A Mokken analysis of the Dark Triad 'Dirty Dozen': Sex and age differences in scale structures, and issues with individual items. (September 2015)

• Record Type: Journal Article
• Authors: Carter, Gregory Louis; Campbell, Anne C.; Muncer, Steven; Carter, Katherine A.
• Is Part Of: Personality and individual differences. Volume 83(2015)
• Page Start: 185
• Page End: 191
• Publication Date: 2015-09
• Subjects: Dark Triad -- Narcissism -- Machiavellianism -- Psychopathy -- Mokken analysis -- Sex differences -- Age differences; Personality -- Periodicals; Individuality -- Periodicals; Individuality -- Periodicals; Personality Development -- Periodicals; Personnalité -- Périodiques; Individualité -- Périodiques; 155.205
• DOI: 10.1016/j.paid.2015.04.012
• Languages: English
• ISSNs: 0191-8869

Abstract

Highlights: The Dirty Dozen is a concise measure of the Dark Triad of personality. Mokken analysis was used to examine sex and age differences in its item structure. Issues with multiple items, aggregate, and trait scales emerged in student samples. All items constituted one hierarchical scale for non-student-aged men and women. Item ordering ('difficulty') was similar in men and women. Abstract: The Dark Triad (DT: Machiavellianism, narcissism, and subclinical psychopathy) have often been measured using a 12-item scale: The 'Dirty Dozen'. Many articles report participants' scale scores as well as their total score because structural models, based on classical test theory analysis, have indicated DT can be represented both as three correlated scales and a single scale. As DT are proposed to underlie a 'male' reproductive strategy of short-term, low-investment mating, sex differences have been of particular theoretical interest. Using two samples – one of student-aged participants; another comprised of a broader national sample – we applied Mokken analysis to investigate whether the same hierarchical structure existed across sex and age. For student women, the exclusion of one psychopathy item produced a single hierarchical DT scale. For student men, items formed a three-item narcissism scale and a six-item Machiavellianism–psychopathy scale. For non-student women and men, all twelve items constituted a unidimensional DT scale. Across all groups, item 'difficulty' was similar: Narcissism items were most easily endorsed and psychopathy items had the lowest rate of endorsement. Results are discussed in relation to the problematic empirical status of the Dirty Dozen psychopathy subscale, and in relation to sex and age differences.