Investigating the Properties of the General Factor (M) in Bifactor Models Applied to Big Five or HEXACO Data in Terms of Method or Meaning

Abstract

This project applied a bifactor model, which specifies a general factor that accounts for the common variance among all scale items, and group factors that reflect additional common variance among clusters of items. This general factor is designated as “M” because of a presumption in the research literature that its origins are to be found in method. The model was applied in eight samples using nine datasets and across three different personality measures, including the Big Five and the HEXACO. Inclusion of M significantly increased model fit and increased the variance explained of items. Evidence showed that M did not reflect aspects of method such as random error or an acquiescent response bias. M correlated positively with variables suggesting psychological adjustment and negatively with variables pointing toward maladjustment. M showed unique relationships with constructs suggesting psychological adjustment over and beyond the Big Five. Data supported an interpretation of M as a synthesizing function within the self and thus suggested that M was a construct revealing substantive psychological meaning rather than mere method.