Can Your Darkness be Measured? Validating the Dark Core of Personality in Swedish
(2022) PSYP01 20221Department of Psychology
- Abstract
- Objective: The Dark Factor of Personality (D) measures the underlying core of all aversive traits. The present study evaluated the psychometric properties of the Swedish version of three item sets D70, D35, and 16 in regard to their factor structure, item quality and functioning, reliability, and convergent and discriminant validity.
Method: A convenience and prolific sample (N = 294) was analyzed using CFA and SEM with Maximum Likelihood Estimation and an IRT with a Graded Response Model. Results: The bifactor model for D70 and single-factor models for D35 and D16 showed a good fit to the data. However, H and FD indicated that the five themes of D should not be modeled in the context of SEM or as subscales when using manifest scores.... (More) - Objective: The Dark Factor of Personality (D) measures the underlying core of all aversive traits. The present study evaluated the psychometric properties of the Swedish version of three item sets D70, D35, and 16 in regard to their factor structure, item quality and functioning, reliability, and convergent and discriminant validity.
Method: A convenience and prolific sample (N = 294) was analyzed using CFA and SEM with Maximum Likelihood Estimation and an IRT with a Graded Response Model. Results: The bifactor model for D70 and single-factor models for D35 and D16 showed a good fit to the data. However, H and FD indicated that the five themes of D should not be modeled in the context of SEM or as subscales when using manifest scores. Moreover, the relatively high ECV and PUC together indicated that despite the bifactorial theory, the D70 can be collapsed in a unidimensional measure without introducing much relative bias. Based on omega and Cronbach’s alpha, the very high reliability values of D70 and D35 could indicate item redundancies. IRT showed similar item quality and functioning to other languages. D provides more information about trait levels above zero. Regarding construct validity, D had high latent trait correlations to Psychopathy and Machiavellianism, no links were found to Narcissism. The correlation pattern with the HEXACO yielded similar results to previous studies in other languages. The especially high correlations with Agreeableness and Honesty-Humility are discussed in light of current research. Conclusion: The Swedish translation of D70, D35, and D16 is psychometrically sound and recommended for use in future research. Despite the better fit of the bifactor model for D70, based on FD, ECV, and PUC, in applied contexts we recommend the use of a single-factor model for D70. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9095305
- author
- Kurtz, Lara LU and Streckert, Nico LU
- supervisor
- organization
- course
- PSYP01 20221
- year
- 2022
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Dark Factor of Personality, CFA, IRT, Reliability, Validity, SEM
- language
- English
- additional info
- Both authors contributed equally to the thesis.
- id
- 9095305
- date added to LUP
- 2022-07-05 12:57:28
- date last changed
- 2022-07-05 12:57:28
@misc{9095305,
abstract = {{Objective: The Dark Factor of Personality (D) measures the underlying core of all aversive traits. The present study evaluated the psychometric properties of the Swedish version of three item sets D70, D35, and 16 in regard to their factor structure, item quality and functioning, reliability, and convergent and discriminant validity.
Method: A convenience and prolific sample (N = 294) was analyzed using CFA and SEM with Maximum Likelihood Estimation and an IRT with a Graded Response Model. Results: The bifactor model for D70 and single-factor models for D35 and D16 showed a good fit to the data. However, H and FD indicated that the five themes of D should not be modeled in the context of SEM or as subscales when using manifest scores. Moreover, the relatively high ECV and PUC together indicated that despite the bifactorial theory, the D70 can be collapsed in a unidimensional measure without introducing much relative bias. Based on omega and Cronbach’s alpha, the very high reliability values of D70 and D35 could indicate item redundancies. IRT showed similar item quality and functioning to other languages. D provides more information about trait levels above zero. Regarding construct validity, D had high latent trait correlations to Psychopathy and Machiavellianism, no links were found to Narcissism. The correlation pattern with the HEXACO yielded similar results to previous studies in other languages. The especially high correlations with Agreeableness and Honesty-Humility are discussed in light of current research. Conclusion: The Swedish translation of D70, D35, and D16 is psychometrically sound and recommended for use in future research. Despite the better fit of the bifactor model for D70, based on FD, ECV, and PUC, in applied contexts we recommend the use of a single-factor model for D70.}},
author = {{Kurtz, Lara and Streckert, Nico}},
language = {{eng}},
note = {{Student Paper}},
title = {{Can Your Darkness be Measured? Validating the Dark Core of Personality in Swedish}},
year = {{2022}},
}