The negative wording factor of Core Self-Evaluations Scale (CSES): Methodological artifact, or substantive specific variance?

- The optimal factor structure of Core Self-Evaluations Scale (CSES) was investigated.
- Ignoring variance captured by reversed items produces biased parameter estimates.
- The wording factor showed specific relation with negative affect.
- The negative wording factor may contain substantive variance.

Among the instruments used to assess core-self evaluations, the Core Self-Evaluations Scale (CSES; Judge, Erez, Bono, & Thoresen, 2003) is commonly used. A recent study (Gu, Wen, & Fan, 2015) has revealed a method factor associated with the negatively worded items of the CSES. Ignoring this systematic variance produced biased validity and reliability estimates. Our objective was to replicate the findings of Gu et al. in two independent Chilean samples, and further investigate the magnitude of the wording factor and its relationship with external criteria. The wording factor explained one third of the common variance, acquired sufficient model-based reliability for psychometric interpretation, and showed a moderate relation with negative affect, above and beyond the core self-evaluations' main construct. Moreover, ignoring the wording factor produced biased correlations between CSE and negative affect. These results suggest that the negative wording factor of the CSES may represent more than a response artifact, i.e., substantive multi-dimensionality that should be investigated in-depth.