Repelled by virtue? The dark triad and eudaimonic narratives

ABSTRACT

We propose that the dark triad of personality predicts how recipients respond to eudaimonic narratives (stories dealing with purpose in life, the human condition, and human virtue). Matched eudaimonic or noneudaimonic videos were presented via random assignment. The more individuals lack empathy and organize their world around self-promotion—reflected in the so-called dark triad of narcissism, Machiavellianism, and psychopathy—the more they perceived the eudaimonic stories (vs. control) to be inauthentic and corny (perceived corniness). This effect translated to a more negative overall evaluation of the eudaimonic videos (moderated mediation). Self-reported feelings of being touched, moved, and inspired (meaningful affect) were largely unaffected by the dark triad, suggesting that these personality factors do not disable emotional responses to eudaimonic narratives.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. All three major stand-alone measures, the NPI (grandiose narcissism), the Mach IV (Machiavellianism), and the SRP (psychopathy), have shown good reliability and validity in many studies over the past decades (cf. Furnham, Richards, & Paulhus, 2013). Among the three measures, the NPI has been used most extensively to date, and a short form (NPI-16) as well as translations and adapted versions are available in several languages. Whereas all three constructs are often conceived as one-dimensional, different researchers have argued for a varying number of facets (see Furnham et al., 2013, for a summary of previous findings). The psychometric properties of both measures aimed at assessing all three constructs have been discussed intensely. The DDS emphasizes brevity and includes four items per construct. Analyses of its psychometric properties have been mixed. The SD3 incorporates nine items per construct with generally positive results on its reliability and validity. When comparing both measures, Jones and Paulhus (2014) found greater predictive power for the SD3.

2. The invitation to MTurk participants led them to a page which randomly assigned them to our experiment or another, unrelated experiment.

3. The control question asked participants to check strongly agree. Participants who did not respond as requested were excluded. The control question was meant to identify participants that provided answers without reading the item texts.

4. The stimuli are available on request. Please feel free to send an e-mail to Mary Beth Oliver or Markus Appel.

5. Confirmatory factor analyses were conducted with Amos for all measures with continuous response scales. Machiavellianism exhibited an acceptable model fit in a model in which the relationship between the errors of two items was freely estimated. Both items dealt with keeping secrets, χ² = 51.6 (df = 26), p = .002, TLI = .951, CFI = .965, RMSEA = .066. The model fit for psychopathy was acceptable when the relationship between the errors of two items that both described control issues was freely estimated, χ² = 46.6 (df = 26), p = .008, TLI = .943, CFI = .959, RMSEA = .059. Perceived corniness exhibited an acceptable model fit when two correlations between errors were freely estimated. The wordings of the related items were similar in a sense that agreement indicated higher corniness, χ² = 7.6 (df = 3), p = .054, TLI = .969, CFI = .991, RMSEA = .083. The model fit for meaningful affect was acceptable once two correlations between errors were freely estimated, χ² = 58.4 (df = 25), p < .001, TLI = .980, CFI = .986, RMSEA = .077.

6. The survey further included measures on positive and negative affect, which were not analyzed in detail. Moreover, it included an item on the potential sharing of the videos. The overall evaluation and sharing items were strongly related, r(227) = .69, p < .001. Results for the sharing item and the overall evaluation item were similar.

7. The effect size estimates for the significant interaction effects are based on hierarchical regression analyses with the story factor (dummy-coded) and the personality factor (continuous, z-standardized) entered in the first step, and the interaction entered in the second step.