| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 890145 | Personality and Individual Differences | 2015 | 6 Pages |
•The bi-factor model with a negative group factor best represented the CSES structure.•Ignoring the wording effect would overestimate the homogeneity of CSES data.•The impact of wording effect on the composite reliability of CSES was negligible.•The impact of wording effect on the criterion validity of CSES was non-negligible.
“Wording effect” refers to the systematic method variance caused by positive and negative item wordings on a self-report measure. Using the Core Self-Evaluation Scale (CSES) as an example, this study modeled the wording effect, and examined the impact of such wording effect on the reliability and validity of CSES. Five competing models were compared, and the findings suggested that the bi-factor model with a negative group factor best represented the CSES structure. The impact of wording effect on the reliability and validity of CSES was investigated, and the results showed that such impact existed and was non-negligible. Failure to model the wording effect would lead to biased estimates of the reliability and criterion-related validity of CSES. Limitations of the study are discussed, and considerations for future research directions are suggested.
