Abstract
Previous work has demonstrated mostly null to small associations between boldness and externalizing behaviors associated with psychopathy. The present study replicated this finding and examined an alternate manner in which boldness may be relevant to the construct of psychopathy, namely by enhancing psychopathic individuals’ capacity for interpersonal manipulation. Using data collected from a Mechanical Turk sample, we measured persuasiveness using ratings of video-recorded product pitches and examined the relations between psychopathic and general traits and persuasiveness across weaker (improvised) and stronger (scripted) conditions. Boldness exhibited a small, positive relation with perceived persuasiveness in the improvised condition only; conversely, psychopathic traits related to antagonism/meanness and disinhibition exhibited small negative associations with persuasiveness and trust in the scripted condition. The results suggest that boldness may help individuals persuade others so as to achieve desired outcomes, although the effect was quite small in nature, whereas psychopathic individuals who are not high on boldness may need to use other tactics (e.g., intimidation, coercion) to manipulate others.
Psychopathy is one of the most studied and well-validated personality disorders; yet considerable debate remains regarding its underlying structure and the centrality of certain elements within it (e.g., Lilienfeld et al., 2012; Lilienfeld, Watts, Francis Smith, Berg, & Latzman, 2015; Lynam & Miller, 2012; Miller & Lynam, 2012). One of the most debated aspects of psychopathy is boldness (Patrick, Fowles, & Krueger, 2009), also known as fearless dominance in some inventories (e.g., Psychopathic Personality Inventory–Revised; Lilienfeld & Widows, 2005), which is characterized by resilience to stress, fearlessness, and social influence. Boldness’ relevance to psychopathy has been debated in view of its null to small correlations with maladaptive outcomes historically linked to psychopathy (e.g., antisocial behavior, aggression, substance use), as well as its divergence from measures of psychopathy that are based on the Psychopathy Checklist–Revised (PCL-R; Hare, 2003; Miller & Lynam, 2012). However, boldness appears in many non–PCL-R measures of psychopathy (Lilienfeld et al., 2016) and classic descriptions of psychopathy in which psychopathic individuals are described as being callous, egocentric, impulsive, and irresponsible while also appearing well mannered, charming, and agentic (e.g., Cleckley, 1941; Crego & Widiger, 2016; Miller, Lynam, Widiger, & Leukefeld, 2001).
Those who dispute boldness’ relevance to psychopathy regard evidence of null to weak relations with externalizing behaviors (e.g., antisocial behavior [ASB], aggression) as problematic on the basis of the premise that antisociality is the critical criterion that defines psychopathy and is responsible for the tremendous interest in this construct across a number of scientific disciplines (DeLisi, 2009; Hare & Neumann, 2010; Karpman, 1948; Lykken, 1995; Lynam & Miller, 2012; Miller & Lynam, 2012). Others disagree, however, arguing that ASB should not be considered central to psychopathy but, rather, represents an irregular behavioral manifestation arising from it (Lilienfeld, 1998; Skeem & Cooke, 2010), a view that is consistent with certain prominent clinical theorists for whom ASB may have been of secondary importance. For instance, Cleckley (1941) focused on ASB that was “inadequately motivated,” although “long histories of ASB appeared in all 15 of Cleckley’s cases” (p. 342; Lynam & Miller, 2012).
Although boldness may be only weakly linked to overt antisocial behavior, it may exert effects via more subtle forms of interpersonal behavior. In addition to the explicit and overt antisocial behaviors typically associated with psychopathy (e.g., Hare & Neumann, 2008), prominent conceptualizations suggest that psychopathic traits are associated with manipulation—the willingness and ability to deceive others for personal gain (Cleckley, 1941). Most models and measures of psychopathy have explicit representations of such content. Facet 1 of the PCL-R (i.e., Interpersonal; Hare, 2003) includes items related to manipulativeness and superficial charm, as do self-report measures derived from this assessment (e.g., Self-Report Psychopathy Scale; Paulhus, Neumann, & Hare, 2009). Other measures such as the Psychopathic Personality Inventory–Revised (Lilienfeld & Widows, 2005) and the Elemental Psychopathy Assessment (EPA; Lynam et al., 2011) also include relevant scales titled “Machiavellian Egocentricity” and “Manipulation,” respectively. Most of these scales load on a domain that is both theoretically and empirically linked with trait disagreeableness/antagonism (Lynam & Miller, 2015). There is also some evidence to suggest that boldness may be linked to charm and manipulativeness, given small to moderate relations with the PCL-R Interpersonal facet (Venables, Hall, & Patrick, 2014); weak to moderate relations with manipulativeness, as indexed by clinical self-report and interview measures (e.g., Strickland, Drislane, Lucy, Krueger, & Patrick, 2013); and small to moderate inverse relations with five-factor model Straightforwardness (e.g., Miller, Lamkin, Maples-Keller, & Lynam, 2016).
Boldness and Persuasiveness
Boldness-related traits may be relevant to interpersonal manipulativeness if such traits make psychopathic individuals more persuasive, allowing such individuals to penetrate deeper into social networks before their more clearly aversive traits (i.e., antagonism, disinhibition) are detected. One study found that boldness exhibited a curvilinear relation with sales performance, such that increases in boldness until –.39 standard deviations were associated with increased sales performance, after which point, boldness was associated with decreased performance (Titze, Blickle, & Wihler, 2017). Although there is little research available on the relation of boldness and persuasiveness per se, there is some research available, particularly within the industrial-organizational psychology literature, linking the main personality correlates of boldness—extraversion and (low) neuroticism (Miller & Lynam, 2012). For instance, extraversion has a generally positive relation to sales performance, sales figures, and supervisor ratings of sales success (Barrick & Mount, 1991; Hurtz & Donovan, 2000; Vinchur, Schippmann, Switzer, & Roth, 1998).
It bears noting that persuasiveness does not in itself imply manipulativeness and/or antisociality in that individuals can persuade others to behave in more prosocial ways (e.g., using famous individuals to encourage others to vote or get vaccinated). If boldness evinces a substantive relation to persuasiveness, it is possible that bold individuals use persuasiveness in the service of prosocial motives in their interpersonal relationships (e.g., persuading a loved one to make a positive life change), given findings linking boldness/fearless dominance to some prosocial qualities including empathy, emotion recognition, and sociability (Gatner, Douglas, & Hart, 2016) as well as altruism under conditions of physical or social risk (e.g., heroism; Smith, Lilienfeld, Coffey, & Dabbs, 2013).
Examining Persuasiveness in Weak and Strong Situations
To study an individual difference like persuasiveness, it is important to consider the extent to which the situation may affect the degree to which the trait is observed. Theory and research on the expression of individual differences in weak versus strong situations (Mischel, 1977; Snyder & Ickes, 1985) are useful for designing studies that optimize the likelihood of observing individual differences. Situations that are relatively ambiguous and fail to specify behavioral norms are considered weak because no particular response is dictated by the situation. Conversely, situations that provide individuals with salient cues for a circumscribed set of behaviors are considered strong (e.g., how to behave during a standardized test). Previous research has suggested that traits are typically more predictive of behavior in weak situations (e.g., Meyer, Dalal, & Bonaccio, 2009) because they are more conducive to variable expression of dispositional traits (e.g., Monson, Hesley, & Chernick, 1982).
Present Study
The present study examined the associations among psychopathic traits, general traits, and persuasiveness on the basis of ratings of video-recorded product pitches, in which participants were incentivized to persuade raters to buy a specific product. We examined this association across two conditions: a weak condition, in which participants were asked to video-record themselves giving an improvised pitch for a new smartphone, and a strong condition, in which participants were asked to video-record themselves giving a scripted pitch of a photo editing smartphone app. We expected that the improvised condition would yield stronger effect sizes for psychopathy than the scripted condition.
Our first aim was to replicate previous findings on the relations among psychopathic traits and self-reports of externalizing behaviors; we expected traits related to antagonism/meanness and disinhibition to show stronger associations with externalizing outcomes than traits related to boldness. Second, we examined relations between psychopathic and general personality traits and ratings of persuasiveness across improvised and scripted conditions. As a third aim, we examined relations between psychopathic traits and ratings of trust in an exploratory fashion to investigate whether trust building may be another manner in which boldness aids psychopathic individuals. 1
Method
Participants
A power analysis indicated that a sample of at least 300 would be sufficiently powered (at .95) to detect correlations as small as .20 using a p ≤ .01 threshold for significance. Six hundred thirty-three participants were recruited from Amazon’s Mechanical Turk, of which 139 were excluded because of invariant responding to more than 85% of EPA items or Big Five Inventory (BFI) items and/or invalid responding based on EPA validity scales. One hundred eighty-seven participants were excluded because of failure to submit at least one video, leaving a final sample of 307 (165 women and 145 men; age: M = 31.39 years, SD = 9.19; 74% White, 11% Black, 6% Asian, and 10% Hispanic). Independent-samples t-test analyses were conducted to test for differences between validly responding participants who did and did not submit videos. Participants who submitted videos exhibited lower levels of antagonism- and disinhibition-related traits, Triarchic Psychopathy Measure (TriPM) meanness, d = −0.64; TriPM disinhibition, d = −0.56; EPA antagonism, d = −0.54; EPA disinhibition, d = −0.43; EPA narcissism, d = −0.26; BFI agreeableness, d = 0.38; BFI conscientiousness, d = 0.23; BFI openness, d = −0.60. It is important that the groups did not differ on traits related to boldness (i.e., boldness, emotional stability, extraversion, neuroticism), except for EPA narcissism. Institutional review board approval was obtained for all aspects of the study.
Video-recording persuasive pitch procedure
Participants were asked to video-record themselves giving two 30-s to 2-min presentations of a consumer product. In the first condition (improvised; i.e., weak condition), participants were asked to improvise a “persuasive pitch for a new smartphone” using elements from a list of eight pros (e.g., “good low light performance”) and six cons (e.g., “heavier than chief competitor”). Participants were asked to use at least two cons (for the list, see the Supplemental Material available online). In the second condition (scripted; i.e., strong condition), participants were asked to recite a script detailing the attributes of a new mobile app (for the script, see the Supplemental Material). Participants were directed to read from the script to reduce variation in persuasiveness due to possible effects of reciting from memory. To incentivize performance, participants were told that the five most persuasive participants in each condition would receive a bonus of $20.00. Participants uploaded their videos to a digital storage database using a file transfer link.
Predictor measures
Triarchic Psychopathy Measure
The TriPM (Patrick, 2010) is a 58-item self-report measure of psychopathy composed of three scales: boldness (19 items; α = .88), meanness (19 items; α = .88), and disinhibition (20 items; α = .87). Scale intercorrelations are reported in Table S1 in the Supplemental Material. TriPM data were available for 307 participants.
Elemental Psychopathy Assessment–Short Form
The EPA–Short Form (EPA-SF) is an 88-item version of the EPA (Lynam et al., 2013). It assesses 18 subscales of psychopathy as well as two validity scales (Infrequency and Virtue). The EPA can be aggregated into four higher order factors (Few, Miller, & Lynam, 2013): antagonism (α = .90), emotional stability (α = .90), disinhibition (α = .90), and narcissism (α = .76). Intercorrelations are reported in Table S1 in the Supplemental Material. EPA data were available for 306 participants.
Big Five Inventory
The Big Five Inventory (BFI; John, Donahue, & Kentle, 1991) is a 44-item scale that measures each Big Five domain of personality. Alphas ranged from .81 to .89. Correlations are reported in Table S1 in the Supplemental Material. BFI data were available for 307 participants.
International Cognitive Ability Resource
The International Cognitive Ability Resource (ICAR; Condon & Revelle, 2014) is a measure of cognitive ability. Eight items from the ICAR Sample Test were used to assess cognitive ability, including items related to Verbal Reasoning (four items) and Letter Number Sequence (four items; α = .60). ICAR data were available for 303 participants.
Criterion measures
Persuasiveness and trust ratings
Six research assistants served as raters and rated one video from each participant on seven bipolar characteristics (i.e., unpersuasive-persuasive, dishonest-honest, unintelligent-intelligent, immoral-moral, phony-genuine, untrustworthy-trustworthy, doesn’t care–cares about me) using a 7-point Likert-type scale. Raters were blind to participants’ psychopathy scores as well as the purpose of the study. Items were adapted from McCroskey and Teven’s (1999) scale measuring goodwill. Exploratory factor analysis using the parallel analysis method of Horn (1965) and the minimum average partial method of Velicer (1976) of the latter six items suggested a one-factor solution that accounted for 94% of the variance in the improvised condition and 95% of the variance in the scripted condition. Thus, the latter six characteristics were averaged to create a composite for each video condition to capture the perceived trustworthiness of each participant. Raters judged only one video from each participant (i.e., scripted or unscripted) to ensure that ratings were not biased by exposure to the other video; three raters rated each video. Interrater reliability (IRR) analyses were conducted (Hallgren, 2012). IRR intraclass correlations (ICCs) were calculated, specifying a one-way model, absolute agreement, and an average unit of analysis, using the irr package in the R programming environment (Gamer, Lemon, & Singh, 2012). IRR results for persuasiveness indicated ICCs of .56 (for improvised video condition) and .64 (for scripted video condition). Aggregation of the Trust component ratings resulted in a linear-composite reliability of .86 (for improvised video condition) and .87 (for scripted video condition; Nunnally, 1978).
Attractiveness ratings
Three separate undergraduate research assistants rated the attractiveness of one still frame picture of each participant using a 10-point scale ranging from very unattractive (1) to very attractive (10). 2 Each still frame picture was edited to conceal distinctive features of the environment. IRR ICCs were calculated, specifying a two-way model, absolute agreement, and an average unit of analysis, using the irr package in the R programming environment (IRR ICC = .78; Gamer et al., 2012).
Crime and Analogous Behavior Scale–BRIEF
The Crime and Analogous Behavior Scale–BRIEF (CAB-BRIEF; Miller & Lynam, 2003) is a 25-item self-report inventory that assesses a variety of externalizing behaviors. An antisocial behavior count was created by giving participants a 1 for every antisocial act they endorsed using (10 items; M = 1.52, SD = 1.14). An intimate partner violence variety count was created by giving participants a 1 for every act of interpersonal violence they endorsed using (six items; M = 0.89, SD = 1.51). A substance use variety count was created by giving participants a 1 for every substance they endorsed using (five items; M = 1.81, SD = 1.35).
Results
Examining relations between psychopathic traits and externalizing behavior
Given the number of tests of statistical significance conducted, an alpha level of p ≤ .01 was set for all analyses. The correlations between psychopathic and general traits, as well as gender, IQ scores, and attractiveness, with ASB, intimate partner violence, and substance use, were examined (see Table 1). 3 Gender was significantly related only to intimate partner violence, with women reporting higher levels. IQ scores and attractiveness ratings were not related to any of the three externalizing behaviors. Five of seven psychopathic traits were significantly associated with ASB (exceptions: TriPM boldness, EPA emotional stability), as were Big Five agreeableness and conscientiousness. In relation to intimate partner violence, only TriPM and EPA disinhibition as well as EPA narcissism were significantly positively related. Finally, with regard to substance use, only TriPM and EPA disinhibition as well as openness were significantly related.
Bivariate Relations Between Trait and Externalizing Behaviors
Male = 1; female = 0.
p ≤ .01.
Correlations between psychopathic and general traits and persuasiveness and trust
Improvised video condition
Ratings of persuasiveness were significantly positively related to IQ, attractiveness, and ratings of trust (see Table 2). 4 From the perspective of psychopathic and general traits, persuasiveness was significantly predicted only by boldness (r = .17) and extraversion (r = .16); both effects were small in nature. Ratings of trust were positively related to ratings of persuasiveness (r = .60) but not significantly related to gender, IQ, attractiveness, or any of the psychopathic or general personality traits.
Bivariate Relations Between Traits and Persuasiveness and Trust
Note: IQ = International Cognitive Ability Resource score.
Male = 1; female = 0.
p < .01.
Scripted video condition
Persuasiveness was negatively correlated with male gender and video length and positively correlated with IQ scores and ratings of trust. Two of three TriPM scales (i.e., meanness, disinhibition) and two of four EPA factors (i.e., antagonism, disinhibition) were significantly negatively related to ratings of persuasiveness in the scripted condition. None of the general personality traits were significantly related to ratings of persuasiveness. Ratings of trust were positively related to ratings of persuasiveness (r = .70) and IQ scores and negatively related to male gender and video length. Ratings of trust were significantly negatively related to two of three TriPM scales (i.e., meanness, disinhibition) and two of four EPA scales (i.e., antagonism, disinhibition); all effect sizes were small in nature. None of the general traits was significantly related to ratings of trustworthiness.
Discussion
There is substantial debate surrounding the relevance of boldness to the broader psychopathy construct (e.g., Lilienfeld et al., 2012; Lynam & Miller, 2012; Miller & Lynam, 2012) due, in part, to its null to small relations with externalizing behaviors, as replicated here. Given this, we tested an alternate manner in which boldness may be relevant to outcomes typically associated with psychopathy, namely, by enhancing psychopathic individuals’ capacity for interpersonal manipulation. We investigated this question using a unique methodology with a number of strengths. First, we measured persuasiveness and trust using ratings of video-recorded product pitches. Second, we examined the relation between boldness and persuasiveness across weaker and stronger contexts. Third, we collected a sample that was large enough to detect reasonably small effect sizes and incentivized performance by compensating top performers.
Our analysis yielded two key findings, each with implications for the conceptualization of psychopathy. First, boldness bore a small association with perceived persuasiveness in the weaker context (i.e., improvised condition), suggesting that boldness may enhance an individual’s ability to persuade (or manipulate) others to make decisions that may not be in their interest or persuade others with prosocial motives (e.g., to behave kindly or charitably). The relation for TriPM boldness was found only in the improvised (weak) condition. Extraversion exhibited a similarly sized association with persuasiveness, consistent with the robust correlation between the two constructs (r = .62 in the current sample). It is possible that both boldness and extraversion are connected to persuasiveness via a desire to obtain social and monetary rewards, greater ease at social interactions, and/or greater practice or skill in social interactions and persuasion. Boldness’ relation with persuasiveness may support the claims of contemporary scholars that boldness underlies Cleckley’s “mask” of sanity (e.g., Edens, Poythress, Lilienfeld, Patrick, & Test, 2008), although the null linear effect on trust as well as null moderational results (see Note 1 and the Supplemental Material) suggest that boldness does not underlie trustworthiness or buffer the effects of meanness or disinhibition on perceived persuasiveness or trust. These results may provide some tentative evidence against the mask hypothesis, but more research is needed under conditions designed to elicit trustworthiness rather than persuasiveness more narrowly and where power is optimized to interpret a null finding. The present findings may also provide one possible avenue for understanding the successful (or noncriminal) psychopathic individual. Boldness has been linked to effective leadership, at least in the short term (e.g., Boddy, Ladyshewsky, & Galvin, 2010), and perceptions of good communication and strategic thinking in the workplace (Babiak, Neumann, & Hare, 2010). More research is needed to determine to what degree some of these adaptive characteristics may be mediated by persuasiveness.
Second, psychopathic traits related to antagonism/meanness and disinhibition exhibited small negative associations with persuasiveness and trust in the scripted condition, where we generally expected fewer effects, suggesting that being disinhibited and antagonistic may actually impair individuals from manipulating others in a socially appropriate way or gaining access to social networks. These findings suggest that some psychopathic traits may be related to a different style of interpersonal manipulation from that generally considered. Psychopathic individuals’ ability to manipulate others may not always involve the use of charm, ingratiation, or trust inducement in a socially appropriate way but, rather, may involve the concealment of intentions or the use of coercion and intimidation. antagonism/disinhibition may be associated with some styles of speech underlying persuasiveness and trust but not with others (e.g., not associated with persuasiveness via diction and syntax, negatively associated with persuasiveness via tone/intonation).
Limitations and Future Directions
Despite the study’s many strengths, some limitations must be acknowledged. First, recruiting from Mechanical Turk was associated with some challenges because of invalid responding and failures to upload videos; our final sample exhibited lower mean levels of antagonism and disinhibition than non-video-submitting participants, which may have resulted in some range restriction and attenuation of effect sizes for those traits. The reasons for attrition are not clear, but we suspect that it was due, in part, to the novelty of our request for video-recording (vs. the more typical task of completing questionnaires), the onerousness of complying with technologically detailed instructions, and the atypicality of our request to expose the participants’ faces and bodies, which may have deterred some participants from following through with the video task. It is important, however, that the sample was not lower on boldness-related traits, which was the construct of greatest interest in this study. Future studies should examine these questions in samples that have higher mean levels of psychopathic traits and greater diversity in terms of age, race/ethnicity, and socioeconomic status. In addition, a substantial proportion of participants (~30%) was excluded because of invariant or invalid responding. Although Mechanical Turk workers are generally thought to provide data of equal or higher quality than many other samples (e.g., Chandler & Shapiro, 2016; Miller, Crowe, Weiss, Maples-Keller, & Lynam, 2017), in this case there was a significant minority of participants whose data were of questionable validity and thus were excluded. Second, our observations of psychopathic traits were derived from self-report data; future studies should compare self-report data on these scales with data collected from other modalities (e.g., informant reports, interviews), although self- and informant reports of psychopathy in research settings tend to converge at reasonably high levels (e.g., Miller, Jones, & Lynam, 2011). Third, we were not powered in this study to detect correlations below .20 with a reasonable level of confidence, although we wonder about the importance of trait-behavior correlations smaller than .20, which would indicate that they shared less than 4% of their variance. Fourth, interrater reliability was not optimal in the case of one of our primary outcomes of interest (i.e., persuasiveness) in the improvised condition. According to Cicchetti’s (1994) benchmarks, the associated ICC was considered only fair, although persuasiveness in the scripted condition was considered good. This lower reliability likely attenuated the size of the relations reported here. To counter this, we reported these associations disattenuated for unreliability in Table S5 in the Supplemental Material.
In closing, the current study replicated previous work demonstrating that boldness is not a robust correlate of the externalizing behaviors often associated with psychopathy. However, boldness did manifest a small positive relation with perceived persuasiveness in the nonscripted condition, suggesting that boldness may allow psychopathic individuals to manipulate others, although it is important to note that this manipulation may be found in the service of both antisocial and prosocial goals.
Supplemental Material
WeissSupplementalMaterials – Supplemental material for Psychopathy and Ratings of Persuasiveness: Examining Their Relations in Weaker and Stronger Contexts
Supplemental material, WeissSupplementalMaterials for Psychopathy and Ratings of Persuasiveness: Examining Their Relations in Weaker and Stronger Contexts by Brandon M. Weiss, Donald R. Lynam and Joshua D. Miller in Clinical Psychological Science
Supplemental Material
Weiss_Open_Practices_Disclosure – Supplemental material for Psychopathy and Ratings of Persuasiveness: Examining Their Relations in Weaker and Stronger Contexts
Supplemental material, Weiss_Open_Practices_Disclosure for Psychopathy and Ratings of Persuasiveness: Examining Their Relations in Weaker and Stronger Contexts by Brandon M. Weiss, Donald R. Lynam and Joshua D. Miller in Clinical Psychological Science
Footnotes
Action Editor
Scott O. Lilienfeld served as action editor for this article.
Author Contributions
All the authors contributed to the study design. Data collection and data analysis were performed by B. M. Weiss under the supervision of J. D. Miller and D. R. Lynam. B. M. Weiss drafted the manuscript, and J. D. Miller and D. R. Lynam provided critical revisions. All the authors approved the final manuscript for submission.
Declaration of Conflicting Interests
The author(s) declared that there were no conflicts of interest with respect to the authorship or the publication of this article.
Open Practices
All materials have been made publicly available via the Open Science Framework and can be accessed at https://osf.io/gk3zb/. The complete Open Practices Disclosure for this article can be found at https://journals-sagepub-com.web.bisu.edu.cn/doi/suppl/10.1177/2167702618783733. This article has received the badge for Open Materials. More information about the Open Practices badges can be found at
.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
