Psychopathic Bankers and Rationality.
Name: Mai Berger
Student number: 10897267 Supervisor: Dr. Audrey Hu
Second examiner: Dr. Joep Sonnemans Date: 17 December 2015
Master of Science, Economics Faculteit Economie en Bedrijfskunde
Abstract
3
1. Introduction
4
2. Related Literature
6
2.1 Measuring psychopaths 6 2.2 Successful psychopaths 7 2.3 Unsuccessful psychopaths 9 2.4 Rationality 102.5 Contribution of this study 11
3. Methodology
13
3.1 Sample 13 3.2 Procedure 13 3.3 Measures 13 3.4 Hypotheses 144. Results
16
4.1 Coding 16 4.2 Summary statistics 184.3 First set of results 20
4.4 Second set of results 21
4.5 Limitations 24
5. Discussion and Conclusion
25
6. Appendix
28
Abstract
50 voluntary banker participants answered 43 questions separated into three parts. The first part was the Levenson’s Self-Report Psychopathy test (LSRP), used to determine the
subjects’ level of psychopathy. The second part was a collection of Allais’ paradox questions regarding money, health, and leisure. This part was used to determine the subjects’ level of rationality. The third part was a collection of control questions about the subjects’ background. Using this data, this thesis aimed to find out whether psychopath bankers are more rational and whether bounded rationality changes as psychopathy levels change. Several regressions were run using ANOVA, where rationality was the dependent variable and the control variables and psychopathy level were the fixed factors and
covariates. There were no significant effects of psychopathy or the control variables on rationality. The implications of this research, its merits, and its weaknesses, are all discussed.
1. Introduction
In economics, there is an assumption of the perfect, rational economic man upon whom our economic models are based. He is the homo economicus who acts in an entirely
self-interested fashion, only concerned with his own payoffs. He is also entirely fictional with numerous studies finding that people, even of different cultures, deviate from the canonical assumption (Henrich et al., 2001). After all, the homo economicus is there for the purpose of facilitating economic studies, deriving models, and verifying theories. However, sometimes these assumptions of the economic man can be relaxed in order to see whether there are similarities between our species and this fictional one. For instance, there are many overlaps in the characteristics of the economic man and a psychopath. A psychopath is also self-interested, though they may be willing to get a higher payoff at the expense of another person, whereas the economic man may care about the effect of their choices on others. But, psychopathy runs on a continuous scale. Those who are most similar to homo
economicus would be the more ruthless, able to adapt, “successful” psychopaths. Another psychopath on the psychopathy dimension is the incarcerated kind. These “unsuccessful” psychopaths are less adaptable and their rebellion against social norms, which is normal for a psychopath, has been exposed leading to their incarceration. Whether a successful psychopath is more or less extreme on the scale than an unsuccessful psychopath is unknown, but what is known is that they are two different kinds of psychopaths, both exhibiting different behavior.
There have been investigations into the behavior of psychopaths, but little is known about the behavior of socially successful psychopaths. This is because it is easier to have a pool of incarcerated, known psychopaths, than it is to find “successful” psychopaths. The search for successful psychopaths have led researchers to believe that they occupy some of the more high-positioned professions in the market (Mullins-Sweatt et al., 2010). Among these high-positioned professions are attorneys, physicians, businessmen, and aristocrats (Hall and Benning, 2006). Thus, this thesis will be looking at the behavior of successful
psychopaths, particularly of the business kind. To be more precise, this paper aims to study
whether psychopath bankers are more rational and whether bounded rationality will change as psychopathic traits differ.
This is a relevant topic for research as successful psychopaths are very under investigated. Past research has dictated that more research regarding the socially successful
psychopaths is badly needed (Mullins-Sweatt et al., 2010). Furthermore, this thesis will illuminate any possible differences or similarities between the successful and unsuccessful psychopaths. Additionally, this thesis will be looking at people with different levels of
psychopathy. This will allow conclusions to be drawn about any relationship or correlation that may exist between psychopathy and rationality. Lastly, there is a possibility that additional information can be deducted regarding the similarities between homo economicus and psychopaths.
To answer the research question, the thesis will first review the existing literature, then explain the methodology, followed by an analysis of the results and ending with the formation of a conclusion.
2. Related Literature
The most important related literature for this research pertains to the topic of successful psychopaths. However, since little is known about successful psychopaths, this study investigates into a matter where the work done previously is scarce. Nevertheless, what is known will be discussed, as well as the existing literature about unsuccessful psychopaths. Additional relevant literature is about the rationality of humans and what is commonly found in modern day society. The related literature will serve as a justification behind the choices made in this research, in addition to elucidating any possible holes in research. Moreover, the existing studies will be critically evaluated in order to ascertain what the dos and don’ts are in this sort of research.
Psychopaths have certain traits differentiating them from the general population. These include: a lack of conscience and incapability of empathy, guilt, or loyalty to anyone except themselves (Babiak and Hare, 2006). Additionally, Babiak and Hare (2006) state
psychopaths are “antisocial and violent for reasons that often seemed random and
senseless”, “egocentric in the extreme”, and “unable to experience deep human emotions, especially love and compassion”. According to Babiak and Hare (2006), psychopaths are the perfect invisible predator. Another source says psychopaths are unreliable, untruthful, unable to follow any life plan, and of average intelligence (Cleckley, 1988). Some more important traits to take into consideration are their impulsivity (Levenson et al., 1995) and lack of irrationality and anxiety (Cleckley, 1988). Psychopathy is known as one of the three dark triad of personality traits, along with narcissism and Machiavellianism (Paulhus and Williams, 2002). According to Paulhus and Williams (2002), psychopaths have low levels of neuroticism and conscientiousness, and exhibit self-enhancement. Some of these
characteristics seem to conflict with one another. A possible explanation for this is a scale of psychopathy, a sort of dimension where one can be more or less psychopathic than
another. This scale of psychopathy may also be associated with the successfulness of a psychopath.
2.1 Measuring psychopaths
One’s scale of psychopathy can be measured through different means. The most popular and commonly used in the literature is Hare’s revised psychopathy checklist (PCL-R). It is clear that the PCL-R is a robust measure across cultures and also across genders (Hare et
al., 2000), making it a viable option. However, in total, the PCL-R takes two to two and a
half hours to complete, making it less desirable. Additionally, it requires extensive training and is a costly assessment to purchase. Other alternatives include, but are not restricted to, descriptions in terms of the five-factor model (FFM), the updated FFM called the five factor
form, the psychopathic personality inventory (PPI), and the Levenson’s self-report
psychopathy scale (LSRP). A commonly used alternative to PCL-R in the literature, is the LSRP. The LSRP is a 26-item self-report (Levenson et al., 1995), normally taking about 5 minutes, making it far more feasible for this survey. The survey uses a five point Likert scale. The LSRP measures the same as the PCL-R but in a more compact manner and in a non-institutionalized population. As this study focuses on a working class sample group, it is optimal to use a test designed for the non-institutionalized. The LSRP also distinguishes people into different levels of primary and secondary psychopathy, which is a nice distinction to have. It has previously been found that individuals in the primary psychopathy category are “more resilient to stress, emotionally stable, and psychologically healthy than their secondary counterparts” (Skeem et al., 2011). This distinction could provide additional information for the analysis.
2.2 Successful psychopaths
A successful psychopath is defined as someone who “fits the criteria of a psychopath, having certain fundamental traits (e.g., callousness), but largely succeed in their exploitation” (Mullins-Sweatt et al., 2010). It is speculated that these individuals are able to succeed in their exploitations because of these fundamental traits they possess. Their fearlessness and willingness to take risks or make ruthless decisions against other individuals may be the driving factor bringing them to the top of the hierarchy. Successful psychopaths are difficult to find or identify due to their successful nature. Regardless, researchers have tried, some succeeding in finding successful psychopaths and others did not. Mullins-Sweatt et al. (2010) narrowed their search of the successful psychopath to the sample group
psychologists interested in law, attorneys, and clinical psychology professors. This sample group was asked to answer questionnaires about psychopathic traits, all the while keeping a specific person they had encountered in their field of work in mind. For the psychologists interested in law and attorneys, this could be a client. For the professors, this could be a student. The modal person chosen by the psychologists interested in law and professors was a colleague. The most discussed person by the attorneys were clients, which could spark debate since a successful psychopath would not be caught in the position where they would need the assistance of a lawyer. This paper found that the successful psychopath was distinguishable from the unsuccessful psychopath, where the former has a higher conscientiousness. This is an important finding to come across as it verifies the suspicions of the psychopathy scale. A previous study has also distinguished the successful and unsuccessful psychopaths from one another through a physical measure (Ishikawa et al., 2001). However, a flaw in the Mullins-Sweatt study is that the researchers are never in direct contact with the said psychopaths, meaning they never confirm for themselves whether
these people are in fact successful psychopaths - the accuracy of the identification of these successful psychopaths could be questioned as the data is not based on self reporting but rather the subjects reporting on others. The sample group may have perceived the
supposed psychopaths in a way differently to how they are in reality, or the supposed psychopaths may have acted differently around the interviewed people. Alternatively, the sample may have just not known the people well enough to describe them in sufficient detail.
Along with the confirmation that successful psychopaths differ from unsuccessful
psychopaths, the Mullins-Sweatt paper also confirms that many successful psychopaths are found in certain occupations. This paper concludes that successful psychopaths are found as professors or psychologists with an interest in law. Other studies have determined that successful psychopaths can be found as attorneys, physicians, psychiatrists, scientists, businessmen, politicians, and aristocrats (Cleckley, 1988; Hall and Benning, 2006). In his book, Clerkly (1988) speaks of a businessman who takes advantage of his interpersonal charm and risk-taking in order to gain occupational success. Through this finding, the logical jump was made to test a sample group within one of those professions. The chosen sample group for this study was bankers, which is a subgroup of businessmen. Additionally, it is estimated that 0.6-1% of the general population are psychopaths, whereas the
prevalence of psychopathy in the business world is a whopping 3.5% (Babiak and Hare, 2006)
However, not all papers have documented victory in finding successful psychopaths. For instance, Ullrich et al. (2008) investigated whether certain features of psychopathy are related to life-success. Ullrich et al. (2008) collected data from a sample of 411 men from the inner city of South London, 87% of whom were of white British ethnicity. The men went through interviews and participated in the PCL:SV, a psychopathy checklist: screening version derived from the PCL-R. From this psychopathy checklist, the researchers focused on four factors of psychopathy: interpersonal, affective, impulsivity/lifestyle, and antisocial. The researchers also measured their life success through two factors: success and wealth, and successful intimate relationships. Success and wealth was measured through the participants’ social class, income, number of rooms at home, supervision of others in job, and home ownership, and successful intimate relationships was measured through perspective of intimate relationship, living in a relationship, stability, and quality of relationship. The factors of psychopathy and two domains of life-success were then analyzed together in order to find an association. The researchers failed to find an association between the interpersonal factor and either life-success domain; affective deficiency was negatively related to both life-success measures; the impulsivity/lifestyle and
antisocial factors had no association with successful intimate relationships but negatively
predicted status and wealth. This goes to show that of the known psychopathic traits tested in this study, the characteristic either had no relation to how successful one was, or
negatively influenced the likelihood of one’s success. Although this paper could not accept the hypothesis that one’s interpersonal psychopathic traits would positively predict greater success in later life, there are possible explanations for why this is. First of all, the
interpersonal psychopathic traits measured in these men show a resemblance to narcissistic personality disorder. Narcissistic personality disorder has two levels, where the second level is a more adapted version, showing selective empathy. Ullrich et al. claim that it is possible the psychopathic traits measured here are more of the level 1-narcissism, thus showing no association with life success. Additionally, the researchers did not measure the intellectual capabilities of the men, which is a prominent feature of the successful psychopath (Cleckley, 1988). It is suggestible that those with a higher intellectual ability is more likely to succeed in life, including psychopaths. Another limitation in this study was that the researchers were not able to analyze the full psychopathic scale: the maximum total score achieved by the participants was not the possible maximum total score of the test. This meant that the study was possibly dealing with people on a lower scale of psychopathy, thereby explaining their lack of success. To ensure the study of successful psychopath, this thesis research will look at people who have diverse backgrounds. The participants will all be recruited from banks, meaning they are more likely to be psychopaths, but are also more likely to have obtained a higher education. Though the participants will not have to prove their intelligence in any way, they will be asked about their educational and work experience, so a general idea can be obtained.
Thus, previous literature regarding successful psychopaths has taught us that to find successful psychopaths, it is best to recruit based on profession as opposed to a random population and hoping some successful psychopaths can be found. However, it should also be kept in mind that the psychopathy checklists have been confirmed as successful in defining psychopaths, but they are mainly tested on caucasian males, sometimes only with those troubled with the law. Though this may create doubt in the validity of identifying psychopaths in this experiment, there is no better alternative nor one with results as promising.
2.3 Unsuccessful psychopaths
Since creating a triumphant measure of psychopath identification, researchers have
observed institutionalized psychopaths to better understand their behavior and physicality’s. For example, a study by Kiehl et al. (2001) examined the neural differences between
non-incarcerated control participants and the non-incarcerated non-psychopaths, the criminal psychopaths had “significantly less affect-related activity in the amygdala/hippocampal formation, parahippocampal gyrus, ventral striatum, and in the anterior and posterior cingulate gyri” and additionally, they had an “overactivation in the bilateral fronto-temporal cortex for processing affective stimuli” (Kiehl et al., 2001). This is interesting information to know as it tells us that there are structural differences in the brain causing psychopaths to diverge from the norm. Moreover, the limbic abnormalities indicate what kind of behavioral differences we can expect or which behavioral differences are associated with which brain region. For example, the amygdala is known for the critical role it plays in emotion (Gallagher and Chiba, 1996), thus any lesser activity there will mean a difference in emotional behavior between psychopaths and non-psychopaths. This decrease of activity in the amygdala may explain why psychopaths are less responsive when victims of their actions express fear or grief (Swaab, 2014). Some psychopaths have a more extreme malfunction of their
amygdala, inhibiting their ability to see facial expressions (Swaab, 2014). This prevents them from seeing their victims’ suffering and feeling any empathy towards them.
Furthermore, reports have described psychopaths as not only committing a
disproportionate number of crimes (Hare and Jutai, 1983), they are also more violent and aggressive than those of their non-psychopathic counterparts (Hare and McPherson, 1984). Interestingly, Hare and McPherson (1984) found that the relation between psychopathy and violence was not dependent on intelligence. This is fascinating as intelligence has a relation with both psychopathy and violence.
Usually, when one thinks of a psychopath, one imagines a serial killer with no emotion or remorse whatsoever. This reputation that psychopaths have gotten are based on what is known about unsuccessful psychopaths. The research regarding unsuccessful psychopaths have found traits of aggression, lack of empathy, and anti social behavior. Whether this behavior can be extended to describe successful psychopaths is doubtful. For this reason, research is directed much more towards studying successful psychopaths as they are more difficult to identify. Since this thesis is also directed towards the study of successful
psychopaths, it is not necessary to critique the studies regarding unsuccessful psychopaths.
2.4 Rationality
Rationality is a common assumption economic models make about man. It is assumed that people will weigh their options, risks, benefits, and costs and make educated decisions. Rational agents are expected to know their present and future tastes and maximize these interests (Kahneman, 2011). In layman’s terms, a rational man is a reasonable man. In
economics, these are two different adjectives and “the only test of rationality is not whether a person’s beliefs and preferences are reasonable, but whether they are internally
consistent” (Kahneman, 2011). Thus, internal consistency is used as a measure of rationality in this experiment.
In the world of behavioral economics and the study of decision making, simple gambles are used to understand human actions and rationality. This is because the choices made in simple gambles reflect hypothetical choices made in more complicated situations which are harder to study. Through the use of simple gambles, expected utility theory was derived: a “normative model of rational choice” (Kahneman and Tversky, 1979). This theory is based on certain axioms which ensure rational decision making. These axioms are completeness, transitivity, independence, and continuity (Von Neumann and Morgenstern, 2007).
Completeness assumes that an individual has precise preferences between two options and can always decide: for every A and B, either A ≥ B or A ≤ B. Transitivity assumes than an individual decides consistently: for every A, B, and C with A ≥ B and B ≥ C, it must be that A ≥ C. Independence assumes that preference order will not change for two alternatives when a third option is presented: for A, B, and C, where A ≥ B and t ∈ (0, 1], tA + (1-t)C ≥ tB + (1-t)C. Continuity assumes when there are three options, and an individual shows a preference for A over B and B over C, there must be a mixture of A and C which the individual shows no preference for over B: for A ≥ B ≥ C, there exists a probability, p, such that B is equally preferable as pA + (1 - p)C. If all these axioms are satisfied by an individual, this individual is said to be rational. However, there are some choice problems which violate the axioms, one of them being, Allais’ paradox (Allais, 1953). The Allais’ paradox presents a pair of choice problems which causes a violation of the expected utility theory due to a certainty effect. The paradox leads to an inconsistency of choices made by people: what they choose for the first choice is not consistent with the decision made in the second choice. If they are consistent with their preferences, this shows rationality. This line of reasoning was used in this experiment.
The Allais’ paradox has shown itself in different formats. Sometimes the choices must be made between different monetary gains, other times different leisurely gains. In order to ensure a rounded experiment for people who hold different values to different rewards, a variety of choice problems will be presented.
2.5 Contribution of this study
This study hopes to fill gaps in research with regards to the relation between psychopathy and rationality. This exact question has never been asked before, thus making it interesting to study. Psychopathy outside institutions is not studied often, hence little is known about
“successful” psychopaths. This study will provide more information to the little collection of research already conducted. As most of the studies regarding psychopathy are performed with institutionalized participants, their methods cannot be replicated here. This is because institutionalized participants and non-institutionalized participants have different
characteristics. Therefore, the methodology used is an adapted form, mixing the methods used in both studies with institutionalized subjects (because those are greater in quantity) and studies with non-institutionalized subjects (because these are more relevant to this case). This combines the best qualities of both types of research to ensure the best method for this experiment.
When it comes to measuring rationality, this has been done with the involvement of simple gambles or decision making games. As this experiment has certain requirements, such as: the participants must act on their own (read: no interactive games), it must not be too time consuming, etc., this limits the options. Simple gambles are used where the participants gamble between choices of money, health, and leisure. This variation ensures that even those participants who do not care about money, health, or leisure will act rationally, as they encounter all three types of choices.
3. Methodology
3.1 Sample
In order to answer the research question, a survey was created to observe 50 participants. As it has been established that some successful psychopaths are found in banks, all 50 participants are from banks. They are not all from the same bank, nor do they hold the same positions. The majority of them were volunteers from a Japanese bank based in
Amsterdam, thus providing participants of Japanese or Dutch nationality. To be precise, the sample consisted of 26 Japanese, 17 Dutch, 2 British, 2 American, 1 Chinese, and 1
German. Of the 50 participants, 20 were female and 30 were male.
3.2 Procedure
All 50 participants completed the survey composed for this research over a two week
period. Some completed the survey online, others completed the survey in the physical dimension. The survey had three major parts to it, each serving a separate purpose. Once the survey was completed, the participants had completed their role in this study.
3.3 Measures
The first section was to determine the participants’ level of psychopathy. This was done using the Levenson’s report psychopathy scale (LSRP). The LSRP is a 26-item self-report (Levenson et al., 1995), normally taking about 5 minutes. This was one of the reasons it was chosen as some alternatives took as long as 2 and a half hours to complete or
required the purchase of specific programs. It was essential to have a short survey as the participants were all completing the survey voluntarily, with no monetary or competitive incentive. The LSRP has been compared with other psychopathy tests and was not found to be of any lesser quality in determining the psychopathy of a person (Brinkley et al., 2001) and it is both reliable and valid (Lynam et al., 1999). Furthermore, the LSRP also
distinguishes people into different levels of primary and secondary psychopathy, which is beneficial as it provides additional data for the analysis. The 26 items of the test are all statements which the participants read and then rated on a 5-point Likert scale between
disagree strongly and agree strongly. Primary psychopathy, which is measured in the first
part of the LSRP, is concerned with the “callous, manipulative, and selfish use of others”, whereas the secondary psychopathy, measured in the second part of the LSRP, deals with “impulsivity and poor behavioral controls” (Lynam et al., 1999).
The second section of the survey tended to the rationality of the participants. This was done through a number of questions where the participants were given two options and they had to choose which deal they would rather make. Essentially, it was a set up for the Allais' paradox: a choice problem where there is an inconsistency in the observed choices relative
to the expected utility theory predictions. Since different people hold different values to money, the choice problems did not revolve only around money, but also to one’s health and holiday options.
The third and final section was for the control variables: age, gender, nationality, highest level of education, position at bank, years at bank, and charitableness. Charitableness was
included in the control section to learn about one’s level of selfishness, an important trait of a psychopath. The other control variables were there to control for any variances in behavior which could be attributed to the participants’ age, gender, nationality, education, position, or experience.
Altogether, the survey consisted of 43 questions. To see the survey and read the questions, please refer to the appendix.
3.4 Hypotheses
With little existing literature about successful psychopaths, it is difficult to have a substantial hypothesis with a lot of research to back it up. However, from what is known, there are a two possibilities.
One is that as psychopathic traits increase in an individual, they will be more rational relative to those with less psychopathic traits. This is based on previous research where researchers found that subjects more similar to the homo economicus were also scored higher on the psychopathy scale (Yamagashi et al., 2014).
It could also be that those who are of the primary psychopathy are more rational than those who are of the secondary. This is because those of the primary group are more emotionally stable and resilient to stress, relative to their secondary counterparts (Skeem et al., 2011). Moreover, Levenson et al. (1995) found that secondary psychopathy has a negative correlation with GPA. This means that those who score higher on secondary psychopathy are of a lower intellectual capability and less emotionally stable relative to their primary psychopath counterparts. With traits like these, it could be expected that secondary psychopaths are less rational than primary psychopaths. Additionally, psychopaths are known to be more impulsive according to Levenson et al. (1995), but this may only be with respect to secondary psychopaths as the primary kind are supposed to be more
emotionally stable. Thus, one could hypothesize that the primary psychopaths are more rational than secondary psychopaths.
Thus, to sum up, there are two working hypotheses and the null hypothesis:
H1a: As psychopathy level increases, rationality will also increase.
H1b: Primary psychopaths are more rational than secondary psychopaths.
4. Results
To begin the analysis of the results, the survey data had to be coded into analyzable variables. For the psychopathy part, this meant coding the 26 questions into one, two, or three factors. For the rationality part, this meant coding the 8 questions into different levels of rationality per person. The codings of these variables will be discussed first, followed by some summary statistics, first set of results, second set of results and a discussion about limitations.
4.1 Coding
Psychopathy
When the LSRP was first developed, Levenson et al. (1995) ran a factor analysis, with principal component, establishing a two factor model where the 26 questions loaded onto two factors: primary and secondary psychopathy. Over the years, literature has argued that a three factor model may be more appropriate (Sellbom, 2011; Brinkley et al., 2001). To see which model fits best with this data, the models from both Levenson et al. (1995) and Sellbom (2011) were fitted. Model fit can be measured through chi square statistics, CFI, TLI, RMSEA, AIC, and BIC. However, chi square statistics are not considered as viable statistical measures of model fit in this case as they rely heavily on sample size (Kline, 2010). Values of CFI and TLI larger than .90 and RMSEA up to .08 demonstrate an acceptable fit (Kline, 2010). Values of CFI greater than .95 and RMSEA below .05 indicate a very good fit. The AIC and BIC are used as comparative fit indices in order to compare non-nested models. The lower the AIC and BIC values are, the better the fit of the model to the data. The two factor model from Levenson et al. (1995), seen in Figure 1, has sixteen variables loading onto the primary psychopathy factor and ten variables loading onto the secondary
psychopathy factor. One variable loading onto each factor is scaled to one in order to have
a reference point and all the error terms have loadings of one to ensure the model is
identified. Of the 26 variables, six variables had negative loadings. This meant that they had the inverse effect of all the other variables and the scaling had to be reversed. These six reversed variables are those beginning with cp in the diagram. The model does not have a good fit to the data. The CFI and TLI of this model were .524 and .481, respectively,
demonstrating a highly unacceptable fit. The RMSEA was .132, concurring with the CFI and TLI values. The AIC was 656.931 and BIC was 787.022.
Figure 1: The two factor model based on Levenson et al. (1995).
The three factor model based on Sellbom (2011), seen in Figure 2, has ten variables loading onto the factor egocentricity, five variables loading onto antisocial, and four variables loading onto callous. In this model too, the negatively loading variables were reversed in scoring. This model shows an improvement in fit to the data with a CFI of .757, TLI of .721, RMSEA of .114, AIC of 325.583, and BIC of 382.135. However, the model fit is still far from ideal with none of the model fit indices reaching the cut off points.
$
With neither model having an acceptable fit to the data, literature is turned to to decide how to code the variables for psychopathy. Though the three factor model does have a better fit relative to the two factor model, the two factor model was derived on the premise of a non-incarcerated sample whereas the three factor model was derived on the premise of
incarcerated males and females. Since a non-incarcerated sample is being dealt with here, the psychopathy variables were coded into two factors. Moreover, in the current study, Cronbach’s alphas for the LSRP total score (26 items), primary psychopathy (16 items), and
secondary psychopathy (10 items) were .85, .862, and .644, respectively. These values of
alpha are consistent with previous literature (Miller et al., 2008; Levenson et al., 1995; Hicklin and Widiger, 2005; Ross et al., 2004) and also, above .8 (for LSRP total score and primary
psychopathy), thus the factors are determined to be reliable.
Each survey question was given an equal weight towards its respective factor. The first sixteen variables, including the reverse coded ones instead of the unadjusted for those necessary, were summed and then divided by sixteen to find the participants’ values for
primary psychopathy. The mean of the proceeding ten variables were also taken, producing
values for secondary psychopathy. The factors primary and secondary psychopathy were now observed variables, ready for analysis.
Rationality
To code a single variable for rationality, each response to the eight rationality questions were labelled a value. The eight questions consisted of four pairs, where the responses labelled 1 in one of the questions of the pair corresponded with the responses labelled 1 in the other question of the pair. Likewise, the responses labelled 2 in one of the questions of the pair corresponded with the responses labelled 2 in the other question of the pair. The absolute differences in response labels were calculated for each pair of questions and then summed. The summed values could be integers between 0 and 5. The most rational were at the 0 end of the scale and least rational at the 5.
4.2 Summary statistics
Thus, the main variables of this analysis are rationality, primary psychopathy, secondary
psychopathy, and LSRP total score.
Psychopathy
The total LSRP score ranged between 32 and 103, where the highest possible achievable score was 130 and the lowest 26. The mean value was 54.74, with a standard deviation of 12.71. The spread of frequency between 32 and 103 can be seen in Figure 3. The 25th, 50th, and 75th percentile were 46.0, 53.0, and 61.5, respectively.
Figure 3: Histogram of LSRP total score frequencies
The summed primary psychopathy score per participant had a minimum of 16.00 and maximum of 62.00, where the highest possible achievable score was 96.00 and the lowest 16. The mean value was 32.18, with a standard deviation of 9.55. The summed secondary
psychopathy per participant had a minimum of 14.00 and maximum of 41.00, where the
highest possible achievable score was 50.00 and lowest 10.00. The mean was 22.56, with a standard deviation of 5.35. The calculated primary psychopathy score had a minimum of 1.00 and a maximum of 3.88. The mean value was 2.01, with a standard deviation of 0.60. The calculated secondary psychopathy score had a minimum of 1.40 and a maximum of 4.10. The mean value was 2.26, with a standard deviation of 0.53.
Rationality
The value of rationality could range between 0 and 5. The minimum observed was 0 and maximum was 4. The mean value was 1.76, with a standard deviation of 1.04. The 25th, 50th, and 75th percentiles were 1, 2, and 2, respectively. To see the distribution of the rationality levels, see Figure 4 on the following page.
Figure 4: Histogram of rationality level frequencies
4.3 First set of results
In the first set of results, means will be compared and statistical tests of differences between measures will be carried out. The mean values of rationality and LSRP total score were compared across genders. The results are summarized in Table 1.
Table 1: Group statistics of rationality and LSRP total score
Levene’s test for equality of variances determined that equal variances can be assumed between the genders for both rationality and total score, F = .955, p = .333, and F = 2.334, p = .133, respectively. The mean values of rationality between female and male are not
Gender N Mean Standard
deviation
rationality male 30 1.733 980
female 20 1.800 1.152
total score male 30 54.233 14.640
significantly different, t = -.220, p = .827. The mean values of total score between female and male were not significantly different either, t = -.342, p = .734. With no significant difference between the genders, it can be assumed that the model tested is appropriate for both genders and separate models do not have to be tested per gender.
By performing a one-way ANOVA, with rationality as the dependent variable and LSRP total
score as the covariate, it is possible to see the relation between the two variables and
whether there is a significant difference in one’s rationality score depending on their LSRP
total scores. LSRP total score is put in as a covariate as it is not a nominal factor. The
ANOVA test yielded an F value (1, 48) of 0.707 and p value of .405. At a .05 significance level, this test finds an insignificant product, meaning that there is no significant difference in rationality score across different LSRP scores. This is a rather disappointing result as it indicates that as one’s psychopathy level increases, their rationality will not be influenced. However, perhaps if the total score was split into the two factors deduced previously and other control variables were included, they affect the variable rationality.
4.4 Second set of results
In this second set of results section, a regression is run and analyzed. This regression is performed in order to yield a more conclusive result. The regression performed here is a univariate general linear model, with the fixed factors: gender, ethnicity; covariates: total
score, age, education, total years working in banks; and dependent variable: rationality. The
variables are allocated as fixed factors or covariates depending on their measurement type. The variables put in as fixed factors are nominal variables, whereas the covariates are scaled variables. This means we are testing the effect of total score on rationality, controlling for one’s gender, age, ethnicity, education, and years in banks. The regression equation,
regression (1), is as follows:
rationality = a + b*total score + c*gender + d*age + e*ethnicity +f*education + g*total years in banks
After controlling for gender, age, ethnicity, education, and years in banks, the regression analysis found an insignificant effect of total score on rationality, F(1, 41) = 0.580, p = .451. This verifies the insignificant effect of LSRP score on rationality found in the previous section. There is no effect of total score on rationality, or any of the other control variables for that matter. Gender had an F value of 1.114, with a significance value of .297. Ethnicity had an F value of 1.575, with a significance value of .291. The interaction effect between gender and
ethnicity, gender*ethnicity, had an F value of 2.809, with a significance value of .101. Age
0.000, with a p value of .994. Lastly, total years in banks had an F value of 0.261, with a p value of .612.
When the regression is run with two variables for psychopathy, one for primary psychopathy and another for secondary psychopathy, it can be determined whether an affinity for
specifically primary or secondary psychopathy affects rationality. This regression is also a univariate general linear model, with the fixed factors: gender, ethnicity; covariates: primary
psychopathy, secondary psychopathy, age, education, total years working in banks; and
dependent variable: rationality. Whether the coded averages for primary and secondary psychopathy were used or just the summed values does not make any statistical difference. The regression equation, regression (2), is as follows:
rationality = a + b*primary psychopathy + c*secondary psychopathy + d*gender + e*age +
f*ethnicity +g*education + h*total years in banks
After controlling for gender, age, ethnicity, education, and years in banks, the regression analysis found an insignificant effect of primary psychopathy on rationality, F(1, 40) = 0.266, p = .609. Secondary psychopathy is also found to have an insignificant effect on rationality, F(1, 40) = 0.104, p = .748. All the other variables also had insignificant results, see Table 2 for precise values.
How often one donates to charity and feels towards charity was also measured. This was not included in the initial regression as there is a potential correlation between these variables and LSRP total score. As covariates and factors must be independent in an
ANCOVA (what was described above), the correlation between the variables and total score were checked. An ANOVA with total score as the dependent variable and one’s charity work as a factor found a significant effect of how often someone donate’s to charity on their LSRP
total score, F(3, 46) = 4.365, p = .009. This means the variable charWork, how often one
donates to charities/does charitable work in a year, is excluded from the ANCOVA. An ANOVA with total score as the dependent variable and one’s feelings about charity work as a factor found an insignificant effect of how someone feels about charity on their LSRP total
score, F(3, 46) = 1.886, p = .145. This means that the variable charFeel, how one feels
about charity donations, can be included in the ANCOVA.
Thus, another regression was performed with the fixed factors: gender, ethnicity; covariates:
total score, age, education, total years working in banks, charFeel; and dependent variable: rationality. There is no need to run this regression for separate variables for primary and
secondary psychopathy as it has been shown that the independent factors have no effect or significant difference on rationality. This means we are testing the effect of total score on
rationality, controlling for one’s gender, age, ethnicity, education, years in banks, and feelings
about charity. The regression equation, regression (3), is as follows:
rationality = a + b*total score + c*gender + d*age + e*ethnicity + f*education + g*total years in banks + h*charFeel
The effect of total score on rationality was still found to be insignificant, F(1, 40) = 0.472, p = .496. Gender, ethnicity, the interaction term between gender and ethnicity (gender*ethnicity),
age, education, total years working in banks, charFeel also had insignificant results. For the
exact F and p values, see Table 2.
Table 2: F and p values for every variable per regression discussed in this section.
Notes: The dependent variable is rationality.
Regression (1) Regression (2) Regression (3)
F Sig. F Sig. F Sig.
Corrected Model 0.913 .516 0.792 .626 0.794 .624 Intercept 3.391 .073 2.804 .102 2.218 .114 Total score 0.580 .451 - - 0.472 .496 Primary psychopathy - - 0.266 .609 - -Secondary psychopathy - - 0.104 .748 - -Age 0.594 .445 0.574 .453 0.598 .444 Education 0.000 .994 0.000 .987 0.000 .984
Total years in bank 0.261 .612 0.255 .616 0.263 .611
charFeel - - - - 0.018 .893 Gender 1.114 .297 0.991 .325 1.095 .302 Ethnicity 1.575 .219 1.532 .228 1.546 .226 gender*ethnicity 2.809 .101 2.743 .105 2.746 .105 R squared .151 .151 .152 N 50 50 50
4.5 Limitations
The most obvious limitation of this experiment is the sample size. With only fifty participants in this study, the conclusions drawn cannot be drawn with absolute confidence or certainty. The likelihood of error is larger than would be with more subjects. However, despite fifty being a rather small number, it is still considered to be sufficient and not too small a sample. Some studies still conduct statistical analyses with as little as 20-30 participants (think: fMRI studies and their monetary costs). One issue with a small sample is that the results found may not be relatable or generalizable to the greater population. This is because the small sample does not contain the realistic variability found in the outside population. Thus, any conclusions drawn from a small sample as such would be discarded, but it is difficult to know when such a case occurs. Another issue with a small sample is that it does not give you the correct significant results that would be acquired through the analysis of a larger sample. This issue that small samples could lead to insignificant results where the results should be significant can also lead to many false positives, which is why the second set of results is necessary. Luckily, in this case, the results were clearly insignificant in both sets of results, preventing the manifestation of false hope.
The limitation of small sample touches upon the matter of external validity. Firstly, as
mentioned before, a small sample is less likely to be representative of a population, meaning that the conclusions drawn about the small sample may not be externally valid. Secondly, the great majority of the participants of this present study came from the same bank and many were also of the same nationality. This could form some sort of exclusion bias which would also affect the external validity of results. Thirdly, there may be a slight self-selection bias. Some of the participants were emailed the survey which they could then answer online. This choice of whether to participate or not had no consequences nor
compensation, thus the decision to partake may show that the subjects all share a common trait.
Another limitation could be the medium through which participants answered the survey. Some answered online whereas others answered in paper form. Those answering the survey on paper would have to return the surveys to a person, meaning their answers could potentially be linked to their identity. This could affect the ways in which they answered questions and their responses may not have been their true feelings. There may have been people concerned with social norms and deviating from such norms, whether they
responded online or in person. These concerns would not allow them to answer honestly. The potential of this occurring is quite high in this sample as there were quite a few
5. Discussion and Conclusion
To conclude, the answer to whether psychopath bankers are more rational and whether bounded rationality will change as psychopathic traits differ is that there is no statistical difference between those with different psychopathic traits, and psychopathy does not significantly predict rationality. This is what was found in this study, however the main findings in the field are somewhat mixed. On the one hand, there is evidence in existing literature that those who have psychopathic traits are more rational and most similar to the economic man (Yamagashi et al., 2014). On the other hand, papers have also found unexpected results where psychopathy did not show the anticipated relation with other life characteristics. For instance, Ullrich et al. (2008).
Thus, there are limitations on how to interpret these findings in existing literature and in this thesis. Firstly, all of these studies with conflicting findings are set in different places, with different kinds of people and different goals. The heterogeneity of these experiments makes them difficult to compare. It cannot be known that what counts for the experiment in Japan will also be valid for the experiment in America. Hence, it is questionable whether the
interpretation of the findings are universally transferrable. There is the problem of
generalizability and the extent to which it is possible. Secondly, the majority of research concerning psychopaths is conducted within institutions. This is a limitation as what is applicable inside institutions may not be applicable in the outside world. This is especially the case regarding psychopaths as those institutionalized are even labelled differently to those who are not, hence the terms unsuccessful and successful psychopaths. So, this limitation also pertains to the issue of generalizability. Thirdly, the identification of
psychopaths has always proven to be a challenge, but it is even more so when attempting to compare results. What one may classify as a psychopath with a certain psychopathy test may yield different results with another psychopathy test. For this reason, the LSRP was used in this case as it has proven effective in research situations. However, this may not allow the comparison of results and traits from studies where a different psychopathy test was used. Additionally, most psychopathy tests have been around for over twenty years. Though many psychological tests have existed for decades, the progressions within the research regarding psychopathy may call for an updated collection of psychopathy tests. Lastly, the findings are difficult to interpret as the sample size is smaller than most studies. As discussed in the results section, there are issues associated with small sample sizes, making it difficult to be more conclusive.
Nevertheless, conclusions were derived in the results section. The main conclusion was that psychopathy level has no significant effect on rationality. This indicates that the null
hypothesis is accepted: there is no relationship between the levels of psychopathy and rationality of a person. Accordingly, the alternative hypotheses are rejected. To recapitulate, the rejected hypotheses are that as psychopathy levels increase, rationality will also increase and that primary psychopaths are more rational than secondary psychopaths. This research also found a lack of relation between psychopathy and rationality even when one’s
background was controlled for. Regressions were run including the participants’ gender, age, ethnicity, education, and years of experience. Even then the effect of psychopathy had a severely insignificant result on rationality. To see whether the rationality of participants may have depended more on primary or secondary psychopathy rather than the entirety, the two were entered as separate variables into the regression, replacing total level of psychopathy. This did not produce any significance in the model, with both primary and secondary psychopathy having statistically insignificant effects on rationality. If one was to analyze the effects of the variables on rationality, regardless of the significance level, one would find that the effects are all positive and in the direction expected. Unfortunately, the lack of statistical significance denies us from drawing any directional relationships regarding psychopathy and rationality.
With the lack of statistical significance in this study, there are definitely steps for future research. The first, and most obvious, modification would be to conduct the experiment with more participants. This would lead to a more conclusive result where even if the relation is insignificant, it is known for a fact that it is due to the relation actually being insignificant rather than the possibility of the sample size giving a false negative. A further step would be to conduct the experiment using neurological measures. For instance, instead of using the LSRP to identify psychopaths, one could use fMRI scans to identify them through their limbic abnormalities. This would allow a clearer identification of psychopaths with less reliance on the subjects to be honest. A measurement of their physical being is not something which can be modified: it is a clear image of something that is there or that is not. The potential issue with this is that the past research regarding the limbic differences between psychopaths and non-psychopaths was conducted with institutionalized
psychopaths (Kiehl et al., 2001). There is a possibility that successful psychopaths may not be the same neurologically as unsuccessful psychopaths. Another possible step for future research would be to conduct this study with multiple measures of rationality. This study relied solely on the Allais’ paradox to measure rationality, but if multiple measures had been used, then the participants’ rationality would be measured with less space for error.
Additionally, if some gambles had been included where the participants were paired with one another, they may have made choices differently as they would take the effect they have on their partner into consideration. The inclusion of these paired gambles would show the
compassion and empathy of a subject, allowing us to see whether their decisions in the psychopathy test are also made in real life situations. This was, unfortunately, not possible in this experiment as most subjects were already acquainted with one another and all the different relations between the subjects would be difficult to control for. Furthermore, an interesting addition to this experiment would be to test the relation between psychopathy and rationality in all directions. This paper specifically focused on whether psychopathy could predict rationality, but future research could focus on the opposite relation: whether rationality can predict psychopathy. For this to be possible, the study would have to find people of all rationality levels. Lastly, all the aforementioned steps could also be combined. All in all, this paper has looked into something never researched before, and so even though the results were not significant statistically, they were significant in other measures.
7/13/2015 Thesis research 6. In today's world, I feel justified in doing anything I can get away with to succeed. * Mark only one oval. 1 2 3 4 5 Disagree Agree 7. People who are stupid enough to get ripped off usually deserve it. * Mark only one oval. 1 2 3 4 5 Disagree Agree 8. Looking out for myself is my top priority. * Mark only one oval. 1 2 3 4 5 Disagree Agree 9. I tell other people what they want to hear so that they will do what I want them to do. * Mark only one oval. 1 2 3 4 5 Disagree Agree 10. I would be upset if my success came at someone else's expense. * Mark only one oval. 1 2 3 4 5 Disagree Agree 11. I often admire a really clever scam. * Mark only one oval. 1 2 3 4 5 Disagree Agree 12. I make a point of trying not to hurt others in pursuit of my goals. * Mark only one oval. 1 2 3 4 5 Disagree Agree
7/13/2015 Thesis research 13. I enjoy manipulating other people's feelings. * Mark only one oval. 1 2 3 4 5 Disagree Agree 14. I feel bad if my words or actions cause someone else to feel emotional pain. * Mark only one oval. 1 2 3 4 5 Disagree Agree 15. Even if I were trying very hard to sell something, I wouldn't lie about it. * Mark only one oval. 1 2 3 4 5 Disagree Agree 16. Cheating is not justified because it is unfair to others. * Mark only one oval. 1 2 3 4 5 Disagree Agree Rate each statement on how much you agree with it on a scale between strongly agree and strongly disagree. 17. I find myself in the same kinds of trouble, time after time. * Mark only one oval. 1 2 3 4 5 Disagree Agree 18. I am often bored. * Mark only one oval. 1 2 3 4 5 Disagree Agree
7/13/2015 Thesis research 19. I find that I am able to pursue one goal for a long time. * Mark only one oval. 1 2 3 4 5 Disagree Agree 20. I don't plan anything very far in advance. * Mark only one oval. 1 2 3 4 5 Disagree Agree 21. I quickly lose interest in tasks I start. * Mark only one oval. 1 2 3 4 5 Disagree Agree 22. Most of my problems are due to the fact that other people just don't understand me. * Mark only one oval. 1 2 3 4 5 Disagree Agree 23. Before I do anything, I carefully consider the possible consequences. * Mark only one oval. 1 2 3 4 5 Disagree Agree 24. I have been in a lot of shouting matches with other people. * Mark only one oval. 1 2 3 4 5 Disagree Agree 25. When I get frustrated, I often "let off steam" by blowing my top. * Mark only one oval. 1 2 3 4 5 Disagree Agree
7/13/2015 Thesis research 26. Love is overrated. * Mark only one oval. 1 2 3 4 5 Disagree Agree Would you rather.. 27. Choose between * Mark only one oval. €2,400 with certainty €2,500 with probability .33, €2,400 with probability .66, and €0 with probability .01 28. Choose between * Mark only one oval. 50% chance to win a threeweek tour of France, Italy, and Spain A oneweek tour of France, with certainty 29. Choose between * Mark only one oval. €4,000 with probability .8 and €0 with probability .2 €3,000 with certainty 30. Choose between * Mark only one oval. Living for 12 years in full health then death with a probability of .89, living for 18 years in full health then death with a probability of .10, and immediate death with probability of .01 Living for 12 years in full health then death with certainty 31. Choose between * Mark only one oval. €2,500 with probability .33 and €0 with probability .67 €2,400 with probability .34 and €0 with probability .66 32. Choose between * Mark only one oval. 10% chance to win a oneweek tour of France 5% chance to win a threeweek tour of France, Italy, and Spain
7/13/2015 Thesis research 33. Choose between * Mark only one oval. Immediate death with probability .89 and living for 12 years in full health then death with probability .11 Immediate death with probability .90 and living for 18 years in full health then death with probability .10 34. Choose between * Mark only one oval. €3,000 with probability .25 and €0 with probability .75 €4,000 with probability .2 and €0 with probability .8 Tell me about yourself 35. Gender * Mark only one oval. Male Female 36. Age * Mark only one oval. 21 30 31 40 41 50 51 60 37. Nationality 38. Highest level of education * Mark only one oval. High school diploma Bachelor's degree Master's degree Doctorate degree Other: 39. How do you feel about charity donations * Mark only one oval. 1 2 3 4 5 Useless Important
7/13/2015 Thesis research Powered by 40. Current position at bank * 41. Years at current bank * 42. Total years working in banks * 43. How often do you donate to charities/do charitable work in a year * Mark only one oval. never 12 times 34 times 5 or more times
7. References
Allais, M. (1953). Le comportement de l'homme rationnel devant le risque: critique des postulats et axiomes de l'école américaine. Econometrica: Journal of the Econometric
Society, 503-546.
Allais, M. (1979). The so-called Allais paradox and rational decisions under uncertainty. In
Expected utility hypotheses and the Allais paradox (pp. 437-681). Springer Netherlands.
Babiak, P., and Hare, R. D. (2006). Snakes in suits: When psychopaths go to work. New York: Regan Books.
Brinkley, C. A., Schmitt, W. A., Smith, S. S., and Newman, J. P. (2001). Construct validation of a self-report psychopathy scale: Does Levenson’s self-report psychopathy scale measure the same constructs as Hare’s psychopathy checklist-revised?. Personality and Individual
Differences, 31(7), 1021-1038.
Cleckley, H. M. (1988). The mask of sanity: An attempt to clarify some issues about the so
called psychopathic personality. Aware Journalism.
Fan, C. P. (2002). Allais paradox in the small. Journal of Economic Behavior & Organization,
49(3), 411-421.
Gallagher, M., and Chiba, A. A. (1996). The amygdala and emotion. Current opinion in
neurobiology, 6(2), 221-227.
Hall, J. R., and Benning, S. D. (2006). The “successful” psychopath. In Patrick, C. J. (Ed.),
Handbook of Psychopathy (pp. 459-475). New York, NY: Guilford Press.
Hare, R. D., Clark, D., Grann, M., and Thornton, D. (2000). Psychopathy and the predictive validity of the PCL-R: An international perspective. Behavioral sciences & the law, 18(5), 623-645.
Hare, R. D., and Jutai, J. W. (1983). Criminal history of the male psychopath: Some preliminary data. In Prospective studies of crime and delinquency (pp. 225-236). Springer Netherlands.
Hare, R. D., and McPherson, L. M. (1984). Violent and aggressive behavior by criminal psychopaths. International journal of law and psychiatry, 7(1), 35-50.
Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., and McElreath, R. (2001). In search of homo economicus: Behavioral experiments in 15 small-scale societies.
Hicklin, J., and Widiger, T. A. (2005). Similarities and differences among antisocial and psychopathic self‐report inventories from the perspective of general personality functioning.
European Journal of Personality, 19(4), 325-342.
Ishikawa, S. S., Raine, A., Lencz, T., Bihrle, S., and Lacasse, L. (2001). Autonomic stress reactivity and executive functions in successful and unsuccessful criminal psychopaths from the community. Journal of abnormal psychology, 110(3), 423-432.
Kahneman, D. (2011). Thinking, fast and slow. Macmillan.
Kahneman, D., and Tversky, A. (1979). Prospect theory: An analysis of decision under risk.
Econometrica: Journal of the Econometric Society, 263-291.
Kiehl, K. A., Smith, A. M., Hare, R. D., Mendrek, A., Forster, B. B., Brink, J., and Liddle, P. F. (2001). Limbic abnormalities in affective processing by criminal psychopaths as revealed by functional magnetic resonance imaging. Biological psychiatry, 50(9), 677-684.
Kline, R. B. (2010). Principles and practice of structural equation modeling (3rd ed.). New York: The Guilford Press.
Levenson, M. R., Kiehl, K. A., and Fitzpatrick, C. M. (1995). Assessing psychopathic
attributes in a noninstitutionalized population. Journal of Personality and Social Psychology,
68, 151–158.
Lynam, D. R., Whiteside, S., and Jones, S. (1999). Self-reported psychopathy: A validation study. Journal of Personality Assessment, 73(1), 110-132.
Miller, J. D., Gaughan, E. T., and Pryor, L. R. (2008). The Levenson Self-Report Psychopathy Scale: An examination of the personality traits and disorders associated with the LSRP factors. Assessment, 15(4), 450-463.
Mullins-Sweatt, S. N., Glover, N. G., Derefinko, K. J., Miller, J. D., and Widiger, T. A. (2010). The search for the successful psychopath. Journal of Research in Personality, 44(4),
554-558.
Paulhus, D. L., and Williams, K. M. (2002). The dark triad of personality: Narcissism, Machiavellianism, and psychopathy. Journal of research in personality, 36(6), 556-563. Ross, S. R., Lutz, C. J., and Bailley, S. E. (2004). Psychopathy and the five-factor model in a noninstitutionalized sample: A domain and facet level analysis. Journal of Psychopathology
Sellbom, M. (2011). Elaborating on the construct validity of the Levenson Self-Report Psychopathy Scale in incarcerated and non-incarcerated samples. Law and human
behavior, 35(6), 440-451.
Skeem, J. L., Polaschek, D. L., Patrick, C. J., and Lilienfeld, S. O. (2011). Psychopathic personality bridging the gap between scientific evidence and public policy. Psychological
Science in the Public Interest, 12(3), 95-162.
Swaab, D. F. (2014). We are Our Brains: A Neurobiography of the Brain, from the Womb to
Alzheimer's. Spiegel & Grau.
Ullrich, S., Farrington, D. P., and Coid, J. W. (2008). Psychopathic personality traits and life-success. Personality and Individual Differences, 44(5), 1162-1171.
Von Neumann, J., and Morgenstern, O. (2007). Theory of games and economic behavior. Princeton university press.
Yamagashi, T., Li, Y., Takagishi, H., Matsumoto, Y., and Kiyonari, T. (2014). In search of homo economicus. Psychological Science, 25(9), 1699-1711.
