Validity of evaluative factors from Big Five and HEXACO questionnaires

Open Peer Commentary and Author's Response

“…We do, however, disagree with the authors' assertion that "no candidate set of seven or more factors has yet been identified" (p. 11) as a meaningful improvement over the six factors of the HEXACO model. Specifically, we feel that the evaluative factor investigated by Biderman and colleagues (Biderman, McAbee, Chen, & Hendy, 2018;Biderman, McAbee, Hendy, & Chen, 2019) represents an important addition to the HEXACO model (and Big Five) as presented in the focal article. Ashton and Lee (2020a) argued against adding a seventh factor, stating that adding such a factor does not provide more information concerning variance accounted for above and beyond the existing six factors.…”

“…[It should also be noted that Bäckström and colleagues (Bäckström, 2007;Bäckström & Björklund, 2016;Bäckström, Björklund, & Larsson, 2009;Bäckström, Björklund, & Larsson, 2014) have applied similar models to Big Five questionnaires, although these authors ascribe this factor to the social desirability characteristics of the items.] Moreover, Biderman et al (2019) demonstrated that this evaluative factor meaningfully related to an important achievement outcome, college student GPA, and that the evaluative factor was second only to Conscientiousness in predicting variance for this outcome (see Biderman et al, 2019, Table 5). The general take-aways from this collection of studies is that there are aspects of item content that people respond to in addition to the ostensible construct (Extraversion, Agreeableness, etc.)…”

“…Reise, 2012)particularly those estimated using exploratory structural equation modeling (ESEM) procedures (see, e.g. Asparouhov & Muthén, 2009;Marsh et al, 2010;Morin, Arens, & Marsh, 2016)-more appropriately represent the collection of traits that influence responses to personality questionnaires (Biderman et al, 2018(Biderman et al, , 2019.…”

“…Bifactor models, such as those applied by Biderman et al (2018Biderman et al ( , 2019 to HEXACO data, specify an additional factor indicated by all of the items on the questionnaire. Such factors are called general factors in the literature on bifactor models (Reise, 2012).…”

“…Biderman et al (2018) provided evidence that such factors are not merely method factors whose influence is independent of item content; rather, they are analogous to the domain factors and are dependent on the content of items. Specifically, these authors suggested that the aspect of item content captured within the general factor is affectively laden or evaluative item content, which they later termed evaluation (Biderman et al, 2019). [It should also be noted that Bäckström and colleagues (Bäckström, 2007;Bäckström & Björklund, 2016;Bäckström, Björklund, & Larsson, 2009;Bäckström, Björklund, & Larsson, 2014) have applied similar models to Big Five questionnaires, although these authors ascribe this factor to the social desirability characteristics of the items.]…”