A Moderated Mediation Analysis to Further Examine the Role of Verbal Intelligence in the Association Between Psychopathic Personality and Crime

The current study examined the association between psychopathy, criminal behavior, and the role of verbal intelligence. One promising approach is to examine alternative links between psychopathic traits and criminality like moderation and mediation effects by considering the potential relevance of verbal intelligence as a possible moderating variable. We hypothesized that psychopathic traits linearly predict antisocial behavior (ASB) but that a conviction because of ASB is moderated by verbal intelligence. To test a path model of this hypothesis, N = 305 participants (42% women; n = 172 inmates of German correctional facilities) filled in questionnaires to assess psychopathic traits, ASB, criminal behavior, and verbal intelligence. The moderated mediation analysis revealed that high psychopathic traits go along with a higher number of ASB, whereas individuals with higher verbal intelligence were more likely to evade detection, thus being more successful in their antisocial acts. These results sheds further light on the construct of adaptive psychopathy, supporting the notion that also non-incarcerated psychopathic individuals act highly antisocial. Only separate factors like verbal intelligence might mitigate negative consequences. Further implications for the concept of successful psychopathy are discussed.

In his influential monograph "The Mask of Sanity" Cleckley (1941) defined affective and interpersonal detachment as well as an disinhibited behavioral disposition as core constructs of the psychopathic personality disorder.Although there are different conceptualizations of psychopathy (Cleckley, 1941;Hare, 2003;Lilienfeld & Andrews, 1996), all of them agree that there is a strong relationship between (high) psychopathic traits on the one hand and the likelihood of deviant and delinquent behavior on the other hand.Consequently, the prevalence of psychopathy is substantially higher in prison samples than in non-criminal samples (e.g., 15% vs. 1%, Coid et al., 2009;Hare, 1991Hare, , 2003)).However, there are also individuals scoring high in psychopathic traits but are still not sentenced for an official crime (Hare, 1996).These individuals represent a group of non-criminal, adaptive, or sometimes called "successful" psychopaths (Benning et al., 2018) who can succeed in private and particularly in professional life at least from a subjective point of view (Babiak & Hare, 2006;Dutton, 2012).Widom (1977) was the first to study adaptive psychopaths, which she recruited through advertisements in the newspaper.Although not providing a criminal psychopathic control group, Widom (1977) revealed that a high number of the adaptive psychopaths had been arrested (64%), whereas only 18% of the sample reported to ever have been convicted, thus indicating an antisocial but still "successful" lifestyle, in terms of avoiding serious consequences of their antisocial behavior (ASB).Besides this, the group of successful psychopathic individuals showed a higher level of education than their criminal counterparts, whereas the socioeconomic was positioned at the lower end of the spectrum and only somewhat higher than traditionally defined psychopaths.No differences were found in Machiavellism and executive functioning but the majority of the sample chose the inhibitation option in a delay of gratification task.
To systemize the differences between adaptive and criminal psychopathy or between psychopathic personality and crime, respectively, Benning et al. (2018) described three models for successful psychopathy, representing three possible links between psychopathic traits and criminal behavior: First, the non-criminal psychopath is a subclinical manifestation of the criminal psychopath assuming a linear relationship between psychopathy and crime with many shared characteristics, for example, personality correlates and drug use.Second, in the model of moderated psychopathy, there are additional compensatory variables that moderate the association between psychopathy and criminality to separate non-criminal from criminal psychopaths, such as age, intelligence, executive functioning, parenting, socioeconomic status, physiology, and neuroanatomy.Third, in the multiprocess model of psychopathy, there are two different etiologically distinct dimensions of psychopathy, weak defensive (fear) reactivity that is associated with the interpersonal-affective features of psychopathy, such as manipulativeness and low empathy, and deficits in executive functioning that could manifest in relatively high degrees of impulsivity and irresponsibility.
In the model of moderated psychopathy, the search for possible protective or risk factors for criminal behavior in psychopathic individuals is prominent.Intelligence has been found to be a stable correlate of criminal behavior and is negatively related to self-reported delinquent behavior as well as to official records of criminal behavior (Bartels et al., 2010;Beaver & Wright, 2011;Lynam et al., 1993;Schwartz et al., 2015;Ttofi et al., 2016).Based on all of the available literature, intelligence appears to be among the strongest predictors of criminal behavior (Boccio et al., 2018).Some studies exhibit a particularly strong effect for verbal intelligence (Moffitt, 1993;Quay, 1987).These studies point to the fact that people with higher IQ do not engage in antisocial or criminal behavior in general (Schwartz et al., 2015).However, Boccio et al. (2018) focused on successful criminals, meaning people who conduct criminal behavior but are at the same time able to avoid detection and any further legal consequences of their criminal acts.They showed in their study that high intelligent male offenders were able to successfully avoid detention and apprehension of their illegal acts.In the specific field of psychopathy, a study from Boccio and Beaver (2018) investigates, whether psychopaths are able to avoid detection, even after being engaged in serious or frequent criminal behavior.Since psychopaths are characterized as being clever and manipulative, the question was if this enables psychopathic individuals to being successful in evading detection of their criminal acts.Boccio and Beaver (2018) revealed that individuals with high levels of psychopathic traits commit more crimes and report more arrests without any advantage in avoiding arrests.
Taken together, intelligence seems to be an important variable in the successful behavior of psychopathic individuals.In this notion, Wall et al. (2013) revealed in a sample of undergraduates that verbal intelligence moderated the relationship between psychopathy and self-reported ASB, such that those higher on both psychopathy and intelligence had engaged in less criminal behavior.These findings were replicated by Watts et al. (2016), who found that intelligence served as a protective factor against self-reported ASB among high psychopathic persons in a sample of undergraduates.Similarly, Moraga et al. (2019) found small but significant effects in their study indicating that psychopathic traits and verbal intelligence interacted in the prediction of ASB.These results demonstrate that individuals with high levels of psychopathy and high levels of verbal intelligence engaged in less antisocial behavior than their low verbal intelligent counterparts, thus supporting the hypothesis of verbal intelligence as a protective factor.Gao and Raine (2010) as well as Lilienfeld et al. (2015) concluded that protective factors such as superior executive functioning and information processing allow psychopathic individuals to act more socially adaptive or to better conceal their antisocial behavior.
These findings reflect the notion of the first model suggested by Benning et al. (2018) that higher psychopathic traits go along with a higher number and higher severity of criminal acts (Beaver et al., 2017;DeLisi, 2009;Hare, 1996;Vaughn & DeLisi, 2008).However, according to the model of moderated psychopathy by Benning et al. (2018), verbal intelligence might act as a moderator in the relationship between psychopathy and criminal behavior that is detected (and therefore by definition unsuccessful) by law enforcement authorities (i.e., criminal psychopathy).
The current study aims to further investigate the nature of successful psychopathy.Doing so, we want to shed further light on the association between psychopathic traits and criminality, and to investigate the role of the protective factor verbal intelligence.Previous studies have shown that psychopathy is associated with high antisocial and criminal behavior, but verbal intelligence might enable criminals to evade detection from law enforcement facilities.Until now, there is still a scarcity of studies that simultaneously include several models of successful psychopathy (the model of subclinical psychopathy as well as the model of moderated psychopathy) and test these models on all individuals of the relevant sample: non-institutionalized and criminal individuals, males and females combined.Therefore, we aimed to test in the current study the hypotheses that higher psychopathic traits go along with a higher number of antisocial behaviors.Furthermore, these antisocial behaviors are more likely to be officially sanctioned in low verbal intelligent individuals who are not able to evade detection compared to individuals with higher verbal intelligence.

Sample
The present study comprises two samples with a total of N = 305 participants.The first sample consists of n = 132 civilians, the second sample consisted of n = 173 inmates of German correctional facilities who completed the questionnaires (for further details see Etzler & Rohrmann, 2017b).A detailed description of both subsamples can be found in Table 1.As expected, the mean, standard deviation, and mode of psychopathy are higher within the inmate sample compared to the civil sample.

Measures
Questionnaire of Psychopathic Personality Traits (FPP).To assess psychopathic personality traits, the FPP (German: Fragebogen Psychopathischer Persönlichkeitseigenschaften; Etzler & Rohrmann, 2017a, 2017b) was used in its original German version.It comprises 30 items to assess six different facets of psychopathy: Lack of empathy, Fearlessness, Narcissistic Egocentrism, Impulsivity, Social Manipulation, and Power.Items are rated on a six-point Likert scale with 0 = do not agree at all to 5 = agree completely.Exemplary items are: "It has happened that a highly emotional book/ movie made me cry" (Lack of Empathy, inverted scale), "I love it when others do what I tell them to do" (Power).Scores for subscales are obtained by summing up five items each; the total score is obtained by summing up all 30 items.The internal consistency of the subscales ranges between McDonald's ω = .63and .80,for the total scale the internal consistency was McDonald's ω t = .90.The test-retest-reliability for the subscales ranges between r tt = .70and .86,and was for the total score r tt = .85(Etzler & Rohrmann, 2017a).The FPP generally exhibits a six-factorial structure with a hierarchical psychopathy factor, and its measurement properties turned out to be stable in both a civilian as well as an inmate sample.There are high correlations with total scores of the Psychopathic Personality Inventory-Revised (PPI-R; Eisenbarth & Alpers, 2007;Lilienfeld & Widows, 2005), r = .677,p < .001,and significant associations with the number of violent convictions, r = .202,p = .008as well as with different educational and disciplinary sanctions r = .315,p < .001.A further currently conducted validation study of the FPP (Pieschel, 2017) using a sample of N = 82 forensic inpatients revealed a significant correlation between the FPP and the Psychopathy-Checklist: Screening Version (PCL:SV, Hart et al., 1995), r = .282,p = .01.

Checklist of Antisocial Behavior and Criminality (CAV/K).
To assess antisocial behavior and criminality, we used the CAV/K (German: Checkliste für Antisoziale Verhaltensweisen und Kriminalität; Etzler et al., 2017), a self-report checklist which consists of 18 items measuring the frequency of different antisocial behaviors.Exemplary items are: "I beat someone up so strongly that he had to go to the doctor," or "I changed or  forged a certificate (e.g., school certificate)."Items are ordered in increasing severity and range from quite common antisocial behaviors like lying and bullying to severe behaviors like humiliation, rape, and homicide.Behaviors are rated on a six-point scale from 0 = never to 5 = very often.Additionally, each item receives a second criminal behavior rating, whether respondents had been convicted for this antisocial behavior or not (0 = no or 1 = yes).The internal consistency was α = .90for Antisocial Behavior and α = .85for Criminality.The construct validity of the Antisocial Behavior index is supported by correlations with, for example, education, r = −.494,p < .001,and the number of violent offences, r = .222,p < .001.The construct validity of the Criminality index is supported by correlations with education, r = −.546,p < .001, the number of violent offenses, r = .323,p < .001,and the length of prison sentence, r = .246,p < .001.

German Multiple Choice Vocabulary Test (MWT-B).
The MWT-B (German: Mehrfachwahl-Wortschatz-Test-B; Lehrl, 2005) assesses verbal intelligence by using 37 items.One item consists of five words and one of these words represents an existing German word, whereas the other four words are pseudo-words that sound similar to the existing one.Participants are asked to choose the existing word.Items are ordered in increasing difficulty.Scores are obtained by summing up all correct answers giving one point for each correct answer.Total scores can be transformed into IQ values regarding verbal intelligence.The reliability of the MWT-B is high with a test-retest coefficient of r tt = .87(14 months) and a correlation with its parallel version MWT-A of r tt = .84(Blaha & Pater, 1979).The MWT-B correlates highly with a global IQ index in adults r = .72as well as with intelligence measured by using the German adaption of the Wechsler Adult Intelligence Scale, r = .81(Lehrl et al., 1995;Wiessner & Felber, 1981).

Procedure
Participants of the civil population were recruited from the community through personal and online advertisement, two 25€ vouchers were raffled off among the participants for reward.Inmates were recruited by disseminating leaflets containing the relevant information of the present study and the opportunity to participate; participation was voluntary and was rewarded with an expense allowance of 5€.Within a larger study described elsewhere (Etzler & Rohrmann, 2017b) all participants filled in the FPP assessing psychopathy, the CAV/K assessing antisocial as well as criminal behavior, and the MWT-B assessing verbal intelligence.

Statistical Analysis
In order to detect outliers, multivariate normality of all predictor variables was examined through Mahalanobis Distance Indices and accordingly, five participants were excluded from further analyses because their scores strongly deviated from multivariate normality (Stevens, 2012).Missing values were treated with the Full Information Maximum Likelihood (FIML) approach within the ML estimation of the path model (Muthén & Muthén, 1998-2012).
Since the hypotheses represent moderation and mediation effects that include several dependent variables, we calculated a manifest path model which allows estimating these effects simultaneously (see Figure 1).According to Preacher et al. (2007), we analyzed a mediation model in which the indirect effect is moderated by an additional variable (which corresponds to Model 3 in Preacher et al., 2007).The basic mediation model includes the dependent variable crime which has been regressed on the predictor psychopathy (PP), with ASB being the mediator in this pathway.Thus, the total effect of psychopathy on crime can be divided into the direct effect from psychopathy to crime and the indirect effect from psychopathy to crime via ASB.To additionally estimate the moderation effect of verbal intelligence (IQ) on the indirect effect, a product term of IQ × ASB was included in the mediation model (e.g., Aiken et al., 1991), as well as the main effect of IQ on crime (see Figure 1).Thus, the path model consists of two separate regression models: In the first model, ASB was regressed on PP (equation (1)).In the second model, crime was predicted by ASB, IQ, and the continuous interaction term ASB × IQ (equation ( 2)).The conditional indirect effect as moderated mediation effect of psychopathy on criminal behavior is represented by a 1 (b 1 + b 3 × IQ).All parameters in the model were estimated through Maximum Likelihood Estimation (Muthén & Muthén, 1998-2012).With regards to the mediation effect, standard errors of the indirect effect were estimated through bootstrapping (MacKinnon et al., 2004).To interpret effect sizes, standardized coefficients are reported, and total, direct, and indirect effects were calculated.Since the indirect effect is a function of the moderator, we determined the range of moderator values producing significant indirect effects, the so-called region of interest.Additionally, we visualized the indirect effect and its bootstrapped bias-corrected confidence interval band for all moderator values between ± 2 SD.

Descriptive Statistics
Table 2 shows the descriptive statistics of all included variables in the analysis.As it can be seen, psychopathy as well as five out of six subscales turned out to be higher in the inmate sample.Only the psychopathic trait power is slightly higher in the nonincarcerated sample compared to inmates.Antisocial behavior as well as criminal behavior is higher in inmates, whereas verbal intelligence is significantly lower in the inmate sample with almost one standard deviation (M diff = 14.9, t = 9.6, p < .001).Psychopathy shows a low and negative correlation with IQ, on the one hand, and positive correlations with ASB and crime, on the other hand.

Moderated Mediation Analysis
In the first step, we estimated the complete saturated model.As the results show the effect from psychopathy to crime was completely mediated, since the direct effect of psychopathy on crime was not significant (b 4 = −0.072,p = .214,CI [−0.188, 0.039]; χ² diff = 1.60) and was therefore fixed to zero in the following analysis.Indices indicated a good model fit, χ² = 1.60, df = 1, χ²/df = 1.6, RMSEA = 0.045, CFI = 0.999, SRMR = 0.009.Standardized model coefficients and significance tests based on bootstrapped standard errors are shown in Figure 2.
As it can be seen, the first part of the model, following hypothesis 1, reveals that psychopathy has a strong positive effect on ASB: the higher psychopathy, the more often ASB have been reported.Results of the second part of the model show that there is a high direct effect from ASB to crime: the higher the frequency of ASB, the higher the frequency of crime.IQ also has a high direct effect on crime: participants with lower IQ reported more criminal behaviors than participants with higher IQ scores.Additionally, the interaction effect of ASB and IQ on crime was significant and negative, that is, participants with lower IQ have a higher probability of being convicted because of ASB than participants with higher IQ scores.Taking both paths into account according to hypothesis 3, a significant linear indirect effect from psychopathy to crime can be observed with a 1 × b 1 = 0.32, p < .001,that is, the higherpsychopathy, the more often convictions were reported.The interaction effect of ASB and IQ on crime describes a conditional indirect effect: for participants with low IQ, psychopathically influenced ASB led more often to crime than for participants with high IQ.This means that the indirect effect between psychopathy and crime is only significant within a certain range of IQ scores.The model explained 55% of the total variance of crime.The significant interaction effect supported the hypothesis of a moderated mediation.To further examine the conditional effects, total effects from psychopathy to crime were tested for specific values of the moderator IQ (M ± 1 SD).Results of the analysis are listed in Table 3: all indirect effects were significant within ± 1 SD.This implies that the prediction of crime by psychopathy is possible within the IQ range of M ± 1 SD and that the indirect effect from psychopathy to crime is higher for participants with lower IQ values compared to participants with higher IQ scores.
To examine the region of interest (i.e., the range of values of the moderator where the conditional indirect effect does not equal zero), the conditional indirect effect of psychopathy on crime (via ASB) for different IQ scores was examined (see Figure 3).Additionally, a confidence interval band of the indirect effect was calculated (the   confidence interval indicates significance as long as it does not contain 0).As Figure 3 shows, IQ values lower than M + 1.2 SD imply a significant indirect effect from psychopathy to crime via ASB.

Discussion
The study aimed to shed further light on the moderating role of verbal intelligence within the relationship between psychopathic traits and crime.In general, the data of the present study support our hypotheses, leading to two main findings of the study: First, we revealed a high effect size in the prediction of antisocial behavior by psychopathic traits.The standardized path coefficient (a 1 = .65)exceeds r = .300which has been widely found within the consistency debate as being a stable effect of the relationship between trait and behavior (Mischel, 1968).Although one should consider that common method variance probably slightly overestimates this effect (all variables were assessed through self-report), this result might be understood in terms of matching specificity of both variables, psychopathy and antisocial behavior in the sense that psychopathy as a latent superordinate trait (Cronbach & Meehl, 1955;Lilienfeld, 1994) is linked to an infinite variety of possible antisocial behavioral acts.Since psychopathy and antisocial behavior both are on a similar level of specificity, predictive effects might reach a relatively high precision.This finding taps into the notion that psychopathic traits as constructs of interpersonal and emotional deficits (Lynam & Miller, 2012) are strongly linked to antisocial behavior, even though no officially recorded criminal behavior was detected (Benning et al., 2018).Although the myth of the successful psychopath remains vivid (Dutton, 2012;Mullins-Sweatt et al., 2010), especially with an eye on white collar crimes (Babiak & Hare, 2006), the current study supports the notion that psychopathic traits are most likely associated with unfavorable outcomes, such as antisocial behavior.Other studies in the field also find rather unfavorable correlates of psychopathic traits, such as lower economic outcomes in adulthood (such as household income, Boccio & Beaver, 2018), victimization (Boccio & Beaver, 2021) or lower intelligence itself (DeLisi et al., 2010), especially for the self-centered impulsivity aspects (deficits in executive functioning) of psychopathy (Eisenbarth et al., 2018).Second, we revealed a small but still significant interaction effect between antisocial behavior and verbal intelligence, indicating that individuals with low verbal intelligence are more likely to get convicted for their antisocial behavior than individuals with higher verbal intelligence scores.This finding supports the notion that the success in the successful psychopath does not come from psychopathic traits themselves (see also Boccio & Beaver, 2018) but from separate moderating variables such as verbal intelligence as a protective factor against detection and apprehension.In the present study, antisocial behavior which did not count as criminal behavior was either not necessarily illegal or it was illegal but was not detected and officially sanctioned by the law enforcement authorities.Both kinds of behavior were rather shown by individuals with higher verbal intelligence who have been more capable to evade detection or conviction, respectively.
The moderating effect of verbal intelligence in the psychopathy-crime relationship might have the same underlying mechanism as the protective effect of verbal intelligence that can be observed in general for successful criminals.According to Boccio et al. (2018) verbal intelligence might represent the ability of being more successful in general and daily life, and the dark figure of crime strongly supports the idea of individuals being able to evade detection.Furthermore, intelligence might lead to successful criminal behavior in the sense that people with high verbal intelligence avoid specific types of crimes that are associated with higher arrest rates such as violent crimes (Paré et al., 2007).
Having an even more detailed look at the criminal-decision process, several authors presented theories that take into account the contributions of emotions, cognition, and situational factors in the decision whether a person will commit a crime (van Gelder & de Vries, 2014, 2016).In a study by Chatzimike Levidi et al. (2022), the authors found that higher levels on the honesty-humility dimension predicted an increased risk of apprehension and a higher anticipated as well as actual negative effect of this leading to a lower intention of offending.As psychopathic traits have been shown to go along with a relatively high sensitivity toward rewards and low sensitivity toward punishment (Blair et al., 2004), the criminal-decision making process might end in favor of the antisocial choice due to low anticipated and general negative affect.Only the cognitive ability of verbal intelligence as a moderator could inform the decision making process in a way that the favorable result of being successful (e.g., not being convicted) can be achieved and might compensate the lack of an (anticipated) emotional reaction in psychopathic individuals.
The current study indicates that not all psychopathic individuals are successful in evading detection of their crimes but only those individuals who are able to act more adaptive due to higher verbal intelligence.These findings support the first and the second hypotheses suggested by Benning et al. (2018) that higher psychopathic traits go along with higher antisocial behavior and the level of verbal intelligence-as moderator variable-helps to differentiate individuals in either the successful or unsuccessful group.
However, the results of the present study have to be considered in light of several limitations.First, as already mentioned, the assessment of all variables was solely based on self-report so there is a possible overestimation of effects due to shared measurement variance of the variables.In order to directly contrast antisocial behavior with its convicted official records, it was necessary to apply this method because other methods like direct observation or official records were not available for the present study.Taken together, biases in the direction of social desirability or even an imprecise memory were possible to occur in this study.Second, the sampling procedure was slightly different in the civil (mostly students) and the offender sample which limited the comparability of both subsamples.However, the sample of the current study included men, women, offenders, as well as non-offending participants, which led to higher variances in the variables of interest.
Consequently, future research should replicate this analysis with other measures of officially documented criminal behavior.Furthermore, since the variance of criminality could just be partly explained by the variables in the analysis, further predictors and moderator variables have to be considered in future studies.Benning et al. (2018) pointed out that age, executive functioning, socioeconomic status as well as physiology and neuroanatomical characteristics might play a role in protecting psychopathic individuals from getting in contact with law enforcement authorities.Furthermore, it is conceivable that specific crime areas have further relevant moderator variables: For example, Hawes et al. (2013) found in their meta-analysis that sexual deviance serves as an accelerating moderator variable since psychopathic individuals who are also highly sexually deviant exhibit higher risk to sexually reoffend than psychopathic individuals without sexually deviant interests.Additionally, internationally notorious individuals might illustrate this effect, as Ted Bundy as well as Edmund Kemper both assumed with high intelligence, high psychopathic traits, and high sexual deviance, driving them to become serial killers.Additionally, criminological variables like the probability of detection, the role, and functioning of the law enforcement authorities, and the willingness of victims to report crimes interfere and influence this relationship.Last but not least, future studies should investigate more deeply the situational criminal decision-making process with regard to psychopathic personality traits and cognitive abilities such as verbal intelligence, following an integrated model of criminal choice (van Gelder & de Vries, 2014, 2016).
Taken together, the present study adds knowledge to the nature of adaptive psychopathy.Since we found a high effect of psychopathic traits on general antisocial behavior, the investigation of adaptive psychopathy might still be useful in predicting maladaptive behaviors in social contexts.Thus, bullying or cheating on colleagues in companies (Board & Fritzon, 2005;Coyne & Thomas, 2008), or oppositional behaviors in school or family might be likely for children or adults with high psychopathic traits (Fanti & Kimonis, 2012) even though they are not committing serious crimes in the future.Furthermore, the findings of this study could be relevant for the prediction of recidivism in offenders because interaction effects between psychopathic traits and moderation variables can inform risk-related decision-making processes.The predictive power of psychopathic traits could be underestimated by considering them as a single and isolated risk factor for criminality without taking into account the presence of other (protective) factors like verbal intelligence.Therefore, assessing psychopathic traits and using them for purposes of criminal prognosis could be more precise in conjunction with other moderator variables.
Note.LL BCA and UL BCA = Lower level and upper level of the bias-corrected and accelerated bootstrapped confidence interval for α = .05;bootstrapping resamples N = 10,000.

Figure 3 .
Figure 3.A plot of conditional indirect effects of psychopathy to crime via ASB.Note.Indirect effects are displayed for IQ ± 2 SD, with confidence bands.The horizontal line denotes an indirect effect of zero.

Table 1 .
Characteristics of Both Subsamples of the Present Study (Total N = 305).

Table 2 .
Descriptive Statistics of Variables Included in the Analysis.

Table 3 .
Indirect Effects of Psychopathy on Crime via Antisocial Behavior at ±1 SD of the Moderator Intelligence (N = 305).