people with ADHD diagnosis compared to people without ADHD diagnosis.}

(P=0.026)

percent of the woman participating in the labor market in the Netherlands were involved in entrepreneurship, for men, this percentage was 18.0 (CBS, 2011). For the other countries in the database, the percentage for men was higher than for women as well. This would indicate that this significant difference in the sample does not pose a threat to validity of the results of this study.

4.1

D

S

E

I

One of the indicators that is used in this study to determine the entrepreneurial intentions of the participants is the chance that they will become an entrepreneur/start their own business in the next five years.

Figure 1 shows the estimated chance of becoming an entrepreneur for both groups within the sample: for the group with ADHD diagnosis and for the group without the diagnosis, which is the control group.

Figure 1 shows the difference between the two subgroups. We can see that there is a higher fraction of participants in the ADHD group that estimate their chances higher than the control group. Up till an estimated chance of 30%, the control group has a higher percentage of participants, after this 30%, almost all the bars of the ADHD group are higher, except for the bar at 70%. This means that there is a higher fraction of people within the ADHD group that estimate their chances higher than within the control group. Figure 1 might give an indication that people with ADHD are more likely to become an entrepreneur than people without ADHD, but other checks need to be done before we can confirm this hypothesis.

A second measure for entrepreneurial intentions is used to check if the first measure is correct. The second measure is based on the question: “Do you want to be an entrepreneur?”. The answers that were possible and the answers that were given for both the ADHD- and control group are shown in Table 2.

0% 10% 20% 30% 40% 50% 60% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Density

Estimated Chance of Becoming an Entrepreneur

Estimated Chance to Becoming an Entrepreneur N=122

ADHD-Group Control group

Do you want to be an entrepreneur?

ADHD group Control group Total

I am: combined 4 7.84% 4 5.13% 8 6.20%

Yes, but I have no funds

8 14.81% 13 16.67% 21 16.28%

Yes, but I have no skills

9 17.65% 3 3.85% 12 9.30%

Yes, I will combine 7 13.73% 3 3.85% 10 7.75%

No 23 45.10% 55 70.51% 78 60.47%

Total 51 100% 78 100% 129 100%

Table 3: The answers to the second measure for entrepreneurial intentions: Do you want to become an entrepreneur?

Table 3 shows that in the ADHD group, 28 out of 51 would be positive towards becoming an entrepreneur. This leads to a percentage of 54.9%. In the control group, 23 out of 78 have entrepreneurial intentions (29.49%). This is a big difference and hints in the direction of the first hypothesis: that people with ADHD have higher entrepreneurial intentions than people without ADHD. For further analysis, I combined the first 4 options in the table mentioned above to 1 dummy that has the value 1 for entrepreneurial intentions. The 5th _{option in the table} corresponded to 0, meaning no entrepreneurial intentions. The distinction between 1 and 0 makes it a harsh difference: either you have entrepreneurial intentions or you do not. The first dependent variable in the regression model will be the subjective chance of starting your own business within 5 years from now, which can be nuanced a bit. With this measure there is no differentation between really wanting to start your own business or just thinking about it.

Other information that hints in the direction of hypothesis 1 can be seen in a correlation table, which is Table 4. This table can be found on the next page. In this table, it can be seen that there is a positive correlation between ADHD and two of the three dependent variables used to test hypothesis 1. Another result of the table is that education on entrepreneurship is positively correlated to both dependent variables.

Chance Entrepreneur

Entrepreneur-ial Intentions

ADHD Age Male Schooling Parents1 Parents2 Education on Entrepreneur-ship Chance Entrepreneur 1.000 Entrepreneur-ial Intentions 0.564*** [0.000] 1.000 ADHD 0.211** [0.017] 0.254*** [0.004] 1.000 Age -0.188** [0.033] -0.156* [0.079] -0.159** [0.054] 1.000 Male 0.021 [0.813] -0.028 [0.754] 0.083 [0.310] 0.052 [0.523] 1.000 Schooling 0.109 [0.219] -0.018 [0.839] -0.097 [0.235] -0.317*** [0.000] -0.047 [0.622] 1.000 Parents1 0.039 [0.661] -0.027 [0.757] -0.118 [0.147] -0.085 [0.299] -0.047 [0.567] 0.039 [0.637] 1.000 Parents2 -0.112 [0.210] -0.065 [0.426] -0.065 [0.426] 0.094 [0.248] 0.025 [0.762] -0.128 [0.115] -0.145* [0.074] 1.000 Education on Entrepreneur-ship 0.424*** [0.000] 0.268** [0.021] 0.126 [0.121] 0.105 [0.195] 0.105 [0.195] 0.064 [0.434] 0.262*** [0.001] -0.031 [0.708] 1.000

Table 4: Correlation table for all the variables that might influence the dependent variables used to test hypothesis 1. Chance of becoming an entrepreneur is used in model 1a+1b and Entrepreneurial Intentions is used in model 2a+2b. The numbers within the parenthesis are p values.

We can see that ADHD positively influences both measures of entrepreneurial intentions. Education on Entrepreneurship has a positive correlation with the dependent variables too.

Change in Difficulties

ADHD group Control group Total

Much Smaller 1 7.69% 1 9.09% 2 8.33%

Slightly Smaller 2 15.38% 1 9.09% 3 12.50%

About the same 4 30.77% 8 72.72% 12 50.00%

Slightly Bigger 2 15.38% 1 9.09% 3 12.50%

Much Bigger 4 30.77% 0 0.00% 4 16.67%

Total 13 100.00% 11 100.00% 24 100.00%

Table 5: A quick overview of the answers given to the chance in difficulties question since I am working as an entrepreneur.

4.2

D

S

C

D

E

Only the entrepreneurs who participated in the questionnaire are considered for hypothesis 2. The number of participants is therefore rather low, but it will be enough to take a quick look at the way that they perceive the different influence on the difficulties at work scale used to describe the ADHD symptoms; do these difficulties have a smaller or bigger influence on their jobs? First we take a look at Table 5 where all the answers to the question are tabulated and divided into the ADHD- and the control group. We can see that the entrepreneurs in the ADHD group are more divided about the change in difficulties than the entrepreneurs in the control group. In the control group, most of the entrepreneurs state that their difficulties are about the same, whilst in the ADHD group, there is a bigger difference in the answers, although the answers are not pointing in the same direction. Table 5 shows the differences in answers. The difficulty around examining this question is that the average of the answers will be around the neutral answer: about the same for both groups, but there seems to be a difference in distribution of the answers.

Within the ADHD group, there seem to be more entrepreneurs who experience more difficulties at work since they are working as an entrepreneur than less. This is the contrary of what was expected in hypothesis 2. For the control group, a majority experienced no change in difficulties. This result was expected: since it is expected that the participants without ADHD experience less of the symptoms described in the difficulties at work scale.

To check for any correlations between the dependent variable and the explaining variables, a correlation matrix is made. It is shown in Table 6. We can see that ADHD is weakly correlated to the dependent variable. We also see that ADHD is significantly correlated to Mean Score ADHD Scale, which confirms that the score on the scale is higher when the participant has an ADHD diagnosis. If the scale was a perfect measure for ADHD, the two would be perfectly correlated. Here, we can see that that is not the case, making the set-up of this research stronger. ADHD is also weakly negatively correlated with age, which is as expected too, since adult ADHD is only a recent development in the DSM-IV scale.

Increase/Decrease Difficulties for Entrepreneurs

Mean Score ADHD Scale

ADHD Male Age

Increase/Decrease Difficulties

1.000

Mean Score ADHD Scale 0.175 [0.413] 1.000 ADHD 0.290 [0.169] 0.667*** [0.000] 1.000 Male 0.116 [0.587] 0.010 [0.903] 0.083 [0.310] 1.000 Age 0.035 [0.873] -0.250** [0.002] -0.159** [0.050] 0.052 [0.523] 1.000

Table 6: A correlation table for all the variables used to test hypothesis 2. The dependent variable is increase/decrease in difficulties for entrepreneurs. The values within the parentheses are p values.

5. RESULTS

5.1

T

H

1

In the previous paragraph we have seen that after some naked eyeballing, there might be a connection between having the diagnosis of ADHD and becoming an entrepreneur. More rigorous tests are required to confirm this presumption.

First, a t-test is conducted to check if the different subgroups within the sample have different mean values for the dependent variable. This test was chosen because a t-test is a quick way to check for differences in the mean between subgroups. In Table 7, the dependent variable is the estimated chance of becoming an entrepreneur within five years from now. Table 7 shows that there is a significant difference between the means of the two subgroups. The mean of the

ADHD group is significantly higher than the mean of the control group. For the second measure of entrepreneurial intentions, the same test was conducted.

Mean Number of Observations Testing

Entire sample 18.450 [2.323] 129 ADHD group 25.294 [4.374] 51 Control group 13.974 [2.459] 78 Difference -11.320 [4.662] 129 T-value: -2.428 H1: Difference < 0 Pr (T < t) = 0.0083

Table 7: Means of the estimated chance of becoming an entrepreneur by subgroups within the sample. Two-sample t test with equal variances to test for a difference in the mean. The t value shows that the mean of the ADHD group is significantly, at the 1% level, larger than the mean of the control group.

The results of the T test for equal means are shown in Table 8. The result of this test is that the mean of the ADHD group is significantly higher than the mean of the control group, even at a significance level of 1%. After these tests for equal means, a Kolmogorov-Smirnov test is conducted to verify whether there is a significant difference between distributions. This test is shown in Table 9.

Both of the measures for entrepreneurial intentions are significantly different for both groups at a 5% significance level. This means that the distribution of the two groups is significantly different from each other. These results indicate that there might be higher entrepreneurial intentions among the ADHD group, as hypothesis 1 states.

Mean [Standard Error] Observations T-value Combined 0.395 [0.043] 129 ADHD group 0.549 [0.703] 51 Control group 0.295 [0.052] 78 Difference 0.254 [0.086] Pr (T<t) = 0.0018 -2.961

Table 8: The means of both groups compared to each other and tested for significant difference. The ADHD group has a significant larger mean than the control group.

Smaller group D P-value Corrected

Measure 1 Measure 2 Measure 1 Measure 2 Measure 1 Measure 2 Control group 0.237 0.254 0.032 0.019

ADHD group -0.006 0.000 0.998 1.000

Combined K-S 0.237 0.254 0.064 0.037 0.043 0.024 Table 9: Two-sample Kolmogorov-Smirnov test for equality of distribution functions shows us that there is a significant difference between the two distributions.

To check if this is correct, a more rigorous regression analysis is conducted. First a small model is made, with the estimated chance of becoming an entrepreneur as the dependent variable and ADHD, gender and age, as explaining variables.

These variables are considered exogenous because they are independent of the system that the person is in: ADHD is a condition of the brain, as has been described in the literature part and Appendix 2 of this study. ADHD will not chance due to external influences, such as gender and age. Schooling and substance abuse can however be influenced by the environment of a person and these variables are therefore considered endogenous. Endogenous variables that might cause entrepreneurial intentions are added in the second column of Table 10. These variables include the highest completed schooling, if the parents of the participants were once an entrepreneur (Parents1) or if they were once an entrepreneur but that they switched back to being employed (Parents2). Parents1 is expected to have a positive influence to the dependent variable and Parents 2 a negative influence. Another variable added is a dummy for schooling on entrepreneurship, which is expected to have a positive influence on the dependent variables. Schooling ranges between value 1 (High School) to 5 (Bachelor- or Master in Science).

Model 1a:

Chance of Becoming an Entrepreneur within 5 years = 𝛽1+ 𝛽𝐴𝐷𝐻𝐷∗ 𝐴𝐷𝐻𝐷 + 𝛽𝐴𝑔𝑒∗ 𝐴𝑔𝑒 + 𝛽𝑀𝑎𝑙𝑒∗ 𝑀𝑎𝑙𝑒 + 𝜀

Model 1b:

Chance of Becoming an Entrepreneur within 5 years = 𝛽1+ 𝛽𝐴𝐷𝐻𝐷∗ 𝐴𝐷𝐻𝐷 + 𝛽𝐴𝑔𝑒∗ 𝐴𝑔𝑒 + 𝛽𝑀𝑎𝑙𝑒∗ 𝑀𝑎𝑙𝑒 +

𝛽𝑆𝑐ℎ𝑜𝑜𝑙𝑖𝑛𝑔∗ 𝑆𝑐ℎ𝑜𝑜𝑙𝑖𝑛𝑔 + 𝛽𝑃𝑎𝑟𝑒𝑛𝑡𝑠1∗ 𝑃𝑎𝑟𝑒𝑛𝑡𝑠1 + 𝛽𝑃𝑎𝑟𝑒𝑛𝑡𝑠2∗ 𝑃𝑎𝑟𝑒𝑛𝑡𝑠2 + 𝛽𝐸𝑑𝑢𝑐𝑎𝑡𝑖𝑜𝑛𝑂𝑛𝐸𝑛𝑡𝑟𝑒𝑝𝑟𝑒𝑛𝑒𝑢𝑟𝑠ℎ𝑖𝑝+

𝛽𝐸𝑑𝑢𝑐𝑎𝑡𝑖𝑜𝑛𝑂𝑛𝐸𝑛𝑡𝑟𝑒𝑝𝑟𝑒𝑛𝑒𝑢𝑟𝑠ℎ𝑖𝑝∗𝐴𝐷𝐻𝐷∗ 𝐸𝑑𝑢𝑐𝑎𝑡𝑖𝑜𝑛𝑂𝑛𝐸𝑛𝑡𝑟𝑒𝑝𝑟𝑒𝑛𝑒𝑢𝑟𝑠ℎ𝑖𝑝 ∗ 𝐴𝐷𝐻𝐷 + 𝜀

The outcome of these regression analyses are shown in Table 10. For model 1a, we can see that the estimated coefficient for ADHD is positive and significant at a 5% level. This shows that in this model there is a significant positive influence of having the diagnosis for ADHD on entrepreneurial intentions. Age has a negative effect on the entrepreneurial intentions, which is as expected. Being male has a positive influence towards becoming an entrepreneur, which is

endorsed by the numbers of the CBS in 2011, as discussed in Section 4. However, this model has a low R2 _{which indicates that the explanatory power of the model is small. In column 2 of Table}

10, model 1b is presented where the significance of the ADHD coefficient is unfortunately lost. This means that ADHD is not a significant factor in increasing the chance of becoming an entrepreneur in the next five years. We can see that education on entrepreneurship significantly contributes to the chance of becoming an entrepreneur. Another expected sign of the parameter was that of Parents2, but the parameter is insignificant.

Dependent variable: (1a) Subjective probability of becoming an entrepreneur in 5 years (1b) Subjective probability of becoming an entrepreneur in 5 years (2a) Entrepreneurial Intentions (2b) Entrepreneurial Intentions ADHD 10.243** [4.708] 7.726 [5.275] 0.633*** [0.233] 0.649** [0.289] Age -0.381* [0.201] -0.295 [0.203] -0.015 [.010] -0.018 [0.012] Male 1.510 [4.994] -1.401 [4.688] -0.074 [0.251] -0.112 [0.264] Schooling 0.978 [2.438] -0.125 [0.136] Parents1 -5.096 [5.271] -0.308 [0.302] Parents2 -7.832 [8.949] -0.857 [0.539] Education on Entrepreneurship 28.169*** [7.141] 1.216*** [0.408] Education on Entrepreneurship x ADHD -4.535 [10.259] -0.756 [0.568] Constant 27.971** [8.192] 19.920 [12.255] 0.039 [0.411] 0.390 [0.688] N 128 128 128 128 R2 0.072 0.238 n.a. n.a.

Prob>F/ Prob> Chi2 _{0.025 0.000 0.014 0.003}

Table 10: Regression analysis on Subjective Probability of becoming an Entrepreneur in 5 years and on the entrepreneurial intentions question. If a coefficient is marked with one *, the coefficient is significant at the 10% level, if a coefficient is marked with two *, the coefficient is significant at the 5% level, if the coefficient is marked with 3*, the coefficient is significant at the 1% level.

In the last two columns of Table 10, two probit models were conducted with as dependent variable entrepreneurial intentions. The question that was used to build the measure can be found in Appendix 5. A probit model was chosen because of the binary value of the dependent variable: it can either take on the value of 1 or 0. In the first probit model, only the exogenous variables are considered.

Model 2a:

Entrepreneurial Intentions = 𝛽1+ βADHD∗ ADHD + βAge∗ Age + βMale∗ Male + 𝜀

Model 2b:

Entrepreneurial Intentions = 𝛽1+ βADHD∗ ADHD + βAge∗ Age + βMale∗ Male + βSchooling∗ Schooling +

βParents1∗ Parents1 + βParents2∗ Parents2 + βEducationOnEntrepreneurship+

βEducationOnEntrepreneurship∗ADHD∗ EducationOnEntrepreneurship ∗ ADHD + ε

In model 2a, the estimated parameter is significant positive at a level of significance of 1% and in model 1b, the parameter is significant at a level of significance of 5%. The P value for ADHD in model 2b is 0.025, which means that ADHD has a significant positive effect on entrepreneurial intentions. This would confirm the first hypothesis. The other estimated parameters are all but schooling pointing in the same direction as model 1a and 1b did. This way, we can determine that the measure for entrepreneurial intentions was a correct one. Another confirmation for this choice in dependent variable can be found in Table 4, a correlation matrix for the variables used to test hypothesis 1.

The Prob>F measure and Prob>Chi2 _{measure in the Table mention that there are indeed} parameters in the model that explain the value of the dependent variables. The R2 _{value is quite}

low for model 1a, but increases for model 1b. For models 2a and 2b, no R2 _{value is available. The}

improvement in the R2 _{measure means that the model has improved since the variables are}

added.

I also investigated if there was an interaction between ADHD and Schooling on Entrepreneurship. This way, it could be checked if having ADHD increased the chance in following schooling on entrepreneurship and that it contributes through a second channel towards entrepreneurial intentions. The interaction term is not significant in both models, but adding the interaction model does change the estimated coefficients of the other parameters. Adding the interaction term also improves the Prob>F and Prob>Chi2 _measure.

5.2

T

H

2

Before we can confirm the hypothesis, we need to test if there is a significant difference in the mean difference in difficulties of the two groups. A t-test was conducted. The average score of the answers on the scale ranging from 1 (much smaller) to 5 (much bigger) is shown in Table 11.

Mean Number of Observations Testing

ADHD Group 3.462 [0.369] 13 Control Group 2.818 [0.226] 11 Difference Control-ADHD -0.643 [0.453] 24 t value: -1.421*

Table 11: A t-test was conducted to check for equal means. For a significance level of 10%, the mean of the ADHD group was larger than the mean of the control group.

This result is surprising: it states that the difference that the entrepreneurs in the ADHD group experience a slight increase of difficulties on the job, while it was expected that there would be a decrease in difficulties. The increase is only small; on the scale, a 4 would mean a bigger influence and a score of 3 is about the same. The control group however does experience a slight decrease in difficulties at work.

After this t-test, a Kolmogorov Smirnov test was conducted to check if there was a significant difference in distribution between the ADHD group and the control group. There was no significant difference between the two groups, which would mean that there is no significant difference in the answers that were given between entrepreneurs with and without ADHD. This conclusion confirms the conclusion of the first test that was conducted: there is a small significant (at 10% level) difference but the means are still around the neutral choice: 3. This would indicate that entrepreneurs do not experience less difficulties since they are working as an entrepreneur, which would lead to the rejection of hypothesis 2. But before we can draw this conclusion, we have to compute an ordered probit model to estimate the parameters for the variables.

The dependent variable will be the change in difficulties (based on the ASRS-scale) and the participants are the entrepreneurs in the survey. Two regression models are specified in Table 12, the first model, 1a, is again only with the exogenous variables. An ordered probit model was chosen because of the ordinal character of the dependent variable: ranging from 1 to 5.

Model 1a:

In Model 1b, Schooling is added Other endogenous variables are not expected to have an influence on the dependent variable, so they are not added to the model. The regression is an ordered probit model, because the depending variable is a categorized variable ranging from 1 (never influenced by the difficulties) until 5 (always influenced by the difficulties).

Model 1b:

Difference in Difficulties = 𝛽1+ βADHD∗ ADHD + βAge∗ Age + βMale∗ Male + 𝛽𝑆𝑐ℎ𝑜𝑜𝑙𝑖𝑛𝑔∗ 𝑆𝑐ℎ𝑜𝑜𝑙𝑖𝑛𝑔 + 𝜀

In the first model, the estimated parameter of ADHD is not significant, neither are those of gender and age. In the second model, model 1b, the estimated parameter for ADHD is significantly at the 10% significance level. It is however positive and not as expected. Unfortunately, the models are both not strong. They both do not have a significant Chi2_value which means that the variables do not have joint significant explanatory power. This could be due to the low number of entrepreneurs in the sample. It would be interesting for future research to check if there is an improvement in working difficulties if someone with ADHD switches from being employed to becoming an entrepreneur. For now, we can only draw the conclusion that the data in this sample does not confirm the second hypothesis of this study. Entrepreneurs with ADHD do not experience less difficulties when they are working as an entrepreneur compared to when they were working for a boss.

Dependent variable: 1a Difference in Difficulties 1b Difference in Difficulties ADHD 0.786 [0.499] 0.886* [0.514] Age 0.022 [0.022] 0.026 [0.023] Male 0.312 [0.495] 0.391 [0.506] Schooling 0.193 [0.219] N 24 24 Prob > Chi2 _{0.401 0.446}

Table 12: This Table shows the outcome of two ordered probit regression models.

Entrepreneurs in the study were asked to explain why they experience a change in difficulties since they are working as an entrepreneur. Multiple explanations were given, not one of the entrepreneurs who experienced more difficulties than before entered their own reason in

the available open question field. The entrepreneurs could give more than one reason why they experience more difficulties, and most of them did. In Table 13, the reasons that were given for an increase in difficulties are displayed. In this Table, the total number of responses for that reason are given, but the responses are also divided into ADHD- and Control group. It was chosen to do so to check if there are certain characteristics to being an entrepreneur that brings more difficulties to people with ADHD and without ADHD. We can see that all of the reasons given apply more to people with ADHD. In Table 14, the same was done for the reasons why the difficulties decreased. Here we see a more even distribution between the ADHD- and Control group.

Reasons why the difficulties increase Number of times chosen in ADHD group

Number of times chosen in Control group

Total

I postpone difficult tasks or do not finish them correctly

5 1 6

I am not able to handle the freedom 2 0 2

I have difficulties in discipline to finish all the tasks

4 0 4

I am not able to handle the responsibilities that are required

3 0 3

Table 13: The reasons that were given by the entrepreneurs in the sample to the question why they experience more difficulties than when they were working for a boss.

Reasons why the difficulties decreased Number of times chosen by ADHD group

Number of times chosen by control group

Total

I have more freedom organizing tasks 2 1 3

I can take more breaks and switch between tasks

1 0 1

Tasks are more interesting now than when I was employed

2 2 4

The stakes are higher so my motivation has improved

1 0 1

I am enjoying my work more as an entrepreneur

1 1 2

I get less distracted because I can determine my own working environment

1 1 2

I can determine my own working hours 2 2 4

I have more experience now 1 1 2

Table 14: Reasons that were given by the entrepreneurs in the sample why they feel that the difficulties have decreased since they are working as an entrepreneur.

6. CONCLUSION

In this paper I have tested if having ADHD contributes significantly towards becoming an entrepreneur and if becoming an entrepreneur would help coping with the difficulties at work for people with ADHD. The literature that was studied provided enough evidence that ADHD can contribute positively towards entrepreneurial intentions. My contribution to the literature was to test if this is true for people who are in the labor market of the Netherlands and who are officially diagnosed instead of scoring high on a shortened symptom scale. I gathered data from 153 participants in an online study, published using various means: through newsgroups, e-mails and by handing out flyers at the hospital. The first hypothesis stated that people with ADHD diagnosis are more likely to become an entrepreneur. I have found supportive evidence for this hypothesis after using two measures for entrepreneurial intentions. ADHD is positively contributing towards entrepreneurial intentions. The second hypothesis stated that for entrepreneurs with an ADHD diagnosis, the difficulties at work have a smaller influence on their work since they have become an entrepreneur. This hypothesis was not confirmed by the data that I have gathered. There is only suggestive evidence that the difficulties at work increase after becoming an entrepreneur, instead of what the second hypothesis stated. Several reasons are given for the increase in difficulties, like postponing difficult activities or too much freedom. The number of entrepreneurs in the study might be the reason for the outcome that was reached, but further research needs to be done here before any harsh conclusions can be drawn.

R

For future research I would like to make a few recommendations that I would have done different looking back on what I have done in this study. Starting with my questionnaire. It might be interesting to check for other countries that are not as accommodative about ADHD as the Netherlands. Here it is a quite common and acknowledged disorder and therefore, society is a bit more accommodative than would be expected in other countries. It would be interesting to see if the results would differ over these countries, even though it would be harder to find participants with ADHD.

L

The respondents of the survey were for a majority females. However, this does not need to be a problem. For example, Nicolaou et al. (2011) describe in their paper that they used the TwinsUK registry as a source and in this sample, 83 percent of the participants were female (Nicolaou, Shane, Adi, Mangino, & Harris, 2011). Another limitation might be the use of a self-reporting,

single item scale as being an entrepreneur. However, this was also used in multiple studies, including the study of Nicolaou et al. (2011) and the validity of this measure was shown by Nicolaou et al. (2008) and by Kessler, et al. (2005).

Another point to consider is that I cannot describe the non-responders to the study. This might be a threat to external validity of the study. To be able to judge if there is a difference between the non-respondents and the respondents, you can test if there is a difference in answers between the early and late respondents, according to Radhakrishna and Doamekpor (2008). Unfortunately, I have not send a reminder to the people I have invited to participate by e-mail. This way, it cannot be tested if there is a difference in early and late respondents, since there is no way I can distinguish the early and late respondents.

We speak of biased estimates if the statistic is calculated in such a way that it is systematically different from the population parameter of interest. This can be caused by different faults in the study: the selection of participants (discussed in the sampling section of this paper), omitted variables, or by the researcher herself due to ‘observer bias’. Omitted variables can be influencing the results, but the specified regression models have a good Prob>F or Prob>Chi2 _{value, which means that there are indeed parameters in the model that explain the} value of the dependent variables. The observer bias, or observer-expectancy effect, was ruled out as much as possible by using neutral language in the questionnaire.

I do not think there is any reversed causality considering the research questions of the study. ADHD is an anomaly in the brain and is therefore exogenous. Reversed causality would be the case if entrepreneurial intentions would cause ADHD. If the research question focused on the relationship between entrepreneurial intentions and education on entrepreneurship, there would be reversed causality. For the second research question: Do the difficulties on the job decrease for entrepreneurs with ADHD? There is also no reversed causality for the reason given before.

If I could go back in time and change the questionnaire, I would add a question for the entrepreneurs, how they had experienced the difficulties before they were entrepreneurs. This way, I would have had a better way to compare the current situation with the past situation. The formulation of the question might have been better for the participants too. I would have also sought for more participants who are entrepreneurs, so I could build upon more data for the second hypothesis.

REFERENCES

ADHD Institute. (n.d.). DSM-5. Retrieved October 23, 2014, from ADHD Institute.

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 179-211.

Biederman, J. M., Faraone, S. V., Spencer, T. J., Mick, E. S., Monuteaux, M. C., & Aleardi, M. B. (2006, April). Functional Impairments in Adults With Self-Reports of Diagnosed ADHD: A Controlled Study of 1001 Adults in the Community. Journal of Clinical Psychiatry(67), 524-540.

Brockhaus, R. a. (1986). The Psychology of an Entrepreneur. The Art and Science of Entrepreneurship, 25-48.

Centers for Disease Control and Prevention. (2013, November 13). Attention-Deficit/Hyperactivity Disorder (ADHD). Retrieved February 10, 2014, from CDC: http://www.cdc.gov/ncbddd/adhd/data.html

Centraal Bureau van de Statistiek. (2014, April 4). Meer bedrijven gestart dan gestopt. Retrieved October 27, 2014, from Centraal Bureau van de Statistiek: http://www.cbs.nl/nl-NL/menu/themas/bedrijven/publicaties/artikelen/archief/2014/2014-4053-wm.htm Chen, C., Greene, P., & Crick, A. (1998). Does entrepreneurial self-efficacy distinguish

entrepreneurs from managers? Journal of Business Venturing, 295-316.

de Graaf, R., Kessler, R., Fayyad, J., ten Have, M., Alonso, J., Angermeyer, M., et al. (2008, May 27). The prevalence and effects of adult attention-deficit/hyperactivity disorder (ADHD) on the performance of workers: results from the WHO World Mental Health Survey Initiative. Occupational and Environmental Medicin, 65, 835-842.

Dipeolu, A. O. (2011). College Students With ADHD: Prescriptive Concepts for Best Practices in Career Development. Journal of Career Development, 408-427.

Garland, M. R. (2006). The short form adult attention deficit/hyperactivity self-report scale is a useful diagnostic measure. Evidence Based Mental Health, 38.

Gezondheidsraad. (2000). Gezondheidsraad: Diagnostiek en behandeling van ADHD. Den Haag: Gezondheidsraad.

Hamilton, B. (2000, June). Does Entrepreneurship Pay? An Empirical Analysis of the Returns on Entrepreneurship. The Journal of Political Economy, 604-631.

Harrison, A. G., Alexander, S. J., & Armstrong, I. T. (2013). Higher Reported Levels of Depression, Stress, and Anxiety Are Associated With Increased Endorsement of ADHD Symptoms by Postsecondary Students. Canadian Journal of School Psychology, 243-260.