The Predictive Validity of Subtle and Obvious Empirically Derived Psychological Test Items Under Faking Conditions

Faking the MMPI-2: Utility of the Subtle-Obvious scales

“…As was expected, response set did significantly affect profiles for the 10 basic clinical and traditional validity scales, confirming Worthington and Schlottmann's (1986) conclusions. Findings regarding the new Back F scale serve to validate this scale, as subjects in the different response set groupings did indeed differ significantly in the predicted manner from one another in their endorsement of items contained within this new scale.…”

“…For instance, Schretlen (1990) has pointed out that using the overall obvious-subtle difference to identify faked profiles is likely to lead to false positives among clinical populations with severe psychiatric illness, as obvious T-scores increase in direct correlation with elevated clinical scales. Worthington and Schlottmann ( 1986) found that current subtle items on the PD scale were not as resistant to faking as was previously believed. Austin (1992), using the MMPI-2, questioned the utility of Subtle-Obvious items particularly with regard to their ability to detect faking good response sets, despite past research that suggested that this may actually be a particular strength of these scales (Grow, McVaugh, & Eno, 1980).…”

“…Research with the MMPI Subtle-Obvious scales has been consistent only in suggesting that response set of a subject (or client) does indeed affect MMPI profile scores and that it is therefore critical that a clinician be able to identify fake good and fake bad response sets in a profile (cf., Worthington & Schlottmann, 1986). Some researchers have found subtle items to be more resistant to faking than obvious ones and, hence, concluded that these items were more useful in terms of profile validity predictions (e.g., Beal, 1989).…”

Situational Judgment Tests and their Predictiveness of College Students’ Success: The Influence of Faking

“…Some studies have reported that applicants do not fake personality tests, and even if they do, it does not negatively affect their validity (Abrahams, Neumann, & Githens, 1971;Cunningham, Wong, & Barbee, 1994;Ellingson, Smith, & Sackett, 2001;Hough, 1998;Hough, Eaton, Dunnette, Kamp, & McCloy, 1990;McCrae & Costa, 1983;Ones & Viswesvaran, 1998;Ones et al, 1996). Conversely, other studies have found that faking occurs in selection settings and attenuates the criterion-related validity of personality tests (Anderson, Warner, & Spencer, 1984;Douglas, McDaniel, & Snell, 1996;Dunnette, McCartney, Carlson, & Kirchner, 1962;Kluger, Reilly, & Russell, 1991;Rosse et al, 1998;Schmit & Ryan, 1992;Worthington & Schlottmann, 1986;Zickar, 1997). For example, Schmit and Ryan (1992) found that the validity of a test for predicting GPA was lower for students who received an incentive to misrepresent themselves.…”

Faking personality profiles on a standard personality inventory

“…give the worst possible impression, as well as those instructed to be honest in their responses (e.g. Kuntz, 1974;Farley and Goh, 1976;Harvey and Sipprelle, 1976;Holden and Jackson, 1981;Burkhart, Christian and Gynther, 1978;Worthington and Schlottmann, 1986;Furnham and Craig, 1987). The results, while varying to some extent as a function of the specific inventory chosen, generally provide conclusive evidence: Most personality inventories, including standard instruments like the MMPI (Hathaway and McKinley, 1946), the EPQ (Eysenck and Eysenck, 1975), and the PRF (Jackson, 1967), consist of scales that are susceptible to faking and response biases which cast doubts on their utility as valid and reliable instruments in personality diagnostics.…”