The general factor of personality is very robust under faking conditions

The nature of the general factor of personality (GFP) under faking (good and bad) conditions is examined across four independent archival samples. In each sample, participants were randomly assigned to complete a personality measure under honest, fake-bad, or fake-good instructions. The factor structure of the GFP was examined for each condition and, across all four samples, the GFP was most robust under the fake-bad instructions (highest factor loadings and accounted for most of the variance), followed by the fake-good conditions, and the GFP was the weakest in the honest or standard instruction situations. Because the structure and composition of the GFP changes under different test-taking instructions, these results suggest that researchers interested in studying the GFP must consider the impact of test-taking conditions.