Assessment of dark personalities in Iran: Psychometric evaluation of the Farsi translation of the Short Dark Triad (SD3-F)

Personality and Individual Differences 102 (2016) 111–117

Contents lists available at ScienceDirect

Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid

Assessment of dark personalities in Iran: Psychometric evaluation of the Farsi translation of the Short Dark Triad (SD3-F) Mohammad Atari a,⁎, Razieh Chegeni b a b

Department of Psychology, University of Tehran, Tehran, Iran Department of Psychology, Alzahra University, Tehran, Iran

a r t i c l e

i n f o

Article history: Received 24 April 2016 Received in revised form 29 June 2016 Accepted 29 June 2016 Available online xxxx Keywords: Dark Triad Machiavellianism Narcissism Psychopathy Iran

a b s t r a c t The Dark Triad traits- Machiavellianism, narcissism, and psychopathy- have attracted much research attention in the past decade. The Short Dark Triad (SD3) is one of the commonly used short measures of the Dark Triad. In the present study, we translate and adapt the SD3 into Farsi. Two studies (total N = 523) examined the factor structure, internal consistency, and convergent validity of the Farsi SD3. Study 1 suggested that only 20 items had sufficient psychometric quality to be subjected to factor analysis. The subsequent exploratory factor analysis suggested a three-factor solution. All SD3 subscales had adequate convergent validity. Furthermore, internal consistency coefficients of the Farsi SD3 subscales were marginally acceptable with the exception of Machiavellianism subscale for men. Finally, Study 2 supported the 3-factor structure of the 20-item SD3 using confirmatory factor analysis. Therefore, the 20-item SD3-F may be used with marginally acceptable reliability and validity to assess the Dark Triad of personality in Iranian context. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Paulhus and Williams (2002) proposed a concept called the “Dark Triad”. This individual-difference construct is comprised of narcissism, Machiavellianism, and psychopathy. Since introduction of the Dark Triad traits, a relatively large body of literature has developed. A sharp increase has occurred in the number of studies investigating the utility of these traits (Furnham, Richards, & Paulhus, 2013). The Dark Triad of personality has been successful at predicting many social and inter-personal behaviors such as aggressive behavior (Baughman, Dearing, Giammarco, & Vernon, 2012), short-term mating (Jonason, Li, Webster, & Schmitt, 2009), and workplace behaviors (O′Boyle, Forsyth, Banks, & McDaniel, 2012). It has been reported that the similarities between narcissism, Machiavellianism, and psychopathy are rooted in a common interpersonal callousness (Jones & Figueredo, 2013). Callousness may be the key in understanding why the Dark Triad traits are considered as socially aversive (Rauthmann, 2012). Narcissism has a well-developed literature in psychology (Miller, Gentile, Wilson, & Campbell, 2013). Traditional perspective on narcissism suggests that underlying characteristics of narcissism include manipulation, exhibitionism, entitlement, vanity, arrogance, over-bearing, and self-sufficiency (Raskin & Terry, 1988). The research literature on narcissism has considerably relied on the Narcissistic Personality Inventory (NPI; Raskin & Terry, 1988) as the primary measure of narcissism ⁎ Corresponding author at: Faculty of Psychology and Educational Sciences, Guisha, AleAhmad Bridge, Tehran, Iran. E-mail addresses: [email protected], [email protected] (M. Atari).

http://dx.doi.org/10.1016/j.paid.2016.06.070 0191-8869/© 2016 Elsevier Ltd. All rights reserved.

(Cain, Pincus, & Ansell, 2008). More recently, Foster, McCain, Hibberts, Brunell, and Johnson (2015) developed and validated Grandiose Narcissism Scale (GNS) which captures Raskin and Terry's (1988) seven-factor model in a psychometrically sound manner. While there are various types of narcissism (e.g., pathological narcissism), it is grandiose narcissism that is most relevant to conceptualization of the Dark Triad. Machiavellianism was formally introduced to the literature by Christie and Geis (1970). The construct was conceptualized primarily based on writings of Niccolo Machiavelli, an Italian political strategist and philosopher. Christie and Geis (1970) also introduced five measurement tools for assessment of Machiavellian personality (i.e., MACH-I through MACH-V). Of these measures, MACH-IV received much more attention (Fehr, Samsom, & Paulhus, 1992). Items of MACH-IV measure constructs such as lack of morality, cynical worldview, negative view on human nature, and manipulativeness. Jones and Paulhus (2009) added three themes to the Machiavellian personality conceptualization (i.e., planning, coalition-formation, and reputation-building). These three themes are rooted in ideas of Sun Tzu, a first-century military strategist. These three themes have been found to be distinguishing among the Dark Triad traits (Paulhus & Jones, 2014). Psychopathy is strongly associated with traits such as pathological lying, superficial charm, egocentricity, lack of remorse, and callousness. Modern conception of psychopathic personality roots in the work of Cleckley (1941). Self-control problems and callousness were main characteristics of his conception. Psychopathy is conceptually close to Machiavellianism; however, Machiavellians plan ahead, form coalitions, and try to maintain a positive reputation. On the contrary, psychopaths are more impulsive and this impulsivity, paired with callous manipulation,

112

M. Atari, R. Chegeni / Personality and Individual Differences 102 (2016) 111–117

results in a life-time of criminal behavior (Hare & Neumann, 2008). Moreover, psychopaths pay little attention to the impact of their behaviors. This feature seems to be distinguishing between psychopathy and Machiavellianism (Paulhus & Jones, 2014). The 64-item Self-Report Psychopathy (SRP; Paulhus, Neumann, & Hare, 2014; Neal & Sellbom, 2012) scale is said to be the most popular measure of psychopathy. Overall, the body of literature suggests that: (1) ego-enhancement goals and plans drive narcissistic behavior, while instrumental goals drive Machiavellian and psychopathic behaviors; (2) impulsivity distinguishes between Machiavellianism and psychopathy; (3) all Dark Triad traits have a callous core that contributes to interpersonal manipulation (Jones & Paulhus, 2011a). Early works on the Dark Triad traits used previously mentioned measures for each trait independently. Yet, a practical limitation was that using the well-established measures of narcissism, Machiavellianism, and psychopathy (i.e., NPI, MACH-IV, and SRP) required a total of 125 items. Even the shortest available versions of these measures (NPI-16, MACH-IV, and SRP-29) added up to 65 items (Paulhus & Jones, 2014). Therefore, recent research studies have turned to using shorter combinational instruments that psychometrically capture the Dark Triad. There are two popular measures for assessment of the Dark Triad traits (Paulhus & Jones, 2014), i.e., the Dirty Dozen (Jonason & Webster, 2010) and the Short Dark Triad (SD3) scale (Jones & Paulhus, 2014). The Dirty Dozen (DD) is a 12-item scale comprising three 4item subscales. Internal consistency, test-retest stability, factor structure, and convergent validity of the DD have been found to be satisfactory (Jonason & Webster, 2010; Jonason, Li, & Czarna, 2013; Webster & Jonason, 2013). The nomological network and bifactor model of the DD have been confirmed in other cultures (Czarna, Jonason, Dufner, & Kossowska, 2016). Though, Miller et al. (2012) suggested that DD lacks essential content. It has been also reported that high facet-level reliability coefficients are due to repetitive wording of items. The DD Mach subscale has conceptual issues (c.f., Jones & Paulhus, 2014). Furthermore, grandiose narcissism has 7 dimensions and, unfortunately, DD Narcissism subscale has 4 items only. In sum, DD shows satisfactory psychometric properties; however, it has conceptual issues which are inconsistent with the original conception of the Dark Triad (Jones & Paulhus, 2014). The SD3 has been developed in accordance with latest findings in the Dark Triad literature and original conception of the Dark Triad. Four distinct studies were conducted to examine different psychometric aspects of the measure. The SD3 was developed as a 27-item scale comprising three 9-item subscales which are labeled according to classic Dark Triad traits. All subscales of the SD3 proved to be internally consistent across community and student samples (Jones & Paulhus, 2014). Furthermore, all subscales were significantly correlated with relevant measures of the same construct. In regards to sex differences, men consistently scored higher in all subscales of the SD3. Finally, Jones and Paulhus (2014) concluded that the SD3 met psychometric standards while capturing the original conceptions of the Dark Triad traits. The Brevity of the SD3 was also noted as a point of strength. Maples, Lamkin, and Miller (2014) compared the SD3 and the DD and concluded that the SD3 would yield more consistent data. It is important to examine the psychometric properties of the SD3 in different cultural contexts, particularly for cross-cultural comparisons. Most of the previous studies on the utility of the SD3 have been conducted in Western cultures. The present study aimed to examine the psychometric properties of the Farsi SD3 (i.e., SD3-F) in Iranian context. Iran may generally provide a useful population for investigation of the psychometric properties of SD3 in a non-Western setting. We hypothesized that the SD3-F would also replicate the 3-factor model of the Dark Triad (Jones & Paulhus, 2014). The factor structure, convergent validity, and internal consistency of the Farsi SD3 were assessed in community and student samples. Additionally, sex differences were examined and, in line with previous research, we hypothesized that men would score higher in the Dark Triad traits.

2. Study 1 In this study, we provide a Farsi translation of the SD3 using a methodologically robust technique. Then we administer the Farsi translation of the SD3 on a relatively large sample along with three standard measures of Machiavellianism, narcissism, and psychopathy as well as the Dirty Dozen. We conduct an initial item analysis followed by an exploratory factor analysis and evaluation of sex differences and convergent validity. In this study, we also draw a comparison between the SD3-F and Dirty Dozen as alternative measures of the Dark Triad. Reliability coefficients are also discussed in men and women. 3. Method 3.1. Participants Participants consisted of 301 adults recruited from six major universities in Tehran, Iran (Alzahra University, Tehran University of Medical Sciences, Amirkabir University of Technology, Islamic Azad University, University of Tehran, and Tarbiat Modarres University). All participants identified themselves as Iranian. Participants ranged in age between 18 and 56 (M = 25.57, SD = 6.68), 57.8% were male, and 75.4% were single. In terms of highest educational degree, 77 participants had a high school diploma, 168 participants had a bachelor's degree, 42 participants had a master's degree, 2 participants had a doctorate degree, and 12 participants did not report their educational background. 3.2. Measures 3.2.1. Short Dark Triad (SD3) The SD3 (Jones & Paulhus, 2014) has 27 items which are rated along a Likert-type scale ranging from 1 (Strongly disagree) to 5 (Strongly agree). For purposes of the present study, all items were translated into Farsi (a.k.a., Persian) by two professional translators. Then, two independent translators back-translated all items back into English. Therefore, aside from the original scale, two Farsi versions and two back-translations were available. Consequently, authors and one professional translator settled the differences and selected the best Farsi translation for each item. Finally, the preliminary Farsi version of the scale was administered on a sample of 20 university students in a discussion group. Some suggestions were made and three small revisions were made on three items. Therefore, the final Farsi translation of the scale was prepared to be administered. 3.2.2. Dirty Dozen (DD) Dirty Dozed was used in this study as an alternative measure of the Dark Triad. Response options are provided in a 7-point Likert-type scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). Previous studies support its psychometric characteristics (Webster & Jonason, 2013). The Farsi translation of the DD has shown adequate validity and reliability in previous work (Aghababaei, Mohammadtabar, & Saffarinia, 2014). In the present study, alpha coefficients for Machiavellianism, psychopathy, and narcissism were 0.83, 0.61, and 0.85, respectively. 3.2.3. Grandiose Narcissism Scale (GNS) The 33-item GNS is a newly developed measure of grandiose narcissism. The factor structure of the GNS is consistent with Raskin and Terry (1988) and it has shown excellent psychometric properties (Foster et al., 2015). Items are rated on a 6-point Likert-type scale ranging from 1 (Strongly disagree) to 6 (Strongly agree). This scale consists of seven subscales (authority, self-sufficiency, superiority, vanity, exhibitionism, entitlement, and exploitativeness). The Farsi version of this scale has shown acceptable reliability and validity (Atari, 2016). In the present study, internal consistency coefficients ranged from 0.74 to 0.90.

M. Atari, R. Chegeni / Personality and Individual Differences 102 (2016) 111–117

113

Table 1 Descriptive characteristics of all 27 items (N = 301). Item

M

SD

Skewness

Kurtosis

Corrected item-total correlation

Alpha if item deleted (α = 0.77)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

4.07 1.55 2.87 3.61 2.05 2.80 3.98 3.51 2.34 2.74 3.44 2.85 2.93 4.13 3.04 2.45 2.70 4.30 1.74 2.76 2.08 2.28 2.21 2.55 2.57 1.94 2.15

1.20 0.97 1.23 1.16 1.30 1.54 1.17 1.32 1.40 1.21 1.21 1.20 1.25 0.97 1.30 1.25 1.36 1.03 1.12 1.35 1.26 1.22 1.32 1.29 1.38 1.36 1.33

−1.22 1.84 −0.09 −0.59 0.99 0.13 −1.05 −0.45 0.57 −0.04 −0.41 0.05 −0.15 −1.10 0.12 0.32 0.19 −1.62 1.39 0.22 0.88 0.48 0.59 0.31 0.37 1.14 0.87

0.48 2.65 −0.99 −0.46 −0.27 −1.46 0.23 −1.00 −1.04 −1.01 −0.68 −0.74 −0.97 0.92 −1.09 −0.93 −1.24 2.14 0.93 −1.18 −0.30 −0.88 −1.12 −1.02 −1.11 −0.12 −0.51

−0.04 0.35 0.39 0.18 0.44 0.43 0.17 0.13 0.43 0.38 0.07 0.45 0.49 0.16 0.05 0.40 0.24 0.12 0.27 0.05 0.37 0.42 0.41 0.42 0.21 0.37 0.36

0.78 0.76 0.75 0.76 0.75 0.75 0.76 0.77 0.75 0.75 0.77 0.75 0.75 0.76 0.77 0.75 0.76 0.76 0.76 0.77 0.75 0.75 0.75 0.75 0.76 0.75 0.75

3.2.4. MACH-IV MACH-IV consists of 20 items which are rated on a 7-point Likerttype scale (Christie & Geis, 1970). Response options range from 1 (Strongly disagree) to 7 (Strongly agree). Previous studies have suggested different factor structures for the MACH-IV in different languages and different settings (Hunter, Gerbing, & Boster, 1982; Rauthmann, 2013; Williams, Hazleton, & Renshaw, 1975). The Farsi adaptation of the MACH-IV showed that only 17 items had adequate psychometric properties (Khodabakhsh, Chegeni, & Atari, 2016). Moreover, a factor

analysis on an Iranian sample revealed a 2-factor structure (i.e., cynical worldview and tendency to manipulation). In the present study, alphas for cynical worldview and tendency to manipulation were 0.73 and 0.66, respectively. 3.2.5. Short Four-Factor Psychopathy Measure (SFFPM) In order for brevity, we needed a relatively short psychopathy measure. Consistent with the four-factor model of psychopathy, Williams, Paulhus, and Hare (2007) developed a self-report measure for college

Table 2 The pattern matrix of the 22 items subjected to factor analysis (N = 301). Item

SD21 SD23 SD5a SD26 SD19 SD6 SD24 SD22 SD9a SD27 SD25 SD13 SD12 SD17 SD10 SD16 SD11 SD4 SD3 SD7 SD1 SD8 Eigenvalue Explained variance

h2

Factor

Payback needs to be quick and nasty. It's true that I can be mean to others. It's wise to keep track of information that you can use against people later. I enjoy having sex with people I hardly know I like to get revenge on authorities. You should wait for the right time to get back at people. People who mess with me always regret it. People often say I′m out of control. Most people can be manipulated. I′ll say anything to get what I want. I have never gotten into trouble with the law.(R) I know that I am special because everyone keeps telling me so. Many group activities tend to be dull without me. I am an average person.(R) People see me as a natural leader. I have been compared to famous people. I hate being the center of attention. (R) Avoid direct conflict with others because they may be useful in the future. Whatever it takes, you must get the important people on your side. There are things you should hide from other people to preserve your reputation. It's not wise to tell your secrets. Make sure your plans benefit yourself, not others.

F1

F2

F3

0.71 0.59 0.53 0.52 0.46 0.43 0.43 0.42 0.37 0.32 0.14 −0.01 0.00 −0.10 0.02 0.06 0.00 −0.14 0.11 0.06 −0.26 0.04 4.31 19.57

−0.10 0.11 −0.06 −0.07 −0.05 −0.08 0.14 0.27 0.15 −0.02 0.12 0.70 0.64 0.60 0.58 0.58 0.10 0.05 0.11 −0.12 0.02 −0.05 1.95 8.85

−0.13 −0.20 0.14 0.03 −0.06 0.35 0.04 −0.12 0.04 0.24 0.01 0.08 0.10 −0.12 −0.01 −0.03 −0.07 0.47 0.42 0.40 0.34 0.24 1.61 7.30

Note. Loadings N0.3 are bolded; F1 = psychopathy; F2 = narcissism; F3 = Machiavellianism; Reversals are indicated with (R). a Item was discarded.

0.41 0.36 0.33 0.26 0.18 0.38 0.27 0.30 0.22 0.20 0.05 0.53 0.46 0.30 0.35 0.36 0.01 0.20 0.27 0.17 0.12 0.06

114

M. Atari, R. Chegeni / Personality and Individual Differences 102 (2016) 111–117

Table 3 Sex differences in the SD3-F subscales (N = 301). Subscale

SD3-M SD3-N SD3-P

Men

Women

M

SD

M

SD

21.07 17.07 19.13

3.90 4.73 5.86

20.47 17.17 15.33

4.32 4.71 4.81

t-Test statistic

Cohen's d

1.24 −0.17 5.97⁎

0.15 −0.02 0.66

Note. p-Values were Bonferroni-corrected (0.05 / 3 = 0.02); M = Machiavellianism; N = narcissism; P = psychopathy. ⁎ p b 0.02.

students. Items of this scale are rated on a 5-point Likert-type scale ranging from 1 (Strongly disagree) to 5 (Strongly agree). They also presented the 5 highest loading items (Williams et al., 2007) for each factor. Chegeni and Atari (2016) translated these 20 high-loading items into Farsi and factor-analyzed them in order to reach a short four-factor measure of psychopathy. The subscales are labeled as interpersonal manipulation, erratic lifestyle, criminal tendencies, and callous affect. Results indicated a robust four-factor structure reflecting the four-factor model of sub-clinical psychopathy. In the present study, all subscales were internally consistent (0.59 b αs b 0.79). 3.3. Procedure The university's ethics committee approved this study. The aforementioned measures were presented in a questionnaire package in a randomized order. All packages started with demographic questions (age, sex, marital status, and educational background). Potential participants were approached and invited to take part in a psychological study about social behavior. Upon agreement to take part, participants completed the questionnaires. Participation was on a voluntary basis and participants were not remunerated. 3.4. Statistical analysis As proposed by Kline (2005), an item analysis was performed on all items prior to factor analysis. That is, descriptive statistics (including mean, standard deviation, skewness, kurtosis, and item-total correlation) of all items were evaluated. Items which violated adequate qualities were considered for omission before subjecting the whole scale to factor analysis. The factor structure of the SD3 was examined by a principal-axis Exploratory Factor Analysis (EFA) with promax rotation. Since no previously reported cross-cultural evidence in underlying factor structure of the SD3 is available, the three-factor structure of the SD3 (Jones & Paulhus, 2014) was not assumed. Parallel analysis (c.f., Hayton, Allen, & Scarpello, 2004) was used to determine the number of factors in EFA. Pearson correlation coefficients were calculated to examine the convergent validity of the SD3-F as well as subscales' inter-correlations. Finally, three distinct t-tests were used to examine potential sex differences.

4. Results and discussion A comprehensive item analysis was performed on all 27 items before factor analysis. Floor effect, ceiling effect, mean, standard deviation, skewness, kurtosis, corrected item-total correlation, and alpha-if-itemdeleted index were checked to identify potentially problematic items. These statistics are presented in Table 1. It can be seen that items 1, 7, 14, and 18 are candidates for ceiling effect (|skewness| N 1). Moreover, items 2, 19, and 26 are candidates for floor effect (|skewness| N 1). Items 2, 6, 17, 18, and 20 had high kurtosis. Furthermore, items 1, 11, 15, and 20 have low corrected item-total correlations (r b 0.1). Finally, deletion of items 1, 8, 11, 15, 18, and 20 would have increased the full-scale internal consistency (Cronbach's alpha). Consequently, items 2, 14, 15, 18, and 20 were discarded prior to further analyses. A principal-axis EFA was performed on the remaining 22 items. Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was 0.774 and Bartlett's test of sphericity was significant (χ2 (231) = 1219.79, p b 0.001). Initially, seven factors had eigenvalues N1 (λ1 = 4.31, λ2 = 1.947, λ3 = 1.61, λ4 = 1.22, λ5 = 1.11, λ6 = 1.05, and λ7 = 1.01); however, a parallel analysis on 1000 random data sets with 95% confidence interval indicated that only three of the factors should be retained (λ1 = 1.61 [accepted], λ2 = 1.50 [accepted], λ3 = 1.42 [accepted], λ4 = 1.34 [rejected]). Consequently, a promax-rotated EFA was performed with a fixed number of three factors on the 22 items. The factor structure of these items is presented in Table 2. The three-factor solution accounted for 35.72% of the total variance. We labeled the factors according to the content of items and the original factor structure reported in Jones and Paulhus (2014). It can be seen that items 5 and 9 are clearly grouped with psychopathy items. Therefore, these items were discarded in this step. Items 25 and 11 had weak loadings on their corresponding factors. Also, item 6 loaded on psychopathy and Machiavellianism factors. Yet, we did not delete these items as we aimed to conserve as many items as possible. As a result, in this phase, a 20-item version of the SD3 was formed. Items 21, 23, 26, 19, 24, 22, 27, and 25 were grouped together and were collectively labeled as the psychopathy subscale. Items 13, 12, 17, 10, 16, and 11 were grouped together and were collectively labeled as the narcissism subscale. Finally, items 4, 3, 7, 6, 1, and 8 were grouped

Table 4 Convergent validity of the Farsi SD3 subscale in men and women (N = 301). Measure

DD-M DD-N DD-P DD-total score MACH-IV GNS SFFPM

Men

Women

SD3-M

SD3-N

SD3-P

SD3-M

SD3-N

SD3-P

0.21⁎⁎ 0.15⁎ 0.15 0.24⁎⁎ 0.24⁎⁎ 0.37⁎⁎ 0.19⁎

0.16⁎ 0.24⁎⁎ 0.26⁎⁎ 0.29⁎⁎ 0.18⁎ 0.55⁎⁎ 0.21⁎⁎

0.36⁎⁎ 0.17⁎ 0.45⁎⁎ 0.56⁎⁎ 0.48⁎⁎ 0.48⁎⁎ 0.56⁎⁎

0.20⁎ 0.31⁎⁎ 0.29⁎⁎ 0.38⁎⁎ 0.39⁎⁎ 0.39⁎⁎ 0.19⁎

0.21⁎ 0.45⁎⁎ 0.20⁎ 0.39⁎⁎ 0.23⁎ 0.52⁎⁎ 0.22⁎

0.30⁎⁎ 0.09 0.32⁎⁎ 0.22⁎ 0.40⁎⁎ 0.45⁎⁎ 0.59⁎⁎

Note. M = Machiavellianism; N = narcissism; P = psychopathy; DD = Dirty Dozen; GNS = Grandiose Narcissism Scale; SD3 = Short Dark Triad; SFFPM = Short Four-Factor Psychopathy Measure. ⁎ p b 0.05. ⁎⁎ p b 0.01.

M. Atari, R. Chegeni / Personality and Individual Differences 102 (2016) 111–117

115

Table 5 Inter-correlations and reliabilities of the SD3-F subscales (N = 301). Men

M N P

Women

M

N

P

M

N

P

α = 0.40 0.13 0.25⁎⁎

– α = 0.69 0.37⁎⁎

– – α = 0.66

α = 0.62 0.19⁎ 0.28⁎⁎

– α = 0.70 0.34⁎⁎

– – α = 0.63

Note. M = Machiavellianism; N = narcissism; P = psychopathy. ⁎ p b 0.05. ⁎⁎ p b 0.01.

together and were collectively labeled as the Machiavellianism subscale. In order to evaluate sex differences, we calculated the descriptive statistics for the 20-item SD3-F scale. Men had higher scores in Machiavellianism, narcissism, and psychopathy. Three distinct t-tests were used to examine sex differences in the SD3-F subscales (see Table 3). The values of effect size indicated moderate to large differences. For evaluation of the concurrent validity of the SD3-F, the correlation coefficients were calculated between the SD3-F subscales and concurrent standard measures of Machiavellianism, narcissism, and psychopathy. Dirty Dozen was also included in this analysis as an alternative measure of the Dark Triad. Note from Table 4 that each SD3-F subscale is strongly associated with its standard counterpart. Facet-level associations also indicated that the Farsi SD3 subscales had strong associations with their corresponding measures. Inter-correlations and reliabilities of the SD3-F subscales are presented in Table 5. Table 6 displays the convergent validity of the SD3-F and the Farsi DD against standard measures of Machiavellianism, narcissism, and psychopathy. It can be seen that, in general, both measures have adequate associations with standard measures. More specific investigation of the correlation coefficients would reveal that the MACH subscale of the DD has higher convergent validity compared to Machiavellianism subscale of the SD3-F. This can be explained by the notion that Machiavellianism subscale of the SD3 is based on new conception of Machiavellianism which consists of newer themes (e.g., coalition formation). Yet, narcissism and psychopathy subscales of the SD3-F have better convergent validities. 5. Study 2 In order to evaluate the confirmatory factor structure of the Farsi SD3, we secured data from 222 participants. We compared three alternate models: a. the g-factor model which assumes that one general factor underlies the whole 27 items; b. the original model which has been proposed by Jones and Paulhus (2014); c. the EFA model which has been extracted in Study 1. Fundamentally, Confirmatory Factor Analysis (CFA) is used to determine if a measure's factor structure derived from exploratory factor analytic approaches can hold up with another respondent sample (Mvududu & Sink, 2013). 6. Method An independent sample of 222 participants was recruited from university settings in Tehran, Iran. Potential participants were approached by the same examiner in Study 1 and asked to take part in a psychological study regarding social behavior. The university's ethics committee approved this study. All participants provided consent before completing the questionnaires. Participants ranged in age between 18 and 37 (M = 24.59, SD = 4.16). Moreover, 54.1% of the participants were male and 80.2% were single. In terms of highest educational degree, 17 participants had a high school diploma, 94 participants had a bachelor's degree, 73 participants had a master's degree, 26 participants had a doctorate degree, and 12 participants did not report their educational

background. All participants completed the Farsi translation of the SD3 (see Study 1). We used CFA to examine the previously identified factor structure of the scale (see Study 1). All three factors were permitted to co-vary. As fit indices for the CFA, the Chi-square (χ2), the Chi-square over degree-offreedom (χ2/df), the Root Mean Square Error of Approximation (RMSEA), the Tucker–Lewis Index (TLI), Goodness-of-Fit Index (GFI), and the Comparative Fit Index (CFI) were analyzed (Hu & Bentler, 1999). Maximum likelihood was used as the estimation method. The analysis was performed using AMOS 19. 7. Results and discussion We compared three alternative models in terms of model fit. The models' fit indices are presented in Table 7. The first model assumed that a general factor would account for the variance of all 27 items. As expected, this model did not have satisfactory values. The second model was the original 27-item model with three inter-related 9-item subscales. Some indices of this model (e.g., χ2/df and RMSEA) fell within acceptable range; however, other indices (e.g., GFI and CFI) were not acceptable. The third model was based upon Study 1 and had 20 items. It had three inter-related subscales. Most of the fit indices for this model were either completely or marginally acceptable. The two latter models were compared using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) indices. Generally, AIC and BIC are used to compare two models estimated from the same data set, with the model yielding lower values to be preferred. Hereby, the third 20-item model had lower AIC and BIC compared with alternate models (see Table 7). As a result, and consistent with Study 1, the third model was retained to explain the factor structure of the SD3-F in Iran. After confirming the 3-factor 20-item model of the SD3-F, we calculated internal consistency coefficients for Machiavellianism, narcissism, and psychopathy in Study 2 sample. For men, alpha coefficients were 0.50, 0.65, and 0.56 respectively. For women, alpha coefficients were 0.61, 0.69, and 0.60 respectively. All subscales were internally more consistent in women compared to men. 8. General discussion The present research aimed to examine the psychometric properties of the Farsi translation of the SD3 in two distinct samples from Table 6 Convergent validities of the SD3-F and the Farsi DD against standard measures (N = 301). Standard measure

MACH-IV SFFPM GNS

SD3-F

DD

M

N

P

M

N

P

0.31⁎⁎ 0.20⁎⁎ 0.39⁎⁎

0.20⁎⁎ 0.19⁎⁎ 0.53⁎⁎

0.47⁎⁎ 0.63⁎⁎ 0.49⁎⁎

0.49⁎⁎ 0.52⁎⁎ 0.38⁎⁎

0.32⁎⁎ 0.18⁎⁎ 0.43⁎⁎

0.41⁎⁎ 0.47⁎⁎ 0.32⁎⁎

Note. M = Machiavellianism; N = narcissism; P = psychopathy; GNS = Grandiose Narcissism Scale; SFFPM = Short Four-Factor Psychopathy Measure; DD = Dirty Dozen; SD3 = Short Dark Triad. ⁎⁎ p b 0.01.

116

M. Atari, R. Chegeni / Personality and Individual Differences 102 (2016) 111–117

Table 7 Model fit indices for alternate models of the SD3-F (N = 222). Model

χ2

χ2/df

GFI

TLI

CFI

RMSEA

AIC

BIC

g-Factor model Original model EFA model

786.36 525.72 240.32

2.43 1.71 1.51

0.76 0.85 0.91

0.39 0.70 0.81

0.44 0.73 0.84

0.080 0.057 0.048

894.36 665.72 342.33

1078.10 903.90 515.86

Note. GFI = Goodness-of-Fit Index; TLI = Tucker-Lewis Index; CFI = Comparative Fit Index; RMSEA = Root Mean Square Error of Approximation; AIC = Akaike Information Criterion; BIC = Bayesian Information Criterion.

Iran. The first study suggested that only 22 items had sufficient psychometric quality to be subjected to EFA. These 22 items were subjected to an EFA and two items (5 and 9) were not loaded on their corresponding subscale. Therefore, two more items were discarded, leaving a final battery of 20 items. Three underlying factors were named Machiavellianism, narcissism, and psychopathy. Study 2 compared three different structural models. The first model hypothesized that a general factor underlies the whole 27 items. The second model was that of the original report (Jones & Paulhus, 2014). The third model was based upon the Study 1 in the present report. The third model showed better fit, indicating that the 20-item SD3-F explains variance of the items better than other alternative models. Our two studies suggest that the 20-item Farsi adaptation of the SD3 achieves an optimal outcome in terms of factor structure and internal consistency. Overall, 7 items were discarded in Farsi adaptation of SD3. Three of these items belonged to Machiavellianism subscale (2. I like to use clever manipulation to get my way; 5. It's wise to keep track of information that you can use against other people later; and 9. Most people can be manipulated). Three of these items belonged to narcissism subscale (14. I like to get acquainted with important people; 15. I feel embarrassed if someone compliments me; and 18. I insist on getting the respect I deserve). Only one item belonged to psychopathy subscale (20. I avoid dangerous situations). The psychometric insufficiency of these items (2, 5, 9, 14, 15, 18, and 20) may have different explanations. First, the present translation may not be appropriate; however, considering the methodologically robust technique in our translation, it seems very unlikely. Second, a cross-cultural factor may underlie these differences. For example, item 14 cannot necessarily be considered as “dark” in Iranian culture. In less developed countries, being acquainted with important people may be not only desirable, but also necessary for one's economic survival. Therefore, the unique role of culture should be taken into account in cross-cultural adaptation of personality measures (c.f., Hofstede & McCrae, 2004). As another example, structure of narcissism, as a member of the Dark Triad, has been shown to be different in collectivistic and individualistic societies (Foster, Campbell, & Twenge, 2003). Of note, cross-loadings of items in Machiavellianism and psychopathy may be due to conceptual similarity between the two constructs (Paulhus & Jones, 2014). It has been suggested that impulsivity makes the distinction between Machiavellianism and psychopathy (Jones & Paulhus, 2011b). To our knowledge, this report is the first to adapt the SD3 in a nonWestern society. Therefore, we cannot draw a comparison between different cultural contexts. Previous work suggests that some measures may have a different factor structure in non-Western cultures such as Iran (e.g., Atari, Akbari-Zardkhaneh, Mohammadi, & Soufiabadi, 2015). It is strongly recommended for future research to examine the factorial validity of the SD3 in different cultural contexts. Establishing cross-cultural factorial validity and invariance of psychometric measures would make cross-cultural comparisons possible. Study 1 also supported convergent validity of the SD3-F. All SD3-F subscales were significantly correlated with a convergent measure. That is, Machiavellianism subscale was significantly associated with the MACH-IV; narcissism subscale was significantly associated with the GNS; and psychopathy subscale was significantly associated with the SFFPM. These correlations support the convergent validity of the

SD3 in Iran. We also evaluated sex differences using the SD3-F. Men scored significantly higher in psychopathy subscale of the SD3-F. The other subscales did not show significant sex differences. Additionally, all subscales of the SD3-F were moderately inter-related. All alpha coefficients were marginally acceptable in men and women (0.62 b αs b 0.70), except for Machiavellianism subscale in men (α = 0.40) which was substantially below the acceptable range. Low internal consistency coefficients may be explained by the non-redundant item generation in development of the SD3 (Jones & Paulhus, 2014), accompanied by exclusion of several items in the Farsi adaptation. We also drew a comparison between the 20-item SD3-F and 12item DD in terms of convergent validity. The Machiavellianism subscale of DD had better convergent validity, as it could correlate strongly with a standard measure of Machiavellianism. It was also less strongly associated with psychopathy. Yet, psychopathy and narcissism subscales of the SD3-F showed stronger convergent validity compared with DD subscales. The DD subscales were internally more consistent in this study. This can be explained by similar wording of items in the DD subscales. Study 2 supported the 3-factor structure of the 20-item SD3-F in a distinct sample. The fit indices suggested the factor structure extracted from Study 1 had moderate factor loadings and fit indices. AIC and BIC also suggested that the 3-factor 20-item model had significantly better fit than the original 3-factor 27-item model. In support of this model, internal consistency of the SD3-F subscales were marginally acceptable. More specifically, alpha coefficients for Machiavellianism, narcissism, and psychopathy were marginally adequate for women (0.61 b αs b 0.69) and men (0.50 b αs b 0.65). As mentioned, low alpha coefficients may be explained by the notion that the SD3 items are not conceptually redundant. In addition, we discarded several items that can also decrease the internal consistency of the subscales. Despite providing psychometric support for the reliability, factor structure, and validity of the SD3-F, this research has several limitations. First, the results of the present set of studies are limited to our samples. It is recommended for future research to examine psychometric characteristics of the SD3 in other regions of Iran with higher variability in ethnicity (e.g., Lor, Balooch, Guilaki, Kurd, etc.) and language (e.g., Kurdish, Guilaki, Lori, Dezfuli, etc.). Our sample was recruited from Tehran, the capital of Iran, with low variability in ethnicity and language. Second, we did not assess the temporal stability of the SD3-F in this research. Future investigations could assess the test-retest reliability of the SD3-F. Third, we did not use non-test variables (e.g., interview or informant report) to evaluate the reliability and validity of the SD3-F. It is recommended for future research to assess other forms of reliability and validity using other techniques (e.g., informant report; see Jones & Paulhus, 2014). In sum, our two studies suggested that 20 items of the SD3 had psychometric sufficiency in Iran. Three factors (Machiavellianism, narcissism, and psychopathy) underlie these items. The SD3-F′s subscales had marginally acceptable reliability and good convergent validity. Therefore, the 20-item SD3-F may be used in Iranian context with caution. Yet, as noted by Jones and Paulhus (2014), researchers are encouraged to use standard measures for assessment of the Dark Triad, if they have sufficient time and space in their research.

M. Atari, R. Chegeni / Personality and Individual Differences 102 (2016) 111–117

References Aghababaei, N., Mohammadtabar, S., & Saffarinia, M. (2014). Dirty Dozen vs. the H factor: Comparison of the Dark Triad and Honesty–Humility in prosociality, religiosity, and happiness. Personality and Individual Differences, 67, 6–10. Atari, M. (2016). Reliability and validity of the Persian version of the Grandiose Narcissism Scale. Unpublished Research Report: University of Tehran. Atari, M., Akbari-Zardkhaneh, S., Mohammadi, L., & Soufiabadi, M. (2015). The factor structure and psychometric properties of the Persian version of body appreciation scale. American Journal of Applied Psychology, 3, 62–66. Baughman, H. M., Dearing, S., Giammarco, E., & Vernon, P. A. (2012). Relationships between bullying behaviours and the Dark Triad: A study with adults. Personality and Individual Differences, 52, 571–575. Cain, N. M., Pincus, A. L., & Ansell, E. B. (2008). Narcissism at the crossroads: Phenotypic description of pathological narcissism across clinical theory, social personality psychology, and psychiatric diagnosis. Clinical Psychology Review, 28, 638–656. Chegeni, R., & Atari, M. (2016). Validation of a short four-factor measure of psychopathy among Iranian university students. Rooyesh-e-Ravanshenasi Journal, 4, 105–112. Christie, R., & Geis, F. L. (1970). Studies in Machiavellianism. New York: Academic Press. Cleckley, H. (1941). The mask of sanity. Oxford, England: Mosby. Czarna, A. Z., Jonason, P. K., Dufner, M., & Kossowska, M. (2016). The Dirty Dozen Scale: Validation of a Polish version and extension of the nomological net. Frontiers in Psychology, 7, 445. Fehr, B. A., Samsom, D., & Paulhus, D. L. (1992). The construct of Machiavellianism: Twenty years later. In C. D. Spielberger, & J. N. Butcher (Eds.), Advances in personality assessment. 9. (pp. 77–116). Hillsdale: Erlbaum. Foster, J. D., Campbell, W. K., & Twenge, J. M. (2003). Individual differences in narcissism: Inflated self-views across the lifespan and around the world. Journal of Research in Personality, 37, 469–486. Foster, J. D., McCain, J. L., Hibberts, M. F., Brunell, A. B., & Johnson, R. B. (2015). The Grandiose Narcissism Scale: A global and facet-level measure of grandiose narcissism. Personality and Individual Differences, 73, 12–16. Furnham, A., Richards, S. C., & Paulhus, D. L. (2013). The Dark Triad of personality: A 10 year review. Social and Personality Psychology Compass, 7, 199–216. Hare, R. D., & Neumann, C. S. (2008). Psychopathy as a clinical and empirical construct. Annual Review of Psychology, 4, 217–246. Hayton, J. C., Allen, D. G., & Scarpello, V. (2004). Factor retention decisions in exploratory factor analysis: A tutorial on parallel analysis. Organizational Research Methods, 7, 191–205. Hofstede, G., & McCrae, R. R. (2004). Personality and culture revisited: Linking traits and dimensions of culture. Cross-Cultural Research, 38, 52–88. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. Hunter, J. E., Gerbing, D. W., & Boster, F. J. (1982). Machiavellian beliefs and personality: Construct invalidity of the Machiavellianism dimension. Journal of Personality and Social Psychology, 43, 1293–1305. Jonason, P. K., & Webster, G. D. (2010). The Dirty Dozen: A concise measure of the Dark Triad. Psychological Assessment, 22, 420–432. Jonason, P. K., Li, N. P., Webster, G. D., & Schmitt, D. P. (2009). The dark triad: Facilitating a short-term mating strategy in men. European Journal of Personality, 23, 5–18. Jonason, P. K., Li, N. P., & Czarna, A. Z. (2013). Quick and dirty: Some psychosocial costs associated with the Dark Triad in three countries. Evolutionary Psychology, 11, 172–185. Jones, D. N., & Figueredo, A. J. (2013). The core of darkness: Uncovering the heart of the Dark Triad. European Journal of Personality, 27, 521–531.

117

Jones, D. N., & Paulhus, D. L. (2009). Machiavellianism. In M. R. Leary, & R. H. Hoyle (Eds.), Handbook of individual differences in social behavior (pp. 93–108). New York: Guilford. Jones, D. N., & Paulhus, D. L. (2011a). Differentiating the Dark Triad within the interpersonal circumplex. In L. M. Horowitz, & S. Strack (Eds.), Handbook of interpersonal psychology: Theory, research, assessment, and therapeutic interventions (pp. 249–267). New York: Wiley. Jones, D. N., & Paulhus, D. L. (2011b). The role of impulsivity in the Dark Triad of personality. Personality and Individual Differences, 51, 679–682. Jones, D. N., & Paulhus, D. L. (2014). Introducing the Short Dark Triad (SD3): A brief measure of dark personality traits. Assessment, 21, 28–41. Khodabakhsh, R., Chegeni, R., & Atari, M. (2016). Translation and validation of the Farsi version of the MACH-IV in an Iranian sample. Research Report: Alzahra University [In Farsi]. Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.). New York, NY: Guilford Press. Maples, J. L., Lamkin, J., & Miller, J. D. (2014). A test of two brief measures of the dark triad: The dirty dozen and short dark triad. Psychological Assessment, 26, 326–331. Miller, J. D., Few, L. R., Seibert, A., Watts, A., Zeichner, A., & Lynam, D. R. (2012). An examination of the Dirty Dozen measure of psychopathy: A cautionary tale about the costs of brief measures. Psychological Assessment, 24, 1048–1053. Miller, J. D., Gentile, B., Wilson, L., & Campbell, W. K. (2013). Grandiose and vulnerable narcissism and the DSM–5 pathological personality trait model. Journal of Personality Assessment, 95, 284–290. Mvududu, N. H., & Sink, C. A. (2013). Factor analysis in counseling research and practice. Counseling Outcome Research and Evaluation, 4, 75–98. Neal, T. M. S., & Sellbom, M. (2012). Examining the factor structure of the Hare Self Report Psychopathy scale. Journal of Personality Assessment, 94, 244–253. O′Boyle, E. H., Jr., Forsyth, D. R., Banks, G. C., & McDaniel, M. A. (2012). A meta-analysis of the Dark Triad and work behavior: A social exchange perspective. Journal of Applied Psychology, 97, 557–579. Paulhus, D. L., & Jones, D. N. (2014). Measuring dark personalities. In G. J. Boyle, D. H. Saklofske, & G. Matthews (Eds.), Measures of personality and social psychological constructs. San Diego, CA: Academic Press. Paulhus, D. L., & Williams, K. M. (2002). The Dark Triad of personality: Narcissism, Machiavellianism, and psychopathy. Journal of Research in Personality, 36, 556–563. Paulhus, D. L., Neumann, C. S., & Hare, R. D. (2014). Manual for the Self-Report Psychopathy (SRP) scale. Toronto: Multi-Health Systems. Raskin, R. N., & Terry, H. (1988). A principal-components analysis of the Narcissistic Personality Inventory and further evidence of its construct validity. Journal of Personality and Social Psychology, 54, 890–902. Rauthmann, J. F. (2012). The Dark Triad and interpersonal perception: Similarities and differences in the social consequences of narcissism, Machiavellianism, and psychopathy. Social Psychological and Personality Science, 3, 487–496. Rauthmann, J. F. (2013). Investigating the MACH–IV with item response theory and proposing the Trimmed MACH*. Journal of Personality Assessment, 95, 388–397. Webster, G. D., & Jonason, P. K. (2013). Putting the “IRT” in “Dirty”: Item Response Theory analyses of the Dark Triad Dirty Dozen—An efficient measure of narcissism, psychopathy, and Machiavellianism. Personality and Individual Differences, 54, 302–306. Williams, M. L., Hazleton, V., & Renshaw, S. (1975). The measurement of Machiavellianism: A factor analytic and correlational study of Mach IV and Mach V. Communication Monographs, 42, 151–159. 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, 205–219.