1
The role of psychopathy in predicting PTSD:
A prospective study in a large military sample deployed in Afghanistan.
Can psychopathy be a protective factor against developing post-traumatic stress disorder (PTSD)? Clarifying the role of psychopathy in the development of PTSD is useful for getting a better understanding of PTSD. Theoretical models describe the Interpersonal/Affective facets of psychopathy as a protective factor and the Lifestyle/Antisocial facets as a risk factor for developing PTSD. Earlier research on the role of psychopathy in predicting PTSD is ambiguous. The current prospective study, with 473 military deployed in Afghanistan showed, contrary to predictions, that the affective facet of psychopathy significantly increased PTSD symptoms six months after deployment. Having an impulsive and thrill-seeking lifestyle increased PTSD symptoms as well, however, this was no longer significant after controlling for other known predictors of PTSD. To summarize, no evidence was found for the idea that psychopathy would function as a protective factor for developing PTSD. In contrast, there was some evidence to suggest that psychopathy is a risk factor for PTSD.Cleckley described the psychopath as “Almost as incapable of anxiety as of profound remorse” (Cleckley, 1976, p. 339-340). Therefore, diagnosing an anxiety disorder such as post-traumatic stress disorder (PTSD), as described in the Diagnostic and Statistical Manual of Mental
Disorders (4th ed., text rev.; DSM-IV-TR;
American Psychiatric Association [APA], 2000), in a psychopathic individual may not be something one would expect to happen soon. Nevertheless, these two forms of psychopathology co-occur (Blackburn, Logan, Donnelly, & Renwick, 2003). Identifying certain aspects of psychopathy as risk or resilience factors for developing PTSD can influence selection criteria of the military and other professions at increased risk for trauma exposure, such as the police or the fire department. If some aspects of psychopathy are indeed protective against developing PTSD, people scoring high on these aspects could be selected.
By selecting people with a lower risk of developing PTSD the development of PTSD in these professions could potentially be reduced. This would consequently be associated with less health care and absenteeism costs. Furthermore, since there is a correlation between PTSD and criminality it is assumed that the reduction of PTSD will be accompanied by a reduction of crime rates (Goff, Rose, Rose, & Purves, 2007; Hare & Neumann, 2008).
Psychopathy, as operationalized in the Psychopathy Checklist-Revised (PCL-R), is a combination of certain personality traits and socially deviant behavior (Hare, 2003; Hare & Neumann, 2005). It consists of core personality traits (“someone who is selfish, callous and remorselessly using others”) and behavioral features (“someone with a chronically unstable, anti-social and socially deviant lifestyle”) (Hare, 1991, as cited in Willemsen, De Ganck, & Verhaeghe, 2012). More recent research described the PCL-R as measuring four facets (Hare & Neumann, 2005), i.e., the Interpersonal and the Affective facet - together described as the personality factor (F1) - and the Lifestyle and Antisocial facet - together described as the behavioral factor (F2) (e.g. Hare & Neumann, 2005).
Master’s Thesis Health Psychology: Clinical Forensic Psychology Name: Fleur van Roosmalen Student ID number: 11046430 Supervisor 1: Bruno Verschuere Supervisor 2: Arnold van Emmerik Date: December 14, 2016
2
As for PTSD, it is a condition that can develop after a person has experienced, witnessed, or was otherwise confronted with a traumatic event. The traumatic event should involve actual or threatened death or serious injury, or a threat to the physical integrity of oneself or others, and invoke intense fear, helplessness or horror. To diagnose PTSD, symptoms of all following clusters have to be present for more than one month; persistently re-experiencing the traumatic event, avoiding stimuli associated with the trauma and persistent symptoms of increased arousal (APA, 2000).
Several meta-analyses on risk factors for PTSD revealed several predictors (e.g. post-trauma social support). It is, however, not entirely clear how these factors may serve to influence the development of PTSD (Brewin, Andrews, & Valentine, 2000; Ozer, Best, Lipsey, & Weiss, 2008). Even more unclear is the relation between psychopathy and PTSD. One explanation given for the co-occurrence of psychopathy and PTSD is negative affectivity (Sellbom, 2015). Negative affectivity is a dispositional vulnerability towards anxiety, stress reactivity, anger and other negative emotions (Watson & Clark, 1984). Several prospective studies have linked negative affectivity to PTSD (e.g. Rademaker, Van Zuiden, Vermetten, & Geuze, 2011) and to psychopathy (e.g. Hale, Goldstein, Abramowitz, Calamari, & Kosson, 2004).
As noted before, Willemsen and colleagues (2012) theorized the relation between psychopathy and PTSD as opposite for the personality factor (F1) (i.e., a negative, protective relationship) and the behavioral factor (F2) (i.e., a positive, risk increasing relationship). The idea of a protective role of the personality factor (F1) can be understood by using the low fear hypothesis. Following this hypothesis, one would expect individuals scoring high on F1 to have a reduced capacity or increased threshold for activating fear systems (Anderson & Kiehl, 2012; Brook, Brieman, & Kosson, 2013). Research revealed that F1 is indeed associated with lower anxiety levels (Blonigen et al., 2010; Hicks & Patrick, 2006). Also, F1 is characterized by low fear conditionability while the development of PTSD can in part be conceptualized as a
fear-conditioning process (Birbaumer et al., 2005; Orr et al., 2000). Furthermore, F1 is associated with amygdala dysfunction, such as a hyporeactive amygdala, while PTSD is associated with a hyperreactive amygdala (e.g. Blair & Mitchell, 2009).
Further, the risk enhancing role of the behavioral factor (F2) can be understood from the viewpoint that individuals scoring high on the behavioral factor would have an increased risk of traumatic exposure. The highly impulsive and anti-social psychopathic individuals are said to experience more traumatic exposure as a result of a certain risk-taking lifestyle which increases the risk of encountering stressful life events (Frick, Lilienfeld, Ellis, Loney, & Silverthorn, 1999; Willemsen et al., 2012).
The few earlier studies on the relationship between psychopathy and PTSD have yielded inconclusive results. Some found a negative association between (facets of) F1 and PTSD (Pham, 2012; Sellbom 2015; Willemsen et al., 2012). However, others found no relation between the two (Blonigen, Sullivan, Hicks, & Patrick, 2012; Kubak & Salekin, 2009; Pham, 2012). Furthermore, the Affective facet was found to moderate the relationship between traumatic exposure and later PTSD, which supports the idea of the Affective facet as a protective factor (Willemsen et al., 2012). Similarly, studies on the relationship between (facets of) F2 and PTSD found positive associations (Blonigen et al., 2012; Kubak & Salekin, 2009; Sellbom, 2015), while other studies found no relation (Kubak & Salekin, 2009; Pham, 2012; Willemsen et al., 2012). These earlier studies were mostly (1) cross-sectional studies with (2) small samples. The current prospective study on the predictive value of different aspects of psychopathy in a large military sample therefore makes a useful contribution to existing studies on the relation between psychopathy and PTSD.
Based on the earlier described low fear hypotheses the individuals scoring high on the Affective facet of psychopathy are expected to be less inclined to experience traumatic events as threatening and horrifying than individuals scoring low on this feature. The Affective facet of
3
psychopathy is therefore expected to be a protective factor against the development of PTSD, and thus to moderate the relationship between exposure to traumatic events and PTSD symptom severity. Whereas, the impulsive, irresponsible and anti-social lifestyle of some psychopathic individuals is expected to increase the risk for traumatic exposure and thereby the risk for developing PTSD (Frick et al., 1999). The Lifestyle and Antisocial facets are therefore expected to be a risk factor for developing PTSD. This led to the following hypotheses:
Hypothesis 1: The Affective facet of psychopathy
reduces the development of PTSD symptoms after exposure to traumatic events (Figure 1).
Figure 1. The Affective facet as a protector. Hypothesis 2: The Lifestyle and/or Antisocial facet
of psychopathy increases the development of PTSD symptoms after exposure to traumatic events.
Hypothesis 3: The expected positive association
between the Lifestyle and/or Antisocial facet of psychopathy and later PTSD symptoms is mediated by exposure to traumatic events (Figure 2).
Figure 2. The Lifestyle and/or Antisocial facets as
a risk factor.
Method
Participants and procedure
The current study uses data which is collected in a large longitudinal study of stress-related disorders in deployed military personnel, also called the Dutch Prospection in Stress Related Military Research (PRISMO) (Rademaker et al., 2011). The institutional review board of the University Medical Center Utrecht approved the PRISMO study (Rademaker et al., 2011). It was conducted in a cohort of 1032 soldiers who, as part of the International Security Assistance Force (ISAF), took part in missions to Afghanistan between 2005 and 2008. Participants were recruited on military bases in the Netherlands. They were provided with a verbal and written description of the study. The military personnel volunteered by signing a written consent form before their mission. They were asked to fill in several self-report questionnaires before and after their four-month deployment to Afghanistan. A part of the data collected in this large longitudinal PRISMO study is used in the current study. The statistical analyses of the current study are performed on a sample consisting of 473 participants.
The sample size of the current study is determined based on a two-tailed power calculation with an alpha of .05 and power of .80. Sellbom (2015) found significant correlations between the factors of psychopathy and PTSD (Factor 1 and PTSD -.41 and factor 2 and PTSD .31). This resulted in an average effect size of r = .36. Furthermore, Willemsen and colleagues (2012) found significant bivariate associations between some facets of psychopathy and PTSD (Interpersonal facet and PTSD -.14 and Affective facet and PTSD -.13). This resulted in an average effect size of r = .13. A calculation with G*power, using an in-between effect size (.18) which was based on effect sizes found in earlier similar studies, resulted in a required minimum sample of 237 people to have a power of .80 (235 df) (Faul, Erdfelder, Lang, & Buchner, 2007).
Applying exclusion criteria on the original
dataset resulted in a final sample size. At first, women (n = 93) are excluded from the dataset to
4
increase homogeneity. Only a small part (9%) of the total dataset consisted of women. Furthermore, gender is found to be a risk factor for developing PTSD (Brewin et al., 2000) and there are gender differences found in psychopathy (e.g. being more prevalent in men; Vitacco, Neumann, & Jackson, 2005). Secondly, military men who during their mission had not experienced any potentially traumatic event (n = 84), as defined by the
DSM-IV-TR A1 criterion of PTSD, are also excluded
(APA, 2000). Thirdly, participants who failed to complete the psychopathy measure and the post-traumatic stress measure six months after deployment (n = 382) are excluded, as data on these questionnaires is required to draw conclusions about the degree of psychopathy and PTSD. Participants who failed to complete the demographic questionnaire and/or the measure of early trauma are not excluded. This resulted in the final sample of 473 participants.
Measures
Traumatic exposure
The measurement of exposure to (potential) traumatic events is based on the deployment exposure scale (DES) (Reijnen, Geuze, & Vermetten, 2015). It contains 19 items rated true or false and was administered one month after deployment. Items refer to combat experiences (e.g. shooting aimed at you), suffering of other experiences (e.g. being rejected by locals), lack of control (e.g. insufficient means to intervene), and negative feelings about the mission (e.g. the feeling that the mission is pointless) (Appendix A). The items about lack of control and feelings about the mission (i.e., items 9, 15, 16, 17 and 18), are not included in this study because they measure deployment experiences which are not defined as potentially traumatic by the DSM-IV-TR A1 criterion of PTSD (APA, 2000). That leaves items 1-8, 10-14 and 19 as (potentially) traumatic experiences, therefore, these items are used in this study to measure trauma exposure.
No validation of the DES exists yet (Reijnen et al., 2015). However, most of the questions of the questionnaire are derived from validated exposure lists, like the Combat Exposure Scale (CES)
(Koopmans, 2009). The CES has an internal consistency of .85 (Coefficient alpha) and a test-retest reliability of r = .97 (Keane et al., 1989). The Cronbach’s alpha for the trauma exposure checklist is questionable (Cronbach’s ɑ = .66; Nunnally & Bernstein, 1994, as cited in Iacobucci & Duhachek, 2003).
Psychopathy
The Temperament and Character Inventory— Short Form-Psychopathy (TCI-P) scale is used as a proxy for psychopathy and was administered before deployment to Afghanistan. The TCI-P consists of a subset of items of the Dutch short-form version of the Temperament and Character Inventory (TCI-SF; Duijsens & Spinhoven, 2006; Duijsens, Spinhoven, Verschuur, & Eurelings-Bontekoe, 1999). This self-report personality questionnaire is used because it is less expensive and time consuming than the PCL-R and, furthermore, personality questionnaires are found to be capable of measuring psychopathy (e.g. Snowden & Gray, 2010). Earlier research suggests that the TCI-SF in particular can be used to measure (aspects of) psychopathy (Huijzer, 2015). In the study of Huijzer, data was obtained from 84 male participants to construct the subscales of the TCI-P. These participants had completed different questionnaires including the TCI-SF and the Self-Report Psychopathy Scale III (SRP-III; Williams, Paulhus, & Hare, 2007). The subscales measuring psychopathy (TCI-SF-P, 30 items) are constructed based on correlations between the subscales of the SRP-III and individual items of the TCI-SF. Appendix B contains the details of the construction of the psychopathy subscales (TCI-SF-CA and TCI-SF-ELS) as used in this study. The Cronbach’s alpha is questionable for the TCI-SF-ELS subscale (Cronbach’s ɑ = .58) and acceptable for the CA subscale (Cronbach’s ɑ = .75) (Nunnally & Bernstein, 1994, as cited in Iacobucci & Duhachek, 2003).
The SRP-III subscales are said to capture psychopathy in a similar manner as the PCL-R (Mahmut, Menictas, Stevenson, & Homewood, 2011). That implies that the subscales as constructed in this study capture the four-factor structure as well (Hare, 2003; Williams et al.,
5
2007). Following this line of reasoning the four-factor structure of psychopathy can be measured with the constructed TCI-P subscales (Hare & Neumann, 2005; Neumann, Hare, & Johansson, 2013).
PTSD Symptoms
The measurement of PTSD symptoms is
conducted using the Dutch version of the Self-Rating Inventory for PTSD (SRIP) (Hovens, Bramsen, & Van Der Ploeg, 2000, 2002; Hovens et al., 1994). The self-report procedure is used instead of a structured interview because it is a less expensive and time consuming but adequate manner to measure PTSD (Sijbrandij et al., 2013). The SRIP consists of 22 items on a 4-point scale, ranging from 1 (not at all) to 4 (very much) (Hovens et al., 1994). Participants are asked to report on the last four weeks (Hovens et al., 2000). The scale is based on the description of PTSD in the DSM-IV-TR (Bramsen, Dirkzwager, Van Esch, & Van Der Ploeg, 2001). It generates a total score (SRIP Total) and three subscales measuring intrusion (SRIP-I, 6 items), avoidance (SRIP-A, 9 items) and increased arousal (SRIP-IA, 7 items) (Van Der Ploeg, Bramsen, & Reuling, 2000). The items are phrased in a trauma-independent way which makes it applicable for use with all populations (Van Der Ploeg et al., 2000).
A total score of 52 points or higher appeared to be a good predictor of the clinical diagnosis of PTSD (Hovens et al., 2000). However, in later research a cutoff of 38 was used, since this provided the highest sensitivity and specificity for a PTSD diagnosis according to DSM-IV-TR (Reijnen et al., 2015; Van Zelst et al., 2003). Test-retest reliability (r = .92), internal consistency (Cronbach’s ɑ = .90) and concurrent validity of the SRIP with other PTSD measures (r = .80 and .82) have been reported to be adequate (Hovens et al., 2000). Den Boer (2013) found an acceptable Cronbach’s alpha (.86) for the SRIP prior to deployment and a good Cronbach’s alpha (.91) after deployment (Nunnally & Bernstein, 1994, as cited in Iacobucci & Duhachek, 2003).1
1 The used dataset contained SRIP total scores and no
SRIP item scores, therefore, Cronbach’s alpha could not
In the current study PTSD symptoms are
measured before, and one, six, twelve and twenty-four months after deployment to Afghanistan. According to DSM-IV-TR criteria, PTSD symptoms have to be present for more than one month before somebody can be diagnosed with PTSD (APA, 2000). As time passed after deployment, the number of drop-outs on the PTSD questionnaire increased. Due to these two reasons, it is decided to use the registration of PTSD symptoms six months after deployment as the primary outcome measure.
Early traumata
Exposure to (potential) traumatic experiences before the age of 18 is conducted using the Dutch self-report short-form version of the Early Trauma Inventory (ETISR-SF) (Bremner, Bolus, & Mayer, 2007;Rademaker, Vermetten, Geuze, Muilwijk, & Kleber, 2008). It was administered before deployment to Afghanistan. The checklist consists of 27 items (rated true/false) forming four scales: general traumas after the age of 18 (11 items, e.g. did you ever see someone murdered), physical punishment before the age of 18 (5 items, e.g. were you ever burned with hot water, a cigarette or something else), emotional abuse before the age of 18 (5 items, e.g. were you often told you were no good) and sexual events (6 items, e.g. did anyone ever have genital sex with you against your will). Scores on each scale represent the total number of endorsed items (Rademaker et al., 2011).
The ETISR is found to be a valid instrument for the measurement of childhood physical, emotional, and sexual abuse, as well as general traumatic events (e.g. natural disaster.) Bremner and colleagues (2000, 2007) found most item scores, domain scores and total scores to highly correlate with measures of PTSD symptomatology. The ETISR showed high levels of internal consistency within the individual domains (Bremner et al., 2007).
be calculated for the SRIP. A Cronbach’s alpha of the SRIP in another study with a similar sample is mentioned to give an indication of Cronbach’s alpha in the current study (Den Boer, 2013).
6
Looking at the psychometric properties of the Dutch translation of the short form of the ETISR, some subscales proved adequate internal consistency: i.e. physical abuse (Cronbach’s ɑ = .76) and emotional abuse (Cronbach’s ɑ = .83). However, others did not: i.e. general trauma (Cronbach’s ɑ = .48) and sexual abuse (Cronbach’s ɑ = .53) (Rademaker et al., 2008). The Dutch short form of the ETISR hasn’t been validated yet. 2
Statistical analysis
Participants who completed pre-deployment psychopathy assessments (SF-CA and TCI-SF-ELS) and post-deployment PTSD assessment (SRIP 6 months after deployment) (N = 473) but who failed to complete other questionnaires (PTSD at baseline: n = 71; Trauma Exposure: n = 67; ETISR-SF: n = 1) are excluded from the analyses using pairwise deletion. Data of participants who completed the pre-deployment psychopathy assessment (CA and TCI-SF-ELS) and post-deployment PTSD assessment (SRIP 6 months) are compared with those who did not complete the post-deployment PTSD assessment (SRIP 6 months) using chi-squared analysis or independent sample Student’s t tests. Next, data is examined for outliers and other potential violations of assumptions of analyses.3
Based on the possibility that detected outliers can represent actual high levels of PTSD symptoms and based on the result that deletion of detected outliers did not result in significant differences, detected outliers are not deleted. Kurtosis and skewness are examined and data is transformed
2 The used dataset contained ETISR total scores and no
ETISR item scores, therefore, Cronbach’s alpha could not be calculated for the ETISR. A Cronbach’s alpha of the ETISR in another study with a similar sample is mentioned to give an indication of Cronbach’s alpha in the current study (Rademaker et al., 2008).
3 The examination for outliers was done in three
different ways. First, a univariate outlier analysis was performed. A score on the dependent variable (the total score of PTSD symptoms six months after deployment) was defined as outlier when it had an absolute z-score greater than 3.29 (Field, 2009, p. 102 - 103).
when appropriate. Analyses are performed with and without transformations, which yielded no meaningful differences in the results or the conclusions regarding the hypotheses. The increase in explained variance of the tested models when using the transformed data of PTSD scores six months after deployment is negligible.4 Based
on these analyses, the results of this study are based on the untransformed data, and outliers are not deleted.
The main analyses are performed as follows. First, bivariate Pearson correlations are performed to examine relationships between psychopathy scores and PTSD scores. Second, regression analyses are performed, to see if psychopathy scores predict PTSD scores. Additionally, all regression analyses are performed a second time to see if the psychopathy scores remained significant predictors of PTSD scores when potential confounders (rank, early trauma, trauma exposure and PTSD symptoms at baseline) are taken into account. Third, the moderation model is examined following the recommendations of Frazier, Tix and Barron (2004). This means using a regression model with the centered Callous Affect scores, the centered scores of the Trauma Exposure variable and the centered scores of the Interaction of Callous Affect and Trauma Exposure. Finally, the mediation model is examined following the recommendations for mediation analysis as described by Baron and Kenny (1986). The moderation and mediation analysis are both performed a second time with the addition of potential confounders (Early trauma and PTSD at baseline).
Second, a multivariate outlier analysis was done. Simple scatterplots were performed on the independent variables, the dependent variable and the potential confounders. Third, a multiple hierarchical regression analysis was conducted with rank, early trauma, trauma exposure and PTSD symptoms at baseline in block 1. Total ELS and total CA score were added in block 2. When a standardized residual had an absolute value above 3.29 it was defined as an outlier. In total, 11 outliers were identified.
7
Results
Descriptives and attrition analysis
Demographic variables of participants who completed the assessment of psychopathy prior to deployment and the PTSD assessment after deployment (responders) are displayed in Table 1.
Non-responders (participants who only completed the assessment of psychopathy prior deployment but not the PTSD assessment) reported on average more exposure to general traumas after the age of 18 (M = 2.27, SD = 1.73) as compared to responders (M = 1.96, SD = 1.63),
t(631) = -2.06, p < .05, Cohen’s d = 0.18 (equal
variances assumed). Furthermore, non-responders
scored higher on the psychopathy assessment of erratic lifestyle (TCI-SF-ELS) (M = 4.33, SD = 2.22) as compared to responders (M = 3.86, SD = 2.08), t(633) = -2.48, p < .05, Cohen’s d = 0.22 (equal variances assumed). Finally, responders had been on more missions before the current mission (M = 0.97, SD = 1.26) in comparison to
non-responders (M = 0.65, SD = 0.93), t(350) = 3.30, p < .05, Cohen’s d = 0.29 (equal variances not assumed). No other differences are observed.
Predicting PTSD symptoms
The regression model (including covariates) significantly predicted PTSD symptoms, F(6,395) = 29.92, p < .001 and accounted for 30% in the explained variance within PTSD scores six months after deployment.
Hypothesis 1: Is the Affective facet a protector?
In contrast to hypothesis 1, there is a positive significant bivariate (Pearson) correlation between Callous Affect and PTSD symptom severity six months after the mission (r = .28, p < .001). Thus, participants scoring higher on Callous Affect reported higher PTSD symptom severity.
In line with the correlation analysis, the
regression analysis did not support hypothesis 1 but instead found that Callous Affect is positively related to PTSD symptoms 6 months after deployment (Table 2). Callous Affect remained a significant predictor of PTSD when covariates are included in the model (Table 3). This suggests that callous affect is a risk, rather than a protective factor for developing PTSD.
A second set of regression analyses is performed to examine whether Callous Affect is a moderator of the relation between traumatic experiences and post-deployment PTSD symptoms. In contrast with hypothesis 1, a significant positive main effect of Callous Affect and a significant positive effect of the interaction term was found, indicating that there is indeed a moderator effect (Table 4). Instead of being protective, the results show that Callous Affect enhances the risk on the development of PTSD symptoms after exposure to traumatic events (Figure 3).
Table 1
(Demographic) Variables and Rank of Responders (N = 473)
Variable M SD
Age 29.53 9.64
Early trauma exposure 3.45 2.94
General trauma 1.96 1.63
Physical 0.95 1.31
Emotional 0.46 1.01
Sexual 0.09 0.45
PTSD at baseline 26.98 5.27
Number of previous missions 0.97 1.26
Erratic Lifestyle 3.86 2.08 Callous Affect 5.81 3.27 n % Rank Troop 182 38.5 Corporal 84 17.8 NCO-officerᵃ 137 29.0 Sub-officer 46 9.7 Staff officer 24 5.1 Education Secondary school 18 3.9
Lower general education 164 35.2
Intermediate vocational education 174 37.3 Higher general secondary
education
54 11.6
Higher professional education 34 7.3
University 21 4.5
Other 1 0.2
8
Callous Affect is found to be risk enhancing even when other known predictors (early trauma and PTSD at baseline) are added into the model (Table 5). All results indicate Callous Affect being a risk, rather than a protective factor for developing PTSD.
Table 3
Hierarchical Multiple Regression of Covariates and Psychopathy Measures on Post-deployment PTSD (N = 473)
B SE B β Block 1 Constant 10.21 1.65 Rank -0.26 0.24 -.05 Early Trauma 0.29 0.10 .13** Traumatic Experiences 0.41 0.13 .14** Pre-deployment PTSD 0.57 0.05 .46*** Block 2 Constant 8.65 1.71 Rank 0.02 0.26 .00 Early Trauma 0.25 0.10 .11** Traumatic Experiences 0.35 0.12 .12** Pre-deployment PTSD 0.53 0.06 .43*** Callous Affect 0.24 0.09 .12* Erratic Lifestyle 0.24 0.14 .07
Note: R² = .29 for Block 1, ΔR² = .02 for Block 2 (p < .05), * < .05, ** < .01, *** < .001.
Table 2
Forced Entry Regression of Psychopathy Measures on Post-deployment PTSD (N = 473) B SE B β Constant 23.22 .73 Callous Affect .50 .10 .24*** Erratic Lifestyle .41 .15 .13** Note: R² = .09, * < .05, ** < .01, *** < .001. Table 4
Hierarchical Multiple Regression of the interaction of Callous Affect and Traumatic Experiences on Post-deployment PTSD (N = 473) B SE B β Block 1 Constant 27.70 0.32 Callous Affect 0.51 0.10 0.25*** Traumatic experiences 0.46 0.14 0.16** Block 2 Constant 27.53 0.32 Callous Affect 0.49 0.10 0.24*** Traumatic experiences 0.45 0.14 0.15**
Callous Affect x Traumatic experiences 0.13 0.04 0.15**
9
Figure 3. Mean PTSD scores after traumatic exposure for low vs high Callous Affect individuals.Table 6
Pearson Correlations of Self-Report Measures
1 2 3 4 5 6 7 8 9 10 1. Callous Affect - 2. Erratic Lifestyle .30** - 3. PTSD total at baseline .26** .13** - 4. Trauma Exposure .17** .13** .09 - 5. PTSD 6 months .28** .20** .50** .20** -
6. Total Early Trauma .10** .15** .23** .10* .25** -
7. Early General Trauma .09* .15** .13** .17** .22** .77** - 8. Early Physical Abuse .03 .05 .18** -.01 .09 .74** .30** -
9. Early Emotional Abuse .13** .11** .21** .03 .22** .62** .24** .31** - 10. Early Sexual Abuse -.03 .05 .04 -.00 .05 .31** .09* .19** .09* -
Note. * p < .05, **p < .01.
Table 5
Hierarchical Multiple Regression of the interaction of Callous Affect with Traumatic Experiences on Post-deployment PTSD, including Covariates (N = 473) B SE B β Block 1 Constant 12.38 1.48 Early Trauma 0.27 0.10 .12** PTSD at baseline 0.53 0.06 .43*** Callous Affect 0.27 0.09 .14** Traumatic experiences 0.37 0.12 .13** Block 2 Constant 11.99 1.46 Early Trauma 0.28 0.09 .13** PTSD at baseline 0.54 0.05 .43*** Callous Affect 0.25 0.09 .13** Traumatic experiences 0.35 0.12 .12**
Callous Affect x Traumatic experiences 0.14 0.04 .17***
10
Hypothesis 2: Is an Erratic Lifestyle a risk for developing PTSD symptoms?Regarding hypothesis 2, there is a positive significant bivariate (Pearson) correlation between Erratic Lifestyle and PTSD symptom severity six months after mission (r = .20, p = .000; Table 6). Thus, participants scoring higher on Erratic Lifestyle reported higher PTSD symptom severity. In line with the correlation analysis, the regression analysis did support hypothesis 2 as well and revealed that Erratic Lifestyle is positively related to PTSD symptoms six months after deployment (Table 1). However, when covariates are included in the model, this effect did not remain significant (Table 2). Taking into account other known predictors of PTSD revealed that no support was found for the idea of an erratic lifestyle being a risk factor for developing PTSD symptoms.
Hypothesis 3: Do individuals with an Erratic Lifestyle experience more traumatic events?
Results partly support hypothesis 3, i.e. the relation between Erratic Lifestyle and PTSD symptom severity can be partly explained by increased traumatic experiences (Table 7), even when other known predictors are included in the model (Table 8). However, including the number of traumatic experiences in the regression analyses does not make that much of a difference. This indicates that the relation between an erratic lifestyle and PTSD symptom severity can be explained by the number of traumatic experiences. At the same time, by explaining only a small part of the relation it shows that there are probably other factors more essential in explaining the relation between Erratic Lifestyle and PTSD symptom severity. The positive association between a pre-trauma erratic lifestyle and later PTSD symptom severity is partly mediated by trauma exposure.
Table 7
Hierarchical Multiple Regression of Erratic Lifestyle and traumatic experiences on Post-deployment PTSD (N = 473)
B SE B β
Step 1
Erratic Lifestyle on PTSD 0.63 0.14 .20***
Step 2
Erratic Lifestyle on Traumatic experiences 0.14 0.05 .13**
Step 3
Erratic Lifestyle on PTSD 0.56 0.15 .18***
Traumatic experiences on PTSD 0.52 0.14 .18***
Note: R² =.04 for Step 1, R² =.02 for Step 2, R² =.07 for Step 3, * < .05, ** < .01, *** < .001.
Table 8
Hierarchical Multiple Regression of Erratic Lifestyle and traumatic experiences on Post-deployment PTSD Scores, including Covariates (N = 473) B SE B Β Step 1 Early Trauma 0.28 0.10 .13** PTSD at baseline 0.57 0.05 .46*** Erratic Lifestyle on PTSD 0.38 0.14 .12** Step 2
Erratic Lifestyle on Traumatic experiences 0.14 0.05 .13**
Step 3
Early Trauma 0.26 0.10 .12**
PTSD at baseline 0.56 0.05 .45***
Erratic Lifestyle on PTSD 0.33 0.14 .10*.
Traumatic experiences on PTSD 0.39 0.12 .14**
11
Discussion
This study did not find any evidence that
psychopathy protects against developing PTSD symptoms. On the contrary, it indicates that having traits of psychopathy increases the likelihood of developing postdeployment PTSD symptoms. First, predeployment Callous Affect and Erratic Lifestyle are found to be significantly related to postdeployment PTSD symptoms and moreover to be a significant predictor of postdeployment PTSD symptoms. However, when there is controlled for known predictors like early trauma, traumatic experiences and PTSD symptoms before deployment, only Callous Affect remains a significant predictor. These results indicate a risk-increasing role of having little empathy and flat affect with regards to developing postdeployment PTSD symptoms. Second, PTSD symptom severity can be explained by an interaction of trauma exposure and Callous Affect. When an individual is exposed to many traumatic events, the likelihood of developing more severe PTSD symptoms is higher for individuals scoring high on Callous Affect. Third, the relation between Erratic Lifestyle and postdeployment PTSD symptoms is found to be partly explained by the number of traumatic experiences, with individuals having an impulsive, thrill-seeking lifestyle experiencing more traumatic events and subsequently developing more (severe) PTSD symptoms.
The found risk-enhancing effect of Callous Affect is not in line with earlier research (Willemsen et al., 2012; Sellbom, 2015). Earlier research showed the Affective facet to protect against developing PTSD. This incongruent finding could be explained by the psychopathy measure used for this study. One could question whether the used psychopathy instrument indeed measures psychopathy traits and specifically whether the Callous Affect subscale measures threat sensitivity. This is relevant because, based on the low fear hypothesis, the protective role of psychopathy is explained by low threat sensitivity for individuals scoring high on the Affective facet. As said before, the construction of the psychopathy measure used in this study was based on correlations between items of a personality questionnaire (TCI-SF) and subscales of a
validated psychopathy self-report questionnaire (SRP-III). In recent literature, researchers questioned the idea of low anxiety as being part of the psychopathy construct. Since the role of anxiety in psychopathy is unclear, the anxiety items were removed from subsequent versions of the SRP-III. This means, the SRP-III has no items assessing lack of anxiety (Visser, Ashton, & Pozzebon, 2011). This means, there are no anxiety related items in the Callous Affect subscale. The absence of anxiety related items in the psychopathy measure used in this study could explain why no protective role of Callous Affect is found. Therefore, having low threat sensitivity in psychopathic individuals could still be a protective factor with regard to developing PTSD symptoms.
The finding that Erratic Lifestyle does not
significantly predict PTSD when other known predictors are taken into account, may be explained by the weak statistical reliability of this subscale. When constructing the Erratic Lifestyle, using a dataset with 84 male participants, the Cronbach’s alpha was .721. However, in our dataset, with several hundred military trained participants, the Cronbach’s alpha was only .589.
The military conducts a strict selection and
training before they decide to send someone on a military mission. It could be possible that the sample which is used in this study – selected and trained individuals – does not have an impulsive or thrill-seeking lifestyle or that they are trained to not act impulsive on a mission. This could mean that it is not possible to differentiate individuals with an erratic lifestyle from individuals without an erratic lifestyle in this sample. However, it is still possible that having an erratic lifestyle predicts PTSD symptoms in normal society or at least in a setting where individuals are not strictly selected and trained. The idea that specific characteristics inherent to the used sample explain why no robust effect of Erratic Lifestyle on PTSD symptoms is found, is supported by our data. The data showed only a small difference between the participants scoring ‘high’ on Erratic Lifestyle and the participants scoring ‘low’ on Erratic Lifestyle. Participants could score between 0 and 12 on the Erratic Lifestyle subscale. The mean score of 635 participants was 3.98, with a standard deviation of
12
2.12. It would be interesting to see if an erratic lifestyle would predict PTSD symptoms in a sample reporting more impulsivity and thrill-seeking than the sample used in this study.
Despite these limitations, the strong points of the present study are that the relationship between psychopathy and PTSD is prospectively examined, while a large sample is used. The fact that the present study failed to provide evidence to support the idea of aspects of psychopathy being protective against developing PTSD symptoms and instead showing these aspects to be risk increasing, indicates that individuals scoring high on these aspects (Callous Affect) could better be selected for professions in which one will not experience a lot of (potentially) traumatic events instead of professions with increased exposure to traumatic events. Another implication of this study is that it shows the importance of knowing what the Callous Affect measure used in this study really measures. Items in the Callous Affect measure were for example: “I talk openly about my experiences and feelings with friends instead of keeping them to myself” or “In general I am cold and distant in relation to other people”. Does Callous Affect measure little empathy and flat affect, negative affectivity or does it measure a more general deficit to experience and describe emotions, also known as alexithymia (Sellbom, 2015). One could argue that not being able to describe emotions in an adequate manner could lower social support. Receiving no post-trauma social support is known as an important risk factor for developing PTSD symptoms (Brewin et al., 2000; Ozer et al., 2008). Following this line of reasoning, therapists or social workers should help individuals scoring high on Callous Affect to describe their emotions after a traumatic event has occurred. Expression of emotions is expected to enhance received social support which in turn prevents the development of PTSD symptoms.
In order to examine whether above described line of reasoning as suggested by the results of this study is true, it is recommended for further research to study what the Callous Affect measure really measures. Furthermore, it is recommended to investigate whether the Callous Affect measure correlates with received post-trauma social support
or with negative affectivity. At last, it is recommended to perform a prospective study with an extensively validated psychopathy measure like the SRP-III to see if above findings can be replicated.
References
American Psychiatric Association. (2000).Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC:
Author.
Anderson, N. E., & Kiehl, K. A. (2012). The psychopath magnetized: Insights from brain imaging. Trends in Cognitive Sciences, 16(1), 52-60.
Baron, R. M., & Kenny, D. A. (1986). The
moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6),
1173-1182.
Birbaumer, N., Veit, R., Lotze, M., Erb, M., Hermann, C., Grodd, W., & Flor, H. (2005). Deficient fear conditioning in psychopathy.
Archives of General Psychiatry, 62, 799-805.
Blackburn, R., Logan, C., Donnelly, J., & Renwick, S. (2003). Personality disorders,
psychopathy and other mental disorders: Co-morbidity among patients at English and Scottish high-security hospitals. The Journal of Forensic
Psychiatry & Psychology, 14, 111-137.
Blair, R. J. R., & Mitchell, D. G. V. (2009).
Psychopathy, attention and emotion.
Psychological Medicine, 39, 543-555.
Blonigen, D. M., Patrick, C. J., Douglas, K. S., Poythress, N. G, Skeem, J. L., Lilienfeld, S. O., . . . Krueger, R. F. (2010). Multi-method assessment of psychopathy in relation to factors of internalizing and externalizing from the Personality Assessment Inventory: The impact of method variance and suppressor effects.
Psychological Assessment, 22(1), 96-107.
Blonigen, D. M., Sullivan, E. A., Hicks, B. M., & Patrick, C. J. (2012). Facets of psychopathy in relation to potentially traumatic events and post-traumatic stress disorder among female prisoners: The mediating role of borderline personality disorder traits. Personality
13
Disorders: Theory, Research, and Treatment, 3,406-414.
Bramsen, I., Dirkzwager, A. J. E., Van Esch, S. C. M., & Van Der Ploeg, H. M. (2001). Consistency of self-reports of traumatic events in a population of Dutch peacekeepers: Reason for optimism? Journal of Traumatic Stress, 14(4), 733-740.
Bremner, J. D., Bolus, R., & Mayer, E. A.
(2007). Psychometric properties of the Early Trauma Inventory – Self Report. Journal of
Nervous and Mental Disease, 195, 211-218.
Bremner, J. D., Vermetten, E., & Mazure, C. M. (2000). Development and preliminary psychometric properties of an instrument for the measurement of childhood trauma: The Early Trauma Inventory. Depression and anxiety, 12, 1-12.
Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Meta-analysis of risk factors for post-traumatic stress disorder in trauma-exposed adults. Journal of Consulting and Clinical
Psychology, 68(5), 748-766.
Brook, M., Brieman, C. L., & Kosson, D. S.
(2013). Emotion processing in Psychopathy Checklist – assessed psychopathy: A review of the literature. Clinical Psychology Review, 33, 979-995.
Cleckley, H. M. (1976). The Mask of Sanity (5th ed.). Georgia, USA: Mosby CO.
Davis, M. H. (1983). Measuring individual
differences in empathy: Evidence for a multidimensional approach. Journal of
Personality and Social Psychology, 44, 113-126.
Den Boer, H. (2013). Pre-trauma personality traits prospectively investigated in relation to PTSD symptoms in a military sample. (Master’s thesis, University of Amsterdam, Amsterdam,
Netherlands). Retrieved from http://www.scriptiesonline.uba.uva.nl/
Duijsens, I. J., Spinhoven P., Verschuur, M. J., & Eurelings-Bontekoe, E. H. M. (1999). De ontwikkeling van de Nederlandse Verkorte Temperament en Karakter Vragenlijst (TCI-105) [The development of the Dutch Shortened Temperament and Character Inventory
(TCI-105)]. Nederlands Tijdschrift voor de
Psychologie, 54, 276-283.
Duijsens, I.J., & Spinhoven, P. (2006).
Handleiding van de Nederlandse Verkorte Temperament en Karakter Vragenlijst. Leiderdorp, Netherlands: Datec.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior
Research Methods, 39, 175-191.
Field, A. (2009). Discovering statistics using
SPSS (3rd ed.). London, England: SAGE
Publications Ltd.
Frazier, P. A., Tix, A. P., & Barron, K. E.
(2004). Testing moderator and mediator effects in counseling psychology research. Journal of
Counseling Psychology, 51, 115-134.
Frick, P. J., Lilienfeld, S. O., Ellis, M., Loney, B., & Silverthorn, P. (1999). The association between anxiety and psychopathy dimensions in children. Journal of Abnormal Child Psychology,
27, 383-392.
Goff, A., Rose, E., Rose, S., & Purves, D.
(2007). Does PTSD occur in sentenced prison populations? A systematic literature review.
Criminal Behaviour and Mental Health, 17,
152-162.
Hale, L. R., Goldstein, D. S., Abramowitz, C. S., Calamari, J. E., & Kosson, D. S. (2004). Psychopathy is related to negative affectivity but not to anxiety sensitivity. Behaviour Research
and Therapy, 42, 697-710.
Hare, R. D. (2003). The Hare Psychopathy
Checklist-Revised (2nd ed.). Toronto, ON,
Canada: Multi-Health Systems. Hare, R. D., & Neumann, C. S. (2005). Structural Models of Psychopathy. Current
Psychiatry Reports, 7, 57-64.
Hare, R. D., & Neumann, C. S. (2008). Psychopathy as a clinical and empirical
construct. Annual Review of Clinical Psychology,
4, 217-246.
Hemphill, J. F. (2003). Interpreting the magnitudes of correlation coefficients. American
Psychologist, 58, 78-79.
Hicks, B. M., & Patrick, C. J. (2006).
Psychopathy and negative emotionality: Analyses of suppressor effects reveal distinct relations with emotional distress, fearfulness, and
14
anger-hostility. Journal of Abnormal Psychology,
115(2), 276-287.
Hovens, J. E., Bramsen, I., & Van Der Ploeg, H. M. (2000). De zelfinventarisatielijst
Posttraumatische Stressstoornis: ZIL-handleiding [Selfreport measure for PTSD symptoms: SIP-manual]. Lisse/Leiden,
Netherlands: Swets & Zeitlinger.
Hovens, J. E., Bramsen, I., & Van Der Ploeg, H. M. (2002). Self-rating inventory for post-traumatic stress disorder: Review of the psychometric properties of a new brief Dutch screening instrument. Perceptual and Motor
Skills, 94, 996-1008.
Hovens, J. E., Van Der Ploeg, H. M., Bramsen, I., Klaarenbeek, M. T. A., Schreuder, J. N., & Vladar Rivero, V. (1994). The development of a Dutch Self-Rating Inventory for Post-Traumatic Stress Disorder. Acta Psychiatrica Scandinavica,
90, 172-183.
Huijzer, D. (2015). Psychopathie en
persoonlijkheidstesten: is de VTCI in staat om psychopathie te meten? [Psychopathy and personalitytests: Is de TCI-SF capable of measuring psychopathy?] (Master’s thesis, University of Amsterdam, Amsterdam, Netherlands). Retrieved from http://scriptiesonline.uba.uva.nl/
Iacobucci, D., & Duhachek, A. (2003).
Advancing Alpha: Measuring reliability with confidence. Journal of consumer psychology,
13(4), 478-487.
Keane, T. M., Fairbank, J. A., Caddell, J. M., Zimmering, R. T., Tayler, K. L., & Mora, C. A. (1989). Clinical evaluation of a measure to assess combat exposure. Psychological Assessment: A
Journal of Consulting and Clinical Psychology, 1(1), 53-55.
Koopmans, J. A. (2009). The cortisol
awakening response and combat exposure during military deployment as predictors for burnout: Prospective Research into Stress in Military Operations (Master’s thesis, University Utrecht, Utrecht, Netherlands). Retrieved from
http://dspace.library.uu.nl/handle/1874/259493
Kubak, F. A., & Salekin, R. T. (2009).
Psychopathy and anxiety in children and adolescents: New insights on developmental
pathways to offending. Journal of
Psychopathology and Behavioral Assessment, 31,
271-284.
Levenson, M. R., Kiehl, K. A., & Fitzpatrick, C. M. (1995). Assessing psychopathic attributes in a noninstitutionalized population. Journal of
Personality & Social Psychology, 68, 151-158.
Mahmut, M. K., Menictas, C., Stevenson, R. J., & Homewood, J. (2011). Validating the factor structure of the Self-Report Psychopathy Scale in a community sample. Psychological Assessment,
23, 670-678.
Neumann, C. S., Hare, R. D., & Johansson, P. T. (2013). The Psychopathy Checklist-Revised (PCL-R), low anxiety, and fearlessness: A structural equation modeling analyses.
Personality Disorders: Theory, Research, and Treatment, 4(2), 129-137.
Orr, S., Metzger, L., Lasko, N. B., Macklin, M. L., Peri, T., & Pitman, R. K. (2000). De novo conditioning in trauma-exposed individuals with and without post-traumatic stress disorder.
Journal of Abnormal Psychology, 109(2),
290-298.
Ozer, E.J., Best, S.R.., Lipsey, T.L., & Weiss, D.S. (2008) Predictors of post-traumatic stress disorder and symptoms in adults a meta-analysis.
Psychological Bulletin, 129, 2-73.
Patton, J. H., Stanford, M. S., & Barratt, E. S. (1995). Factor structure of the Barratt Impulsiveness Scale. Journal of Clinical
Psychology, 51, 768-774.
Pham, T. (2012). Psychopathy and traumatic stress. Journal of Personality Disorders, 26(2),
2123-225.
Rademaker, A. R., Van Zuiden, M., Vermetten, E., & Geuze, E. (2011). Type D personality and the development of PTSD symptoms: A prospective study. Journal of Abnormal
Psychology, 120(2), 299-307.
Rademaker, A. R., Vermetten, E., Geuze, E., Muilwijk, A., & Kleber, R. J. (2008). Self-reported early trauma as a predictor of adult personality: A study in a military sample.
Journal of Clinical Psychology, 64, 863-875.
Raine, A., Dodge, K., Loeber, R., Kopp, L., Lynam, D., Reynolds, C., … Liu, J. (2006). The Reactive-Proactive Aggression (RPQ)
15
Questionnaire: Differential correlates of reactive and proactive aggression in adolescent boys.
Aggressive Behavior, 32, 159-171.
Reijnen, A., Geuze, E., & Vermetten, E. (2015). The effect of deployment to a combat zone on testosterone levels and the association with the development of post-traumatic stress symptoms: A longitudinal prospective Dutch military cohort study. Psychoneuroendocrinology, 51, 525-533. Sellbom, M. (2015). Elucidating the complex
associations between psychopathy and post-traumatic stress disorder from the perspective of trait negative affectivity. International Journal of
Forensic Mental Health, 14, 85-92.
Sijbrandij, M., Reitsma, J. B., Roberts, N. P., Engelhard, I. M., Olff, M., Sonneveld, L. P., & Bisson, J. I. (2013). Self-report screening instruments for post-traumatic stress disorder (PTSD) in survivors of traumatic experiences (Protocol). Cochrane Database of Systematic
Reviews 20-13, Issue 6. Art. No.: CD010575.
Snowden, R. J., & Gray, N. S. (2010).
Temperament and character as a function of psychopathy: Relationships between the Psychopathy Checklist – Revised and the Temperament and Character Inventory in a sample of personality disordered serious or repeat offenders. The Journal of Forensic
Psychiatry & Psychology, 21(6), 815-833.
Van Der Ploeg, H. M., Bramsen, I., & Reuling, I. E. W. (2000). Test-retest reliability of the self-rating inventory for post-traumatic stress disorder. Psychological Reports, 87, 735-737. Visser, B. A., Ashton, M. A., & Pozzebon, J. A.
(2011). Is low anxiety part of the psychopathy construct? Journal of Personality, 80, 726-747. Vitacco, M. J., Neumann, C. S., Jackson, R. L.
(2005). Testing a four-factor model of psychopathy and its association with ethnicity,
gender, intelligence, and violence. Journal of
Consulting and Clinical Psychology, 73(3),
466-476.
Watson, D., & Clark, L. A. (1984). Negative
affectivity: The disposition to experience aversive emotional states. Psychological Bulletin,
96, 465-490.
Williams, K. M., Paulhus, D. L., & Hare, R. D. (2007). Capturing the four-factor structure of psychopathy in college students via self-report.
Journal of Personality Assessment, 88(2),
205-219.
Willemsen, J., De Ganck, J., & Verhaeghe, P. (2012). Psychopathy, traumatic Exposure, and lifetime post-traumatic stress. International
Journal of Offender Therapy and Comparative Criminology, 56(4), 505-524.
Zelst, W. H. van, de Beurs, E., Beekman, A. T. F., Deeg, D. J. H., Bramsen, I., & van Dyck, R. (2003). Criterion validity of the self-rating inventory for posttraumatic stress disorder (SRIP) in the community of older adults. Journal
of Affective Disorders, 76, 229-235.
Zhang, S., Benson, T., & Deng, X. (2000). A test- retest reliability assessment of the international self-report delinquency instrument. Journal of
16
APPENDICES
Appendix A
Deployment Exposure Scale (DES)
Answer options : yes/no 1. Enemy fire *
(beschietingen niet op u gericht?) 2. Incoming fire *
(beschietingen op u gericht?) 3. Held at gunpoint *
(onder schot gehouden? Wapen op u gericht) 4. Held hostage *
(gijzeling van uzelf) 5. Colleague held hostage *
(gijzeling van collega’s uit uw eenheid) 6. Personal danger *
(persoonlijk gevaar oplopen ten gevolge van oorlogshandelingen, ongelukken of bedreiging) 7. Colleague injured or killed *
(gedood of gewond persoon binnen de eigen eenheid) 8. Physical injuries *
(zelf gewond geraakt) 9. Rejected by locals
(afwijzing door plaatselijke bevolking) 10. Witnessed people suffering *
(het zien van menselijk leed) 11. Witnessed dead *
(aanblik van doden) 12. Witnessed wounded *
(aanblik van ernstig gewonden) 13. Heard people screaming *
(gillen van gewonden) 14. Witnessed others injured/killed *
(getuige geweest van het sneuvelen of ernstig gewond raken van mensen) 15. Insufficient means to intervene
(onvoldoende mogelijkheden hebben om in te grijpen) 16. Insufficient control over situation
(geen controle over situatie hebben) 17. Mission experienced as useless
(het idee dat de missie zinloos was geworden) 18. Memories of earlier deployments
(herinneringen aan eerdere uitzendingen kwamen boven) 19. Traffic accident *
(verkeersongeval)
17
Appendix B
Details on the construction of the TCI-SF-P
As mentioned, the subscales of the TCI-SF-P are based on the correlation of TCI-SF items with the SRP-III subscales in a sample of 84 male participants (Huijzer, 2015). A two-tailed Spearman correlation test was used. Every TCI-SF item was added to one of the four SRP-III subscales based on the size of the correlation. The subscale with the highest correlation with that particular item was selected. Subsequently only those items with a significant (p ≤ .05) correlation of r ≥ .2 were selected. This resulted in the following subscales:
Interpersonal Manipulation (TCI-SF-IPM; 12 items): Cronbach’s alpha = .742.
TCI-SF item 3, 4, 6, 7, 18, 24, 25, 50, 63, 76, 78, 103.
Callous Affect (TCI-SF-CA; 17 items): Cronbach’s alpha = .749.
TCI-SF item 12, 23, 32, 33, 37, 40, 43, 47, 56, 62, 64, 81, 82, 91, 95, 99, 100.
Erratic Lifestyle (TCI-SF-ELS; 17 items): Cronbach’s alpha = .764.
TCI-SF item 1, 14, 26, 27, 31, 39, 42, 55, 66, 70, 75, 79, 86, 87, 88, 94, 97.
Antisocial Behavior (TCI-SF-ASB; 10 items): Cronbach’s alpha = .408.
TCI-SF item 16, 19, 22, 41, 48, 61, 73, 84, 92, 105.
During a consensus meeting with the supervisor, there is decided whether items should be deleted based on an improvement of the content validity with deletion and/or on an increase of Cronbach’s alpha with deletion. Having a closer look at the content of the items of the Interpersonal Manipulation subscale resulted in the conclusion to eliminate this scale. This is because the TCI-SF-IPM items did not seem to measure interpersonal manipulation as defined by Williams and colleagues (2007). The items, based on their content, do not capture characteristics such as pathological lying, conning and manipulating (Williams et al., 2007). However, some items of the abolished TCI-SF-IPM seem to fit the Callous Affect subscale based on their content. Looking at the correlation table again showed these items did not only have a significant (p ≤ .05) correlation of r ≥ .2 with TCI-SF-IPM but also with TCI-SF-CA. Therefore, it is decided to add these six items to the Callous Affect subscale. Other items are deleted from the TCI-SF-CA based on their content. These five deleted items did not seem to measure low empathy, a general lack of concern for other people and other deficiencies in affect including remorse and guilt (Williams et al., 2007). With regards to the Erratic Lifestyle subscale, 8 items are deleted based on their content. They did not seem to reflect unreliability, recklessness and impulsivity (Williams et al., 2007). Furthermore, 4 items from the deleted TCI-SF-IPM and TCI-SF-ASB scale are selected based on their content and added to the TCI-SF-ELS scale. As previously mentioned, the Antisocial Behavior scale is eliminated. The content of the TCI-SF-ASB items did not seem to measure antisocial behavior as defined by Williams and colleagues (2007). The TCI-SF-ASB items did not measure criminal tendencies. Finally, it is checked for a second time whether further deletion of items would result in a higher Cronbach’s alpha or a better content validity. Therefore, item 94 is deleted from the TCI-SF-ELS scale due to the higher Cronbach’s alpha and a better content validity. Other items which would result in a small increase of Cronbach’s alpha are not deleted based on their content. This resulted in the final two subscales:
Interpersonal Manipulation (TCI-SF-IPM): Deleted.
Callous Affect (TCI-SF-CA; 18 items): Cronbach’s alpha = .797.
TCI-SF item 6, 12, 18, 23, 24, 32, 37, 40, 43, 47, 50, 62, 63, 78, 81, 82, 95, 100.
Erratic Lifestyle (TCI-SF-ELS; 12 items): Cronbach’s alpha = .721.
TCI-SF item 1, 22, 26, 41, 42, 55, 66, 70, 75, 76, 86, 105.
18
Validation of the constructed psychopathy subscalesa. Convergent validity
In order to investigate the convergent validity of the Callous Affect (CA) subscale and the Erratic Lifestyle (ELS) subscale the consistency with the Levenson Self-Report Psychopathy Scale (LSRP; Table 1; Levenson, Kiehl, & Fitzpatrick, 1995) and the Self-Report Psychopathy Scales (SRP-III; Table 2; Williams et al., 2007) is examined. This is done using a two-sided Pearson correlation test. A generally accepted guideline for interpreting correlation coefficients is: weak = .1, moderate = .3 and strong = .5 (Hemphill, 2003).
Table 1
Correlations between the LSRP and the TCI-SF-P, inclusive of its subscales (N = 84)
TCI-SF-CA TCI-SF-ELS
LSRP I .614** .216*
LSRP II .241* .578**
LSRP Total .586** .437**
Note. LSRP = Levenson Self-Report Psychopathy Scale, * p < .05, **p < .01.
Table 2
Correlations between the SRP-III and the TCI-SF-P, inclusive of its subscales (N = 84)
TCI-SF-CA TCI-SF-ELS SRP IPM .528** .324** SRP CA .676** .201 SRP ELS .270* .714** SRP ASB .132 .553** SRP Total .526** .590**
Note. SRP-III = Self-Report Psychopathy Scales, * p < .05, **p < .01.
b. Criterion validity
Criterion validity is assessed by looking at the correlation between the CA and ELS subscales and four external criteria. These external criteria are measured using the Reactive-Proactive Aggressive Questionnaire (RPQ; Table 3; Raine, Dodge, Loeber, Kopp, Lynam et al., 2006), the Interpersonal Reactivity Index (IRI; Table 4; Davis, 1983), the Barratt Impulsiveness Scale (BIS; Table 5; Patton, Stanford, & Barratt, 1995) and the International Self-Report Delinquency (ISRD; Table 6; Zhang, Benson, & Deng, 2000).
Table 3
Correlations between the RPQ and the TCI-SF-P, inclusive of its subscales (N = 84)
TCI-SF-CA TCI-SF-ELS
RPQ Reactive -.023 .391**
RPQ Proactive .164 .505**
RPQ Total .051 .495**
Note. RPQ = Reactive-Proactive Agressive Questionnaire, * p < .05, **p < .01.
Table 4
Correlations between the IRI and the TCI-SF-P, inclusive of its subscales (N = 84)
TCI-SF-CA TCI-SF-ELS
19
IRI EC -.526** -.063
IRI PD .102 .037
IRI F -.174 .103
IRI Total -.372** -.004
Note. IRI = Interpersonal Reactivity Index, PT = Perspective Taking, EC = Empathic Concern, PD = Personal Distress, F = Fantasy, * p < .05, **p < .01.
Table 5
Correlations between the BIS and the TCI-SF-P, inclusive of its subscales (N = 84)
TCI-SF-CA TCI-SF-ELS
BIS Attention .156 .670**
BIS Motor .027 .629**
BIS Planning -.049 .726**
BIS Total .042 .788**
Note. BIS = Barratt Impulsiveness Scale, * p < .05, **p < .01.
Table 6
Correlations between the ISRD and the TCI-SF-P, inclusive of its subscales (N = 84)
TCI-SF-CA TCI-SF-ELS
ISRD -.097 -.499**
Note. ISRD = International Self-Report Delinquency, * p < .05, **p < .01.
Based on the above results, the CA and ELS subscales are used in the current study as a proxy for the psychopathy aspects (i.e. Affective facet and Lifestyle facet). For future research it is recommended to perform extensive validating research on the CA and ELS subscales in order to underpin whether or not these subscales measure the Affective and Lifestyle facet of psychopathy.
