This is the author version published as:
This is the accepted version of this article. To be published as :
This is the author version published as:
QUT Digital Repository:
Grieve, Rachel and Mahar, Douglas P. (2010) The emotional
manipulation-psychopathy nexus : relationships with emotional
intelligence, alexithymia, and ethical position. Personality and Individual
Differences, 48(8 ). pp. 945-950.
Copyright 2010 Elsevier
The Emotional Manipulation-Psychopathy Nexus: Relationships with Emotional
Intelligence, Alexithymia and Ethical Position
Queensland University of Technology
Corresponding Author: Rachel Grieve
School of Psychology and Counselling
Faculty of Health
Queensland University of Technology
B Wing, O Block, Kelvin Grove Campus
Victoria Park Road, Kelvin Grove,
Queensland AUSTRALIA 4059
This research examined for the first time the relationship between emotional manipulation,
emotional intelligence, and primary and secondary psychopathy. As predicted, in Study 1
(N=73), emotional manipulation was related to both primary and secondary psychopathy.
Only secondary psychopathy was related to perceived poor emotional skills. Secondary
psychopathy was also related to emotional concealment. Emotional intelligence was
negatively related to perceived poor emotional skills, emotional concealment, and primary
and secondary psychopathy. In Study 2 (N=275), two additional variables were included:
alexithymia and ethical position. It was found that for males, primary psychopathy and
emotional intelligence predicted emotional manipulation, while for females emotional
intelligence acted as a suppressor, and ethical idealism and secondary psychopathy were
additional predictors. For males, emotional intelligence and alexithymia were related to
perceived poor emotional skills, while for females emotional intelligence, but not
alexithymia, predicted perceived poor emotional skills, with ethical idealism acting as a
suppressor. For both males and females, alexithymia predicted emotional concealment. These
findings suggest that the mechanisms behind the emotional manipulation-psychopathy
relationship differ as a function of gender. Examining the different aspects of emotional
manipulation as separate but related constructs may enhance understanding of the construct
of emotional manipulation.
The Emotional Manipulation-Psychopathy Nexus: Relationships with Emotional Intelligence,
Alexithymia and Ethical Position.
Austin, Farrelly, Black, and Moore (2007) operationalised emotional manipulation as
the “dark side” of emotional intelligence (EI). Emotional manipulation is the capability of
individuals to manipulate the emotions of others within a self-serving framework. This
research explored emotional manipulation by investigating relationships with EI and
psychopathy, and then extending the examination to include differential effects of ethical
position and alexithymia as a function of gender.
Salovey and Mayer (1990) originally described EI as including the appraisal and
expression of emotion in self and in others, the regulation of emotion, and the utilisation of
emotion. Subsequently Mayer and Salovey (1997) refined their definition to include specific
facilitative abilities. Individuals’ responses to EI test items appear to reflect genuine
emotional facilitation, with EI unrelated to socially desirable responding (Grieve & Mahar, in
However, conceptualisations of EI are generally positive, as well as facilitative (Austin
et al., 2007). Austin et al. extended existing notions of EI, arguing that the ability to use and
manage emotions could also be used in negative and malicious contexts. The authors
presented a scale tapping emotional manipulativeness. Three factors were identified:
emotional manipulation, perceived poor emotional skills, and emotional concealment. Austin
et al. found that the manipulative trait of Machiavellianism accounted for 16% of the
variance in the emotional manipulation factor.
Psychopathy has also been shown to be associated with manipulative behaviours
(Neumann, Hare, & Newman, 2007). Originally viewed as a homogenous construct, more
recent approaches suggest psychopathy can be differentiated into two related factors; primary
and secondary (e.g. Del Gaizo & Falkenbach, 2008; Levenson, Kiehl, & Fitzpatrick, 1995;
Skeem, Poythress, Edens, Lilienfeld, & Cale, 2003). Primary psychopathy is characterised by
malevolent, manipulative, callous, deceptive, and remorseless behaviour, with reduced affect.
Secondary psychopathy is characterised by impulsivity, anxiety, and antisocial behaviour.
The Current Research and Hypotheses
Two studies were conducted. The first study aimed to empirically investigate
conceptual similarities between emotional manipulation and psychopathy. In order to
examine any associations clearly, the two factors of psychopathy and the three identified
factors of emotional manipulation were considered separately. The second study aimed to
clarify and explore the nature of the emotional manipulation-psychopathy relationship found
in the preliminary study.
Psychopathy as an overarching construct is associated with manipulative and deceptive
behaviour (Neumann et al., 2007). Therefore, it was hypothesised that the general emotional
manipulation factor would be positively related to primary and secondary psychopathy.
Further, as primary psychopathy is particularly associated with manipulation (Levenson et
al., 1995), it was hypothesised that the relationship between emotional manipulation and
primary psychopathy would be stronger than the relationship between emotional
manipulation and secondary psychopathy.
Secondary psychopathy, rather than primary psychopathy, is associated with anxiety
(Levenson et al., 1995; Skeem et al., 2003). In addition, secondary psychopaths have poorer
emotion perception than primary psychopaths (Del Gaizo & Falkenback, 2008). Thus, it was
also hypothesised that the poor emotional skills factor, reflecting rumination regarding
emotional skills, would be correlated with secondary psychopathy, but not with primary
As primary psychopaths are known for their unemotional demeanour (e.g. Levenson et
al., 1995), they may not need to conceal their emotions in order to manipulate others’
behaviour. However, secondary psychopathy is associated with anxiety, which may need to
be concealed for successful manipulation to occur. Therefore, it was hypothesised that only
secondary psychopathy would be related to the emotional concealment factor.
Using Salovey and Meyer’s (1990) operationalisation of EI as a theoretical base, and
derived from Austin et al.’s (2007) findings, two additional hypotheses were generated.
While EI includes the ability to use emotion, and as current operationalisation of EI relies
largely on emotional facilitation in positive contexts, it was hypothesised that emotional
manipulation would be uncorrelated with EI. Further, as EI includes the ability to express and
regulate emotion, it was hypothesised that EI would be negatively correlated with both poor
emotional skills and the ability to conceal emotions. Finally, in line with previous research
(Grieve & Mahar, in press) it was hypothesised that both primary and secondary psychopathy
would be negatively correlated with EI.
Seventy-three Australian undergraduates (58 female, 15 male) with a mean age of
23.96 years (SD= 10.03) participated in return for course credit.
Design and Procedure
Ethical clearance for this and Study 2 was obtained from the university’s ethics
committee. A correlational design was used. The variables were emotional manipulation,
perceived poor emotional skills, emotional concealment, EI, and primary and secondary
psychopathy. After giving informed written consent, participants completed the
questionnaires in random order.
The emotional manipulation variables (emotional manipulation, perceived poor
emotional skills, and emotional concealment) were measured using subscales comprised of
items identified as loading clearly onto factors in Austin et al.’s (2007) investigation, and
showing adequate reliability (Cronbach’s alphas of .88, .66, and .77 respectively). Items were
measured on a five-point Likert scale, with the anchors 1= strongly disagree, 5= strongly
Emotional manipulation. 10 items tapping the tendency to use emotional manipulation
were used. A sample item is I know how to play two people off against each other.
Poor emotional skills. Four items tapping self perceptions of confidence in the ability to
manipulate emotions were used. A sample item is I am not very good at motivating people.
Emotional concealment. Four items tapping the tendency to hide emotional reactions
from others were used. A sample item is When someone has made me upset or angry, I often
conceal my feelings.
Emotional Intelligence. Schutte et al.’s (1998) 33-item self-report of EI was used.
Responses are given on a five-point Likert scale. This measure taps three constructs from
Salovey and Mayer’s (1990) model: emotional appraisal, emotional regulation and emotional
utilisation. The scale has a Cronbach’s alpha of .90 (Schutte et al., 1998).
Psychopathy. Psychopathy was measured using Levenson et al.’s (1995) Primary (16
items) and Secondary (10 items) Psychopathy Scale. Self-reported responses are given in a
four-point Likert format. Cronbach’s alpha for the two scales have previously been shown to
be .82 and .62 respectively.
Results and Discussion
Males scored higher than females on primary psychopathy, however this difference was
not significant (p=.19). This trend was in line with previous studies (Grieve & Mahar, in
press; Levenson et al., 1995). As there were no significant differences between genders on
any other variables (all ps>.67), and as this study was preliminary in nature, all data was
analysed together. Descriptive statistics and bivariate correlations are presented in Table 1.
Scores and distributions on all measures were similar to earlier research (Austin et al., 2007;
Levenson et al., 1995; Schutte et al., 1998). Reliability was good for all tests except the
perceived poor emotional skills subscale.
As predicted, the emotion manipulation subscale was significantly and positively
related to primary and secondary psychopathy. Further, the relationship between emotional
manipulation and primary psychopathy was particularly strong, accounting for 28.1% of
variance, compared to 12.3% of variance in secondary psychopathy. This reflects the
manipulative and malevolent nature of psychopathy (Neumann et al., 2007). However,
importantly, these results suggest that manipulation and psychopathy are not synonymous.
This adds weight to Austin et al.’s (2007) conceptualisation of emotional manipulation as a
The hypothesis that secondary psychopathy, but not primary psychopathy, would be
related to perceived poor emotional skills was supported. This finding aligns with previous
research that secondary psychopaths experience more anxiety (Levenson et al., 1995; Skeem
et al., 2003), and are less skilled at emotion perception (Del Gaizo & Falkenback, 2008), than
primary psychopaths. Perhaps what individuals with secondary psychopathic traits are
actually reporting here is the lack of prosocial skills that are a corollary of secondary
psychopathy. These individuals may be anxious about and aware of their lack of
interpersonal skill. Another possibility is that individuals with primary psychopathic traits are
responding to these particular items strategically. For example, Levenson et al. suggest that
primary psychopaths may see manipulative ability as positive; perhaps they are therefore
unwilling to admit any lack of manipulative skill. Note that the low Cronbach’s alpha for the
perceived poor emotional skills subscale may suggest that these items are somewhat
inconsistent; however an alternative explanation is variations in people’s perceptions of
themselves when completing these items.
As hypothesised, secondary psychopathy was positively related to emotional
concealment, while primary psychopathy was not. There are several possible explanations for
this finding: perhaps primary psychopaths do not experience the same levels of emotion as
secondary psychopaths, or their “natural” demeanor is such that they do not need to
consciously or effortfully conceal their emotions. However, another consideration is that
individuals with primary psychopathic traits are experiencing some degree of alexithymia,
and as a result the concept of emotional concealment is arbitrary to their experience.
The hypotheses that EI would be negatively related to poor emotional skills and
emotional concealment were also supported, accounting for 37.2% and 12.9% of the variance
respectively. In the measure used in the current study, EI represents the ability to identify,
regulate, and utilise emotion (Schutte et al., 1998). Therefore, it is unsurprising that EI was
negatively related to poor emotional skills and emotional concealment. In line with Grieve
and Mahar’s (in press) findings, EI was also negatively related to both primary and
secondary psychopathy. However, only the relationship with secondary psychopathy was
While the preliminary study provides insight into the emotional manipulation-
psychopathy relationship, several questions were unanswered regarding the emotional
manipulation construct. For example, current operationalisations of EI are generally positive
and facilitative (Austin et al., 2007), while the emotional manipulation construct represents a
negative and malicious kind of EI. While in line with expectations, the small bivariate
relationship between EI and the emotional manipulation factor does not appear to reflect the
underlying conceptual similarities between the two constructs, such as the regulation and
management of emotions in order to obtain a particular outcome.
In addition, it is unclear whether someone with the ability to manipulate will actually
choose to engage in manipulative tactics: principles and values may influence an individual’s
use of emotional manipulation. This issue parallels other themes within the EI literature,
where increasingly trait EI is seen as typical performance EI, while ability EI is seen as
maximal performance EI (e.g. Freudenthaler, Neubauer, & Haller, 2008). Similarly,
emotional manipulation void of ethical redress may represent maximal, “dark” EI, while the
measurement of ethically regulated emotional manipulation may represent an individual’s
typical behaviour. This may be particularly relevant in the current context as individuals with
psychopathic traits are known to show less moral concern (Glenn, Iyer, Graham, Koleva, &
Further, the results of the first study suggest that assessing the effects of alexithymia on
emotional manipulation factors may help clarify the role of psychopathy. In addition, the first
study’s modest sample size precluded the investigation of any differences in the emotional
manipulation-psychopathy relationship as a function of gender.
A second study was therefore conducted in order to examine possible effects of ethical
reasoning and alexithymia on the relationships between psychopathic traits, EI, and the three
emotional manipulation factors, using a larger sample and regression techniques. Moreover,
the larger sample allowed relationships between variables to be examined as a function of
Three hypotheses were generated. Firstly, it was hypothesised that primary and
secondary psychopathy in combination with alexithymia and ethical position, would predict
emotional manipulation. Secondly, it was hypothesised that EI, secondary psychopathy and
alexithymia would predict perceived poor emotional skills. Finally, it was hypothesised that
alexithymia would predict emotional concealment. No specific hypotheses were made
regarding gender differences. Instead, as this research was exploratory in nature and in order
to maintain parsimony, the data was simply analysed separately for males and females to
allow identification of gender effects.
Two hundred and seventy-five Australian undergraduates (187 female, 88 male) with a
mean age of 23.53 years (SD= 8.92) participated, some of whom received course credit.
Eighty-three percent of participants reported English as their first language, and 96% were
either currently or previously in paid employment (Mean duration= 6.41 years, SD=7.02). On
average, participants had completed 13.64 years (SD= 1.94) of education.
Design and Procedure
A correlational design was used. The predictor variables were EI, idealistic ethical
reasoning, relativistic ethical reasoning, primary psychopathy, secondary psychopathy, and
alexithymia. Outcome variables were emotional manipulation, perceived poor emotional
skills, and emotional concealment. Ethical reasoning was not included as a predictor for
emotional concealment as it seemed unlikely that any relationship would exist between the
After giving informed written consent, participants completed the questionnaires in
random order. The ethical reasoning questionnaire was given last, in order to minimise any
impact of value laden questions on responses to the other questionnaires
For consistency, the measures used to operationalise EI, psychopathy, and emotional
manipulation were the same as in the first study.
Alexithymia was assessed using the 20 item Toronto Alexithymia Scale (Bagby,
Parker, & Taylor, 1994). Bagby et al. report Cronbach’s alpha for the scale at .80 -.83 over
three samples. Measured on a five-point Likert scale (anchors: 1=strongly disagree,
5=strongly agree), a sample item is I am often confused about what emotion I am feeling.
Ethical reasoning was measured using the Ethics Position Questionnaire (Forsyth,
1980), which assesses two approaches to ethical reasoning: idealism and relativism, with 10
items each. A sample ethical idealism item is One should never psychologically or physically
harm another person. A sample relativism item is What is ethical varies from one situation
and society to another. Responses are given in a nine-point Likert format (1=completely
disagree, 9=completely agree). Reliability for the two subscales is acceptable at .80 and .73
respectively (Forsyth). This approach is regularly used in assessment of ethical reasoning
(e.g. Dubinsky, Nataraajan, & Huang, 2005; Waterman, 1988). Higher scores on idealism
approach reflecting a belief that there is a morally “right” course of action to be taken in any
given situation. Higher scores on relativism reflects less reliance on rules and norms when
making ethical decisions, instead focusing on differences in the situations or individuals
involved (Forsyth; Waterman).
Results and Discussion-Study 2
Descriptive statistics are presented individually for males and females in Table 2,
showing apparent gender differences in emotional manipulation, psychopathy, and ethical
idealism. Analysis via t-test confirmed that these differences were significant, with males
scoring significantly higher than females on emotional manipulation, t(268)=-1.56, p<.001,
and psychopathy, t(271)=-4.28, p<.001. Females scored significantly higher than males on
ethical idealism, t(260)=5.00, p<.001. Internal reliabilities were calculated across the whole
sample and were generally good to excellent. However, reliability for secondary psychopathy
was low (albeit in line with previous research; e.g. Grieve & Mahar, in press), and for the
perceived poor emotional skills and emotional concealment scales was also low, possibly due
to the small number of items in each, or, as suggested in Study 1, varying intrapersonal
conceptions of emotionality.
Separate multiple regressions were conducted for males and females for emotional
manipulation, perceived poor emotional skills and emotional concealment, with missing
cases excluded listwise. All relevant multivariate assumptions were met. Bivariate
correlations are presented separately in Table 3 for males and females.
The first hypothesis was differentially supported as a function of gender. For males, the
linear combination of psychopathy, alexithymia, EI and ethical position significantly predicted
emotional manipulation, R= .64, accounting for 40.5 % of variance, F(6,75)= 8.50, p<.001,
with ƒ²=.69, indicating a very large effect (Cohen, 1992). Within the regression model, higher
levels of primary psychopathy and EI were significant predictors of emotional manipulation.
For females, the linear combination of primary psychopathy, secondary psychopathy,
alexithymia, EI and ethical position also significantly predicted emotional manipulation, R=
.54, accounting for 29.1 % of variance, F(6,163)=11.13, p<.001, with ƒ²=.41, a large effect
size. However for females, in addition to higher levels of primary psychopathy and EI, higher
levels of secondary psychopathy and lower levels of ethical idealism were also significant
predictors of emotional manipulation. Details of the regression are presented separately for
males and females in Table 4.
However, note that for females, EI has a very low bivariate correlation with emotional
manipulation (r=. 07, see Table 3). This suggests that while EI directly predicts emotional
manipulation in males, in females, EI is instead acting as a suppressor variable, significantly
contributing to the psychopathy-emotional manipulation nexus through the suppression of
extraneous variance. For both males and females, alexithymia was not a significant predictor of
emotional manipulation. In females, higher levels of secondary psychopathy also significantly
predicted emotional manipulation. This suggests that antisocial tendencies play a different role
in emotional manipulation for males and females. For men, being emotionally manipulative is
unrelated to secondary psychopathy, however for women, the two constructs appear to be
linked. Further, for females, lower levels of idealistic ethical reasoning were also associated
with emotional manipulation. Hence, it seems that despite the intuitive appeal of ethical
reasoning being influential in the emotional manipulation-psychopathy relationship, this effect
is only germane in females. For females, believing in an ethically “right” course of action is
related to less emotional manipulation, however for males, ethical position is unrelated to
Perceived Poor Emotional Skills
The second hypothesis was partially supported. Again, the relationships evident
differed as a function of gender. For males, the combination of psychopathy, alexithymia, EI
and ethical position significantly predicted perceived poor emotional skills, R= .74,
accounting for 54.6 % of variance, F(6,75)=15.01, p<.001.ƒ²= 1.21, indicating an extremely
large effect size (Cohen, 1992). Within the model, lower EI significantly predicted perceived
poor emotional skills, and higher levels of alexithymia were marginally significant in the
model. For females, the combination of primary psychopathy, secondary psychopathy,
alexithymia, EI and ethical position also significantly predicted perceived poor emotional
skills, R= .56, accounting for 31.4% of variance, F(6,163)=12.41, p<.001. ƒ²=.46, a large
effect size. For females, like males, lower EI was associated with perceived poor emotional
skills. However, unlike males, alexithymia did not significantly contribute to perceived poor
emotional skills. In addition, higher levels of ethical idealism were also significantly
associated with perceived poor emotional skills for females. Further, for females, given the
low bivariate correlation between ethical idealism and perceived poor emotional skills (r=-
14 Download full-text
.02 , see Table 3), ethical idealism appears to have a suppression effect in the model. This
suppression effect, possibly through idealism’s conceptual relationships with the prosocial
aspects of EI, highlights the complexity of these constructs for females. For both males and
females, in contrast to predictions, secondary psychopathy was not related to perceived poor
emotional skills. Details are presented in Table 5.
The limited role of secondary psychopathy suggests two extensions to findings from
Study 1. Firstly, it seems that the inclusion of alexithymia in the regression equation has
allowed a fuller understanding of the bivariate secondary psychopathy-perceived poor
emotional skills relationship, particularly for males, with alexithymia proving to be a much
stronger predictor. Secondly, these results may also indicate that having poor emotional skills
(as evident by higher scores on alexithymia and lower scores on EI) may fundamentally
differ from the perception of having poor emotional skills, and any resulting anxiety ensuing
from this perception. This explanation may also align with the low reliability of the perceived
poor emotional skills measure.
The final hypothesis, that alexithymia would be related to emotional concealment was
supported for both males and females. For males, the combination of psychopathy,
alexithymia, and EI did not significantly predict emotional concealment, F(4,83)=2.36,
p=.06., R= .32, accounting for 10.2% of variance. However, with ƒ²=.11, suggesting a small
to medium effect size (Cohen, 1992), it seems likely that the model was not significant at
α=.05 due to the study’s modest power to detect an effect of this size, rather than indicating a
lack of relationship. Within the model, higher levels of alexithymia significantly predicted
emotional concealment. For females, the combination of primary psychopathy, secondary
psychopathy, alexithymia, and EI did significantly predict emotional concealment, R=.33,
accounting for 10.6% of variance, F(4,175)=5.21, p=.001.ƒ²=.12, representing a small to