PAID 6040 No.

of Pages 7, Model 5G
16 January 2014

Personality and Individual Differences xxx (2014) xxx–xxx

1

Contents lists available at ScienceDirect

Personality and Individual Differences

journal homepage: www.elsevier.com/locate/paid

4
5

3 The Dark Triad, happiness and subjective well-being

6 Q1 Vincent Egan a,⇑, Stephanie Chan b, Gillian W. Shorter c,d
7 Q2 a
Centre for Family and Forensic Psychology, University of Nottingham, Yang Fujia Building, Jubilee Campus, Wollaton Road, Nottingham NG8 1BB, UK
8 Q3 b
Home Team Behavioural Sciences Centre, Home Team Academy, Singapore 698928, Singapore
9 c
Bamford Centre for Mental Health and Wellbeing, University of Ulster, Londonderry, Northern Ireland
10 Q4 d
All Ireland Hub for Trials Methodology Research, University of Ulster, Londonderry, Northern Ireland

11
12
a r t i c l e i n f o a b s t r a c t
1
2 4
7
15 Article history: Narcissism can be expressed in grandiose or vulnerable forms. We examined whether positive 28
16 Available online xxxx psychological states (defined by the Oxford Happiness Inventory (OHI) and the Diener Satisfaction With 29
Life (SWL) scales) assisted differentiation relative to general personality traits and the ‘‘the Dark Triad’’ 30
17 Keywords: (psychopathy, narcissism, and Machiavellianism, measured by the D12 and Short Dark Triad (SD3) indi- 31
18 Dark Triad ces) for 840 persons primarily from the UK, USA and Canada. The best fitting structural equation model 32
19 Happiness comprised two latent variables, one of positive mood (comprising total scores on the OHI and SWL 33
20 Subjective well-being
scales), and another forming a ‘‘dark dyad’’ of Machiavellianism and psychopathy (predicted by low 34
21 Psychopathy
22 Narcissism
agreeableness and lower positive mood), with narcissism regarded as a separate construct correlated 35
23 Machiavellianism with the dark dyad. Latent positive mood was primarily predicted by higher emotional stability and 36
24 Five factor model extraversion. Narcissism was predicted by lower emotional stability, lower agreeableness, and higher 37
25 Latent profile analysis extraversion. Latent profile analysis identified four groups in the data: ‘‘unhappy but not narcissistic’’, 38
26 ‘‘vulnerable narcissism’’, ‘‘happy non-narcissism’’ and ‘‘grandiose narcissism’’. Our results suggest more 39
problematic narcissism can be identified by reference to measures indexing positive mood states and 40
general personality traits. 41
Ó 2014 Published by Elsevier Ltd. 42

43
44
45 1. The Dark Triad, happiness and subjective well-being whereas Agreeableness (A) and Extraversion (E) predict positive 65
emotions (DeNeve & Cooper, 1998; Steel, Schmidt, & Shultz, 66
46 Narcissism differentiates into grandiose and vulnerable forms 2008). Antagonistic interpersonal behaviour is also predicted by 67
47 (Pincus & Lukowitski, 2010). Grandiose narcissists classically personality; persons high in narcissism and Machiavellianism are 68
48 present as confident, self-centred, and other-oblivious, but can be more likely to disrupt the well-being of colleagues (Campbell, Hoff- 69
49 wilful and exhibitionistic. Vulnerable narcissists are similar, but man, Campbell, & Marchisio, 2011; Kessler et al., 2010); and narcis- 70
50 also hyper-sensitive and hostile. Vulnerable narcissists are inter- sistic aggression is underpinned by low A and E (Egan & Lewis, 71
51 personally problematic, whereas grandiose narcissists can be 2011). Vidal, Skeem, and Camp (2010) found persons in the com- 72
52 highly effective leaders (Furnham, 2007). General non-antagonistic munity with higher secondary psychopathy (i.e., neurotic emotions 73
53 personality traits (for example, Agreeableness or low Neuroticism) concurrent with callous and grandiose traits) better at understand- 74
54 and positive mood are moderating mechanisms that help differen- ing subtle differences in the meanings of complex negative 75
55 tiate narcissism (Gruber, Mauss, & Tamir, 2011). The current study emotions (e.g., grief, depression, remorse, misery) than those with 76
56 further tests this notion, examining the degree to which personal- high primary psychopathy (characterised by callousness and fear- 77
57 ity traits and positive emotions such as happiness and subjective less dominance). Similarly, Ali, Amorim, and Chamorro-Premuzic 78
58 well-being differentiate narcissism, relative to Machiavellianism (2009) found persons with high primary psychopathy reported 79
59 and psychopathy (collectively known as the Dark Triad), and how more positive emotions after viewing sad stimuli, whereas those 80
60 these general personality traits underlie the Dark Triad as mea- with high secondary psychopathy experienced negative emotions 81
61 sured by a new omnibus test of the construct. even after viewing neutral stimuli. Del Gaizo and Falkenbach 82
62 Meta-analyses using the Five Factor Model of personality (FFM; (2008) found persons with high primary psychopathy accurately 83
63 McCrae & Costa, 1997) found Neuroticism (N) strongly predicts recognised fearful facial expressions in others, experienced more 84
64 lower life satisfaction, less happiness, and more negative emotions, positive emotions, and reported fewer negative emotions. Persons 85
with greater secondary psychopathy were poor at recognising 86
emotions in others, and experienced more negative emotions, per- 87
Q5 ⇑ Corresponding author. Tel.: +44 1158466627. haps because they were more absorbed in their own feelings. 88
E-mail address: [email protected] (V. Egan).

0191-8869/$ - see front matter Ó 2014 Published by Elsevier Ltd.

http://dx.doi.org/10.1016/j.paid.2014.01.004

Please cite this article in press as: Egan, V., et al. The Dark Triad, happiness and subjective well-being. Personality and Individual Differences (2014), http://
dx.doi.org/10.1016/j.paid.2014.01.004
 PAID 6040 No. of Pages 7, Model 5G
16 January 2014

2 V. Egan et al. / Personality and Individual Differences xxx (2014) xxx–xxx

89 Emotions are often studied using negative constructs. However, 2.2. Materials 149
90 positive emotional constructs such as subjective well-being (SWB;
91 Diener, Suh, Lucas, & Smith, 1999) and happiness – the state of 1. The International Personality Item Pool (IPIP), The IPIP (Gold- 151
150
92 having enjoyable feelings and making positive judgements (Ryan berg, 1999) is a public domain measure of the Big Five person- 152
93 & Deci, 2001) – are straightforward to measure, and help assess ality dimensions; the version used in this study contained fifty 153
94 persons in a more rounded way. SWB and happiness are associated statements (IPIP-50), requiring participants to indicate on a 154
95 with greater physical and mental health, and with better relation- five-point scale how accurately each statement applied to their 155
96 ships at work and in private lives (Argyle, 1987). The two con- own personality. Responding ranges from 1 (‘‘nothing like me’’) 156
97 structs are not synonymous; while SWB is fundamental to to 5 (‘‘very much like me’’). Ten statements corresponded to 157
98 happiness (Diener et al., 1999), happiness also involves social fac- each personality trait with some statements being reverse- 158
99 tors (Kashdan, 2004); for example, happiness is correlated with keyed to avoid response set bias. Goldberg (1999) reports the 159
100 higher E (Argyle, Martin, & Crossland, 1989). mean reliability for each of the 5 scales to be 0.84. Gow, Whit- 160
101 Seeking to elaborate generic constructs associated with eman, Pattie, and Deary (2005) found the factor structure of the 161
102 unpleasant dispositions, Paulhus and Williams (2002) examined IPIP dimensions were very much equivalent to those assessed 162
103 the relationships between the Dark Triad and the FFM. They found by the NEO-family of instruments, although the N dimension 163
104 that all components of the Dark Triad were negatively correlated is inverted and called Emotional Stability (ES), and the O dimen- 164
105 with A; that narcissism correlated positively with E and Openness sion is re-named Intellect (I). Egan and Taylor (2010) found the 165
106 (O); that Machiavellianism and psychopathy correlated negatively internal consistency reliabilities (Cronbach’s alpha) for a UK 166
107 with Conscientiousness (C); and that psychopathy was negatively sample satisfactory; E = 0.84; A = 0.76; C = 0.77, ES = 0.87, and 167
108 correlated with N but positively correlated with E and O. Jakobwitz I = 0.73. 168
109 and Egan (2006) examined the Dark Triad, differentiating primary 2. The Short Dark Triad questionnaire (SD3). Jones and Paulhus (in 169
110 and secondary psychopathy. Their results confirmed the perennial press) describe a brief 27-item measure of Machiavellianism, 170
111 correlation between the Dark Triad and A (Furnham, Richards, & psychopathy and narcissism, all of which have 9-item scales. 171
112 Paulhus, 2013). Principal components analysis of Jakobwitz et al’s Responding is made to a proposition on a 5-point scale ranging 172
113 data revealed that secondary psychopathy was distinct, in that nar- from 1 (‘‘disagree strongly’’) to 5 (‘‘agree strongly’’). The factor 173
114 cissism, Machiavellianism, and primary psychopathy were nega- structure was clear, and the scales reliable; Machiavellianism 174
115 tively correlated with A, whereas secondary psychopathy was a = 0.75; psychopathy a = 0.72; and narcissism a = 0.73 (Jones 175
116 also associated with high N and low C. Similar results followed & Paulhus, in press). 176
117 when Ross, Lutz, and Bailley (2004) mapped primary and second- 3. The Dirty Dozen Scale (D12). The ‘‘dirty dozen’’ test (Jonason & 177
118 ary psychopathy onto the FFM. Narcissism differs from more obvi- Webster, 2010) comprises 12 items to briefly measure the Dark 178
119 ously ‘‘dark’’ personality traits as it has ‘‘brighter’’ elements, and Triad. Each item comprises a proposition which is rated on a 5- 179
120 can be attractive interpersonally (Rauthmann & Kolar, 2012). We point scale from 1 (‘‘strongly disagree’’) to 5 (‘‘strongly agree’’). 180
121 propose attractive aspects of potentially difficult personalities are Internal (test–retest) reliabilities for the two Machiavellian 181
122 identifiable using positive traits. items are a = 0.79 (0.89), six psychopathy items a = 0.77 182
123 This study predicts that general personality traits and positive (0.74), and four narcissism items are a = 0.88 (0.84). 183
124 mood differentiate narcissism more than they do Machiavellianism 4. Oxford Happiness Inventory (OHI). The OHI (Argyle et al., 1989) 184
125 and psychopathy, which are largely driven by low A. We test this is a 29-item scale used to measure happiness in non-clinical 185
126 proposition using a large sample and recruit beyond student co- populations. Responses are scored on a 6-point scale, with 186
127 horts, using multivariate statistics to strongly test the hypothesis. responses ranging from 1 (‘‘strongly disagree’’) to 6 (‘‘strongly 187
agree’’). The scale has an overall internal reliability of 0.91, with 188
subscales being also reliable; mastery (0.80), satisfaction (0.81), 189
social cheerfulness (0.74), vigour (0.67), and social interest 190
(0.65) (Meleddu, Guicciardi, Scalas, & Fadda, 2012). 191
128 2. Method and procedure 5. The Satisfaction with Life Scale (SWLS). The SWLS (Diener, 192
Emmons, Larsen, & Griffin, 1985) is a brief (5-item), highly reli- 193
129 2.1. Participants able (a = 0.87, test–retest reliability = 0.82) and well-validated 194
measure of positive emotions. Persons respond to a proposition 195
130 The study opportunistically recruited 861 persons via Facebook on a 7-point scale ranging from 1 (‘‘strongly disagree’’) to 7 196
131 and a variety of online Internet-based research sites. All included (‘‘strongly agree’’). 197
132 participants were fluent English speakers, and aged 18 years or 198
133 over. To optimise integrity of information, we asked participants In all cases, a higher scores on the measured construct indicated 199
134 to respond to an attentional probe question with an ‘‘agree’’ re- greater endorsement of the given trait. 200
135 sponse part way through the survey, and elsewhere to respond
136 likewise with ‘‘disagree’’. This method identified 21 individuals
137 who had not responded as requested, perhaps because they had
138 not read the question correctly, responded randomly, or because 2.3. Procedure 201
139 they had a response set. When these 21 persons were excluded,
140 the sample comprised 594 females and 246 males, mean partici- Participants were recruited online via a variety of social net- 202
141 pant age = 30.1 years (SD = 12.7). The mean years of education for working web sites, accruing responses via a variety of initiating 203
142 the cohort was 15.0 (SD = 3.8), although 216 (25.7%) had 12 or few- Facebook pages, University bulletin boards, and the Hanover on- 204
143 er years of education, and so were unlikely to have a degree. Per- line Psychological research page. Participants were informed of 205
144 sons were recruited from the UK (375), the USA (306), Europe the nature of the study and those who wished to take part con- 206
145 (70), Asia (42), Canada (30), Australasia (9), and Africa (8). Of the sented online. Subjects then completed a brief series of questions 207
146 cohort, 441 (52.5%) were currently in a relationship, while 399 describing their age, gender, education, nationality, marital status, 208
147 (28.5%) were single. As only 27 participants had prior criminal con- and whether they had prior convictions. The full survey comprised 209
148 victions, testing for forensic effects was unrealistic. 124 questions. 210

Please cite this article in press as: Egan, V., et al. The Dark Triad, happiness and subjective well-being. Personality and Individual Differences (2014), http://
dx.doi.org/10.1016/j.paid.2014.01.004
 PAID 6040 No. of Pages 7, Model 5G
16 January 2014

V. Egan et al. / Personality and Individual Differences xxx (2014) xxx–xxx 3

211 3. Analysis Table 2

Correlations (Pearson’s r) between SD3 and D12 measures of the Dark Triad (n = 840).

212 All items were automatically written to file, allowing item-anal- SD3
213 yses for all scales. Scales were tested for internal reliability using D12 Machiavellianism Narcissism Psychopathy
214 Cronbach’s alpha, and the two Dark Triad scales correlated as a Machiavellianism 0.38 0.27 0.55
215 means of providing convergent reliability. Measures were intercor- Narcissism 0.21 0.50 0.24
216 related to examine the degree of shared variance between the dif- Psychopathy 0.51 0.29 0.59
217 ferent personality measures. Finally, structural equation models
All coefficients one-tailed and significant at P < .001; D12 is the short form 12-item
218 were calculated using AMOS (Arbuckle & Wothke, 2003) using to- measure of the Dark Triad; SD3 the 27-item measure of the same construct.
219 tal scores for the measured variables for all participants. These Underlined coefficients show the agreement for key constructs between test
220 examined how the latent variables of positive mood and the Dark instruments.
221 Triad related to each other and to dark and normal personality
222 traits for the SD3, as compared to the D12. ever, there were slight improvements in fit for a SD3-defined dark 249
dyad (v2 (df) = 92.977 (22); GFI = 0.976; AGFI = 0.943; 250
RMSEA = 0.064; CMIN/DF = 4.427), compared to dark dyad defined 251
by the D12 (v2 (df) = 131.952 (22); GFI = 0.956; AGFI = 0.922; 252
223 4. Results
RMSEA = 0.077; CMIN/DF = 5.998). A combined model using both 253
SD3 and D12 Dark Triad indices combined to increase the indica- 254
224 Summary statistics for the scales used are presented in Table 1.
tors for the relevant latent variables did not sharpen the fit of 255
225 All measures were reliable at a = 0.72 or above except for OHI Social
the model (v2 (df) = 259.97; GFI = 0.948; AGFI = 0.905; 256
226 Cheerfulness (a = 0.63); results were comparable to published val-
RMSEA = 0.079; CMIN/DF = 6.190). The results of the structural 257
227 ues. Exploratory correlations did not find age or education associ-
equation model using the SD3 are presented in Fig. 1. 258
228 ated strongly with any personality construct; although by the
We examined how personality traits related to latent positive 259
229 nature of the large sample size, these small associations (typically
mood and the dark dyad, and if positive mood related to these con- 260
230 0.1) were sporadically significant. Measured construct correlations
structs. Latent positive mood was defined by the total score on the 261
231 between comparable indices on the SD3 and D12 were modest to
OHI and SWLS. Positive mood was associated primarily with ES and 262
232 acceptable (Machiavellianism r = 0.38, P < .001; narcissism
E, with smaller (but significant) influences from the other IPIP 263
233 r = 0.50, P < .001; psychopathy r = 0.59, P < .001). However, there
dimensions. Though not associated with narcissism, latent positive 264
234 was also considerable scale cross-correlation. These associations
mood was negatively associated with the latent ‘dark dyad’ ( 0.37 265
235 suggest that the some Dark Triad constructs and measures overlap
and 0.26 for the SD3 and D12, respectively). These results suggest 266
236 considerably and are non-specific (Table 2).
that positive mood is the product of E and ES, and is primarily asso- 267
237 Structural equation models were fit to total scores, the models
ciated with the more positive expression of narcissism, whereas 268
238 involving latent variables of positive mood (the OHI and SWLS total
the vulnerable elements of narcissism correlate with lower A, low- 269
239 scores) and their relationship with general personality traits, in
er ES, and the dark dyad (itself negatively related to positive 270
240 turn examining how these contributed to the Dark Triad. All mea-
mood). The SD3-defined dark dyad is solely predicted by A at 271
241 surement models comprised continuous scale summary scores for
0.51 ( 0.66 for the D12). Our results suggest the SD3 narcissism 272
242 participants, rather than item-level indicators. The initial AMOS
measure can be interpreted as indicating grandiose or vulnerable 273
243 models tested if Machiavellianism, narcissism, and psychopathy
narcissism if examined alongside general personality and positive 274
244 loaded on a single Dark Triad latent variable. A far better fit was
mood traits. 275
245 obtained by having a ‘dark dyad’ latent variable (psychopathy
Finally, a latent profile analysis was conducted. A series of mod- 276
246 and Machiavellianism), with a separate, directly measured narcis-
els with two to six profiles were run. A four profile model was se- 277
247 sism variable which was allowed to correlate with the dark dyad.
lected the best fit of the data (Table 3). This was decided on the 278
248 Fit statistics for both SD3 and D12 models were acceptable, how-
basis of the Lo–Mendel–Rubin Likelihood Ratio test (2001). Though 279
Table 1 the information criteria did not conclusively suggest an optimal 280
Means, standard deviations and reliability of measurements. model, inspection of the profiles showed a four profile solution of- 281
Measure Mean SD Alpha reliability fered a theoretically sound profile relative to the more empirical 282

IPIP
three profile solution. The first profile represents a group with 283
Emotional stability 28.0 7.9 0.88 the least ‘‘happy’’ characteristics. This group were lowest on E, 284
Extraversion 30.0 7.7 0.88 ES, I, the OHI total, and the SWLS, and comprised 21% of the sam- 285
Intellect 37.4 5.9 0.80 ple. Vulnerable narcissists were similarly low on A but higher in 286
Agreeableness 38.5 6.0 0.85
self-reported I, highest in Machiavellianism, narcissism and psy- 287
Conscientiousness 33.6 6.1 0.78
chopathy, and next most dissatisfied and unhappy relative to the 288
SD3
unhappy non-narcissists: this group represented 23% of the cohort. 289
Machiavellianism 2.83 0.58 0.74
Narcissism 2.70 0.63 0.74 The largest group were the grandiose narcissists, who comprised 290
Psychopathy 2.07 0.62 0.76 31% of the sample. This group were the highest for E, A, C, and 291

D12 ES, and were equivalent to the vulnerable narcissists in higher 292
Machiavellianism 3.2 1.6 0.72 self-rated intellect. They were highest on the OHI and SWL scales, 293
Narcissism 11.2 3.1 0.73 and lower in psychopathy. The happy, non-narcissistic group (25% 294
Psychopathy 14.5 4.5 0.74 of the group) were also higher in A, and were less narcissistic, psy- 295
Positive mood chopathic, or Machiavellian (Fig. 2). 296
Oxford happiness inventory (OHI) total 91.4 17.2 0.92
OHI mastery 31.1 6.9 0.82
OHI satisfaction 20.2 4.9 0.86 5. Discussion 297
OHI vigour 14.9 3.9 0.72
OHI social interest 6.93 1.7 0.63 We tested the degree to which positive emotions are associated 298
OHI social cheerfulness 18.3 3.5 0.77
with the Dark Triad, using general underlying personality traits to 299
Diener satisfaction with life scale 14.6 4.5 0.86
structure this information. While the two Dark Triad scales were 300

Please cite this article in press as: Egan, V., et al. The Dark Triad, happiness and subjective well-being. Personality and Individual Differences (2014), http://
dx.doi.org/10.1016/j.paid.2014.01.004
 PAID 6040 No. of Pages 7, Model 5G
16 January 2014

4 V. Egan et al. / Personality and Individual Differences xxx (2014) xxx–xxx

Fig. 1. SEM of subjective positive mood, the Dark Triad (measured by SD3) and personality (IPIP). All pathways are standardised regressions statistically significant at P < .001.
Narcissism contributes substantially to a latent ‘‘dark dyad’’, but functions independently of the latent variable. The dark dyad, however, is predicted primarily by lower
Q8 positive mood and lower Agreeableness.

Table 3
Fit statistics of latent class analysis for ten personality indicators.

Number of classes Ll (df) AIC BIC SSABIC LRT p

2 21,759.38(31) 43,580.76 43,727.49 43,629.05 1208.22 0.0001
3 21,543.67 (42) 43,171.33 43,370.13 43,236.76 425.68 00001
4 21,424.31 (53) 42,954.63 43,205.50 43,037.19 235.53 0.0398
5 21,337.16 (64) 42,802.31 43,105.25 42,902.01 171.99 0.0842
6 21,262.77 (75) 42,675.54 43,030.54 42,792.37 146.79 0.6784
7 21,202.44 (86) 42,576.89 42,983.96 42,710.85 119.04 0.6343

Note: LL(df), log likelihood value and associated degrees of freedom; LRT, Lo–Mendell–Rubin adjusted likelihood ratio test value; AIC, Akaike information criterion; BIC,
Bayesian information criterion; SSABIC, sample size adjusted Bayesian information criterion.

Fig. 2. Graph presenting standardised (z-score) means for the four groups identified using latent class analysis.

301 correlated, they were also significantly correlated with their into a better-fitting dark dyad of Psychopathy and Machiavellian- 304
302 respective measures of Psychopathy and Machiavellianism. Struc- ism, and a separate narcissism measure. Latent positive mood 305
303 tural equation models indicated better fit if one split the Dark Triad and dark dyad constructs were produced, with personality traits 306

Please cite this article in press as: Egan, V., et al. The Dark Triad, happiness and subjective well-being. Personality and Individual Differences (2014), http://
dx.doi.org/10.1016/j.paid.2014.01.004
 PAID 6040 No. of Pages 7, Model 5G
16 January 2014

V. Egan et al. / Personality and Individual Differences xxx (2014) xxx–xxx 5

307 being fitted to these constructs. Both models were essentially the rather different to most contemporary environments (Gangestad, 373
308 same, though the SD3’s fitness indices were slightly better than Thornhill, & Garver-Apgar, 2005). In this context, grandiose narcis- 374
309 those for the D12; combining both scales for indicators did not sism may have developed as a way to present social dominance, 375
310 strengthen the fit of the model. Positive mood was particularly pre- good-natured and confident pride being a way to compete with 376
311 dicted by ES and E, whereas the dark dyad was predicted by low A, disagreeable persons primarily oriented to selfish and self-centred 377
312 greater narcissism, and lower positive mood. Narcissism was par- behavioural strategies (Dillon, Adair, Wang, & Johnson, 2013). Posi- 378
313 ticularly predicted by E, low A, and greater self-rated happiness. tive states clarify and display this distinction and help observers 379
314 These findings reiterate narcissism’s complexity and expression evaluate threat from narcissistic peers. 380
315 in relatively positive and negative forms (Ackerman et al., 2011). Although our findings derive from good measures, a large and 381
316 Latent profile analyses revealed four patterns of trait expression, reasonably representative sample, and rigorous analysis, the study 382
317 supporting relationships shown in the SEM. These profiles reflected is not without faults. Central to these was that our measurement of 383
318 groups of persons who were unhappy, persons in the normal range narcissism was extremely brief, and did not actively test for differ- 384
319 (i.e., within 0.5 of a standard deviation of the mean for all mea- entiation of a priori scales of grandiose and vulnerable narcissism 385
320 sures), and grandiose and vulnerable narcissists. Grandiose narcis- as indexed by the Narcissistic Personality Inventory (Ackerman 386
321 sists were higher in E, ES, satisfaction with life, and general et al., 2011) or the Pathological Narcissism Inventory (PNI; Pincus 387
322 happiness; vulnerable narcissists were lower in A and ES, and high- et al., 2009); some have suggested that the identification of proso- 388
323 er in psychopathy and Machiavellianism. cial aspects of narcissism is a consequence of using scales with 389
324 Our results suggest differentiating grandiose and vulnerable items that more measure self-esteem than genuine narcissistic 390
325 narcissists may be relatively straightforward, answering the call tendencies (Brown, Budzek, & Tamborski, 2009; Veselka, Schermer, 391
326 Pincus and Lukowitski (2010) made regarding this need. In our & Vernon, 2012). While we accept this possibility, we sought to test 392
327 data, vulnerable narcissists were not gregarious or socially inter- the notion that general narcissism can be grandiose or vulnerable, 393
328 ested, and were relatively anhedonic, even though they were not depending on the configuration of traits it functions within, and 394
329 as markedly introverted, emotionally unstable or unhappy as our our results upheld such a differentiation. It would nevertheless 395
330 most troubled group. In future practice, informant ratings, socio- be helpful to replicate our findings with more extensive measures 396
331 metric evaluations of functioning within groups, and even, as here, of narcissism and on occupational and clinical samples. 397
332 self-report scales are all potentially applicable to making this
333 behavioural differentiation. Our findings reaffirm in that happiness
334 (which we extend by also including life satisfaction) is under- 6. Uncited references 398
335 pinned by ES and E (Robbins, Francis, & Edwards, 2010).
336 We used two novel measures of the Dark Triad to examine Ames, Rose, and Anderson (2006), Board and Fritzon (2005), 399
337 which was more useful. Both scales were reliable, and internal Christie and Geis (1970), Emmons (1984), Hills and Argyle 400
338 and cross-test correlations were comparable. Our SEM models sug- (2002), Lo, Mendel, and Rubin (2001) and McCrae and Costa 401
339 gested the fit of personality and positive mood to the D12’s Dark (2004). Q6 402
340 Triad measures was slightly poorer than that for the SD3. This
341 may be because although the D12 is psychometrically rigorous in
342 some respects (Webster & Jonason, 2013), it focuses on the more References 403
343 callous-unemotional elements of psychopathy (‘primary psychop-
344 athy’) than the interpersonally antagonistic and disinhibited as- Ackerman, R. A., Witt, E. A., Donnellan, M. B., Trzesniewski, K. H., Robins, R. W., & 404
Kashy, D. A. (2011). What does the Narcissistic Personality Inventory Really 405
345 pects of the construct (Miller et al., 2012). Jonason and Luéano Measure? Assessment, 18(1), 67–87. http://dx.doi.org/10.1177/ 406
346 (2013) accept this critique, noting they seek to estimate complex 1073191110382845. 407
347 traits using only a few items per dimension, leading to a loss of res- Ali, F., Amorim, I. S., & Chamorro-Premuzic, T. (2009). Empathy deficits and trait 408
emotional intelligence in Psychopathy and Machiavellianism. Personality and 409
348 olution. The D12 nevertheless remains useful in research where Individual Differences, 47(7), 758–762. http://dx.doi.org/10.1016/ 410
349 participant attention may be relatively transient, as in the case j.paid.2009.06.016. 411
350 for studies involving offenders or the mentally disordered (e.g., Ames, D. R., Rose, P., & Anderson, C. P. (2006). The NPI-16 as a short measure of 412
Narcissism. Journal of Research in Personality, 40(4), 440–450. http://dx.doi.org/ 413
351 Egan & Beadman, 2011; Egan, Kavanagh, & Blair, 2005). 414
10.1016/j.jrp.2005.03.002.
352 Jonason and Luéano (op cit) also found the best-fitting model Arbuckle, J. L., & Wothke, W. (2003). Amos 5.0 users guide. Chicago: SmallWaters. 415
353 for their data was bifactorial: one dimension being psychopathy– Argyle, M. (1987). The psychology of happiness. London, UK: Methuen. 416
Argyle, M., Martin, M., & Crossland, J. (1989). Happiness as a function of personality 417
354 Machiavellianism, the other narcissism. Our SEM analyses repli- 418
and social encounters. In J. P. Forgas & J. M. Innes (Eds.), Recent advances in social
355 cated this structure for the D12, and also for the SD3, underlying psychology: An international perspective (pp. 189–203). North-Holland: Elsevier. 419
356 the separation of narcissism from the manipulative–callous dark Board, B. J., & Fritzon, K. (2005). Disordered personalities at work. Psychology, Crime 420
& Law, 11(1), 17–32. http://dx.doi.org/10.1080/10683160310001634304. 421
357 core of what we called ‘‘the dark dyad’’ (Jones & Figueredo, 2014;
Brown, R. P., Budzek, K., & Tamborski, M. (2009). On the meaning and measure of 422
358 Rauthmann & Kolar, 2012). Our data suggests that when scales of narcissism. Personality and Social Psychology Bulletin, 35(7), 951–964. http:// 423
359 personality and positive mood are included, narcissism itself can dx.doi.org/10.1177/0146167209335461. 424
360 be constructively differentiated, such that grandiose narcissism Campbell, W. K., Hoffman, B. J., Campbell, S. M., & Marchisio, G. (2011). Narcissism 425
in organizational contexts. Human Resource Management Review, 21(4), 426
361 (or at least authentic pride in oneself) can be distinguished from 268–284. http://dx.doi.org/10.1016/j.hrmr.2010.10.007. 427
362 often hubristic and destructive vulnerable narcissism. Christie, R., & Geis, F. (1970). Studies in Machiavellianism. New York: Academic Press. 428
363 Second-order personality constructs such as narcissism and Del Gaizo, A., & Falkenbach, D. (2008). Primary and secondary psychopathic-traits 429
and their relationship to perception and experience of emotion. Personality and 430
364 psychopathy–Machiavellianism emerge out of lower-level person- Individual Differences, 45(3), 206–212. http://dx.doi.org/10.1016/ 431
365 ality traits with some emergent features of their own. These traits j.paid.2008.03.019. 432
366 are adaptive when implemented in particular ecological niches; DeNeve, K. M., & Cooper, H. (1998). The happy personality: A meta-analysis of 137 433
personality traits and subjective well-being. Psychological Bulletin, 124(2), 434
367 expressions of social dominance appear to enhance the effective- 435
197–229. http://dx.doi.org/10.1037/0033-2909.124.2.197.
368 ness of mating effort (Egan & Angus, 2004; Jonason, Li, & Buss, Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life 436
369 2010). Dark Triad traits maintain in the population because of scale. Journal of Personality Assessment, 49(1), 71–75. http://dx.doi.org/10.1207/ 437
s15327752jpa4901_13. 438
370 the efforts persons high in the Dark Triad put into sexual activity,
Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three 439
371 and because the traits they present may have been ancestral mark- decades of progress. Psychological Bulletin, 125(2), 276–302. http://dx.doi.org/ 440
372 ers of biological fitness in a dangerous and unpredictable world 10.1037/0033-2909.125.2.276. 441

Please cite this article in press as: Egan, V., et al. The Dark Triad, happiness and subjective well-being. Personality and Individual Differences (2014), http://
dx.doi.org/10.1016/j.paid.2014.01.004
 PAID 6040 No. of Pages 7, Model 5G
16 January 2014

6 V. Egan et al. / Personality and Individual Differences xxx (2014) xxx–xxx

442 Dillon, H. M., Adair, L. E., Wang, Z., & Johnson, Z. (2013). Slow and steady wins the Jones, D. N., & Figueredo, A. J. (2014). The core of darkness: Uncovering the heart of 506
443 race: Life history, mate value, and mate settling. Personality and Individual the Dark Triad. European Journal of Personality, 27(6), 521–531. http:// 507
444 Differences, 55(5), 612–618. http://dx.doi.org/10.1016/j.paid.2013.05.015. dx.doi.org/10.1002/per.1893. 508
445 Egan, V., & Angus, S. (2004). Is social dominance a sex-specific strategy for Kashdan, T. B. (2004). The assessment of subjective well-being (issues raised by the 509
446 infidelity? Personality and Individual Differences, 36(3), 575–586. http:// Oxford Happiness Questionnaire). Personality and Individual Differences, 36(5), 510
447 dx.doi.org/10.1016/S0191-8869(03)00116-8. 1225–1232. http://dx.doi.org/10.1016/S0191-8869(03)00213-7. 511
448 Egan, V., & Beadman, M. (2011). Personality and gang embeddedness. Personality Kessler, S. R., Bandelli, A. C., Spector, P. E., Borman, W. C., Nelson, C. E., & Penney, L. 512
449 and Individual Differences, 51(6), 748–753. http://dx.doi.org/10.1016/ M. (2010). Re-examining Machiavelli: A three-dimensional model of 513
450 j.paid.2011.06.021. Machiavellianism in the workplace. Journal of Applied Social Psychology, 40(8), 514
451 Egan, V., Kavanagh, B., & Blair, M. (2005). Sexual offenders, personality and 1868–1896. http://dx.doi.org/10.1111/j.1559-1816.2010.00643.x. 515
452 obsessionality. Sexual Abuse, 17(3), 223–240. http://dx.doi.org/10.1177/ Lo, Y., Mendel, N. R., & Rubin, D. B. (2001). Testing the number of components in a 516
453 107906320501700301. normal mixture. Biometrika, 88(3), 767–778. http://dx.doi.org/10.1093/biomet/ 517
454 Egan, V., & Lewis, M. (2011). Neuroticism and agreeableness differentiate emotional 88.3.767. 518
455 and narcissistic expressions of aggression. Personality and Individual Differences, McCrae, R. R., & Costa, P. T. Jr, (1997). Personality trait structure as a human 519
456 50(6), 845–850. http://dx.doi.org/10.1016/j.paid.2011.01.007. universal. American Psychologist, 52(5), 509–516. http://dx.doi.org/10.1037/ 520
457 Egan, V., & Taylor, D. (2010). Shoplifting, unethical consumer behaviour, and 0003-066X.52.5.509. 521
458 personality. Personality and Individual Differences, 48(8), 878–883. http:// McCrae, R. R., & Costa, P. T. (2004). A contemplated revision of the NEO Five Factor 522
459 dx.doi.org/10.1016/j.paid.2010.02.014. Inventory. Personality and Individual Differences, 36(3), 587–596. http:// 523
460 Emmons, R. A. (1984). Factor analysis and construct validity of the Narcissistic dx.doi.org/10.1016/S0191-8869(03)00118-1. 524
461 personality inventory. Journal of Personality Assessment, 48(3), 291–305. http:// Meleddu, M., Guicciardi, M., Scalas, L. F., & Fadda, D. (2012). Validation of an Italian 525
462 dx.doi.org/10.1207/s15327752jpa4803_11. version of the Oxford Happiness Inventory in adolescence. Journal of Personality 526
463 Furnham, A. (2007). Personality disorders and derailment at work: The paradoxical Assessment, 94(2), 175–185. http://dx.doi.org/10(1080/00223891), 2011, 645931. 527
464 positive influence of pathology in the workplace. In J. Langan-Fox, C. L. Cooper, Miller, J. D., Few, L. R., Seibert, L. A., Watts, A., Zeichner, A., & Lynam, D. R. (2012). An 528
465 & R. J. Klimoski (Eds.), Research companion to the dysfunctional workplace examination of the Dirty Dozen measure of Psychopathy: A cautionary tale 529
466 (pp. 22–39). Cheltenham: Edward Elgar. about the costs of brief measures. Psychological Assessment, 24(4), 1048–1053. 530
467 Furnham, A., Richards, S. C., & Paulhus, D. L. (2013). The Dark Triad of personality: A http://dx.doi.org/10.1037/a0028583. 531
468 10 year review. Social and Personality Psychology Compass, 7(3), 199–216. http:// Paulhus, D. L., & Williams, K. M. (2002). The Dark Triad of personality: Narcissism, 532
469 dx.doi.org/10.1111/spc3.12018. Machiavellianism and Psychopathy. Journal of Research in Personality, 36(6), 533
470 Gangestad, S. W., Thornhill, R., & Garver-Apgar, C. E. (2005). Women’s sexual 556–563. http://dx.doi.org/10.1016/S0092-6566(02)00505-6. 534
471 interests across the ovulatory cycle depend on primary partner developmental Pincus, A. L., Ansell, E. B., Pimentel, C. A., Cain, N. M., Wright, A. G., & Levy, K. N. 535
472 instability. Proceedings of the Royal Society B: Biological Sciences, 272(1576), (2009). Initial construction and validation of the Pathological Narcissism 536
473 2023–2027. http://dx.doi.org/10.1098/rspb.2005.3112. Inventory. Psychological Assessment, 21(3), 365–379. http://dx.doi.org/10.1037/ 537
474 Goldberg, L. R. (1999). A broad-bandwidth, public-domain, personality inventory a0016530. 538
475 measuring the lower-level facets of several five-factor models. In I. Mervielde, I. Pincus, A. L., & Lukowitski, M. R. (2010). Pathological Narcissism and narcissistic 539
476 Deary, F. De Fruyt, & F. Ostendorf (Eds.). Personality psychology in Europe (Vol. 7, personality disorder. Annual Review of Clinical Psychology, 6, 421–446. http:// 540
477 pp. 7–28). Tilburg, The Netherlands: Tilburg University Press. dx.doi.org/10.1146/annurev.clinpsy.121208.131215. 541
478 Gow, A. J., Whiteman, M. C., Pattie, A., & Deary, I. J. (2005). Goldberg’s ‘IPIP’ Big-Five Rauthmann, J. F., & Kolar, G. P. (2012). How ‘‘dark’’ are the Dark Triad traits? 542
479 Factor markers: Internal consistency and concurrent validation in Scotland. Examining the perceived darkness of Narcissism, Machiavellianism, and 543
480 Personality and Individual Differences, 39(2), 317–329. http://dx.doi.org/10.1016/ Psychopathy. Personality and Individual Differences, 53(7), 884–889. http:// 544
481 j.paid.2005.01.011. dx.doi.org/10.1016/j.paid.2012.06.020. 545
482 Gruber, J., Mauss, I. B., & Tamir, M. (2011). A dark side of happiness? How, when, Robbins, M., Francis, L. J., & Edwards, B. (2010). Happiness as stable extraversion: 546
483 and why happiness is not always good. Perspectives on Psychological Science, 6, Internal consistency reliability and construct validity of the Oxford Happiness 547
484 222–233. http://dx.doi.org/10.1177/1745691611406927. Questionnaire among undergraduate students. Current Psychology, 29(2), 89–94. 548
485 Hills, P., & Argyle, M. (2002). The Oxford Happiness Questionnaire: A compact scale http://dx.doi.org/10.1007/s12144-010-9076-8. 549
486 for the measurement of psychological well-being. Personality and Individual Ross, S. R., Lutz, C. J., & Bailley, S. E. (2004). Psychopathy and the Five Factor Model 550
487 Differences, 33(7), 1073–1082. http://dx.doi.org/10.1016/S0191- in a non-institutionalized sample: A domain and facet level analysis. Journal of 551
488 8869(01)00213-6. Psychopathology and Behavioral Assessment, 26(4), 213–223. http://dx.doi.org/ 552
489 Jakobwitz, S., & Egan, V. (2006). The ‘Dark Triad’ and normal personality traits. 10.1023/B:JOBA.0000045337.48535.a5. 553
490 Personality and Individual Differences, 40(2), 331–339. http://dx.doi.org/10.1016/ Ryan, R. M., & Deci, E. L. (2001). On happiness and human potentials: A review of 554
491 j.paid.2005.07.006. research on hedonic and eudaimonic well-being. Annual Review of Psychology, 555
492 Jonason, P. K., Li, N. P., & Buss, D. M. (2010). The costs and benefits of the Dark Triad: 52(1), 141–166. http://dx.doi.org/10.1146/annurev.psych.52.1.141. 556
493 Implications for mate poaching and mate retention tactics. Personality and Steel, P., Schmidt, J., & Shultz, J. (2008). Refining the relationship between 557
494 Individual Differences, 48(4), 373–378. http://dx.doi.org/10.1016/ personality and subjective well-being. Psychological Bulletin, 134(1), 138–161. 558
495 j.paid.2009.11.003. http://dx.doi.org/10.1037/0033-2909.134.1.138. 559
496 Jonason, P. K., & Luéano, V. X. (2013). Walking the thin line between efficiency and Veselka, L., Schermer, J. A., & Vernon, P. A. (2012). The Dark Triad and an expanded 560
497 accuracy: Validity and structural properties of the Dirty Dozen. Personality and framework of personality. Personality and Individual Differences, 53(4), 417–425. 561
498 Individual Differences, 55(1), 76–81. http://dx.doi.org/10.1016/ http://dx.doi.org/10.1016/j.paid.2012.01.002. 562
499 j.paid.2013.02.010. Vidal, S., Skeem, J., & Camp, J. (2010). Emotional intelligence: Painting different 563
500 Jonason, P. K., & Webster, G. D. (2010). The dirty dozen: A concise measure of the paths for low-anxious and high-anxious psychopathic variants. Law and Human 564
501 Dark Triad. Psychological Assessment, 22(2), 420–432. http://dx.doi.org/10.1037/ Behavior, 34(2), 150–163. http://dx.doi.org/10.1007/s10979-009-9175-y. 565
502 a0019265. Webster, G. D., & Jonason, P. K. (2013). Putting the ‘‘IRT’’ in ‘‘Dirty’’: Item response 566
503 Jones, D.N., & Paulhus, D.L. (in press). Introducing the Short Dark Triad (SD3): A brief theory analyses of the Dark Triad Dirty Dozen – An efficient measure of 567
504 measure of dark personality traits. Assessment. doi:10.1177/ Narcissism, Psychopathy, and Machiavellianism. Personality and Individual 568
505 Q7 1073191113514105. Differences, 54(2), 302–306. http://dx.doi.org/10.1016/j.paid.2012.08.027. 569
570

Please cite this article in press as: Egan, V., et al. The Dark Triad, happiness and subjective well-being. Personality and Individual Differences (2014), http://
dx.doi.org/10.1016/j.paid.2014.01.004