Abstract
We examined the effects of warnings and speeding on scale scores and convergent validity of a measure of Conscientiousness in a faking context (N = 329). A completely crossed 2 × 2 experimental design was used in which instructions (no warning or a warning) and speeding (with or without a time limit) were manipulated. No statistically significant effects on scale scores or convergent validity were evidenced for speeding. Warning participants did decrease Conscientiousness scores by almost 1 standard deviation (d = .91). Warnings also moderated the relationship between self- and observer-ratings of Conscientiousness such that the relationship between self- and observer-ratings was statistically significant and positive in the warned conditions (r partial = .29, p < .01, n = 136), but this same relationship was statistically nonsignificant (r partial = −.14, p > .05, n = 148) in the unwarned conditions.
Notes
1We conducted a literature search in the ISI Web of Science database using the search string: Topic=(faking) and Topic=(personality). The literature search was designed in such as way as to examine how personality and faking research had progressed numerically after several seminal personality review articles had reexamined the validity of personality for personnel selection purposes (CitationBarrick & Mount, 1991; CitationHough, 1992; CitationTett, Jackson, & Rothstein, 1991). The following were the number of publications (average per year in parentheses) identified using the above search string for each of the following five year intervals: (1) 1992–1996: 36 (7.2); (2) 1997–2001: 56 (11.2); and (3) 2002–2006: 69 (13.8).
2The original sample consisted of 330 participants. We identified one multivariate outlier which we dropped from analyses. This resulted in a reduction of sample size from 330 to 329 in self-ratings and from 285 to 284 in observer-ratings.
3The job advertisement was adapted from a business general manager listing on the Human Resources Development Canada Job Bank Web page (http://jb-ge.hrdc-drhc.gc.ca) with identifying information on the employer and address made fictitious.
4The original sample consisted of 574 observers. However, two observers provided ratings with no variance across the Conscientiousness items indicating careless responding. We dropped these two observations from all analyses, which did not result in a reduction of sample size of matched pairs.
aThe WPT manual (CitationWonderlic, 1992) reports test–retest coefficients of .82 to .94.
*p < .05.
**p < .01.
***p < .001.
*p < .05.
**p < .01.