Determinants, Detection and Amelioration of Adverse Impact in Personnel Selection Procedures: Issues, Evidence and Lessons Learned

Hough et al. (2001) reported that White people were typically more motivated to do well on tests, more likely to believe tests would influence their future relationship with the company, and had more positive perceptions of the criterion validity of ability tests; women and older people have greater test anxiety.

“White people were typically more motivated to do well on tests, were more likely to believe tests would influence their future relationship with the company that asked them to take them and had more positive perceptions of the criterion validity of ability tests.Women and older people have greater test anxiety.”

Hough et al. (2001) found that culture-free, nonverbal ability tests still showed modest group differences, and attempts to address the confound of ethnicity versus culture by building tests with Black cultural components found little success in overcoming differences.

“They also found evidence suggesting culture-free, nonverbal ability tests still showed modest group differences and that attempts to address the confound of ethnicity vs culture by building tests with Black cultural components found little success in overcoming differences.”

Hough et al. (2001) found little research on test coaching in occupational contexts; some studies showed potential in addressing group differences, but effects were modest and inconsistent, and findings were more consistent where test takers were less familiar with tests.

“Little research had been done in occupational contexts and whilst some studies demonstrated potential in addressing group differences, effects were modest and inconsistent.Findings did seem to be more consistent where test takers were less familiar with tests.”

Hough et al. (2001) suggested that reducing differences in test perceptions would have at best a small effect on test score differences.

“findings were sometimes inconsistent and typically suggested reducing differences in test perceptions would have, at best, a small effect on test score differences.”

Hough et al. (2001) reported that group differences in test performance lowered when motivation scores were partialled, test motivation had a moderating effect on test validity, and validity perceptions positively correlated with test performance.

“Group differences in test performance lowered when motivation scores were partialled, test motivation had a moderating effect on test validity and validity perceptions positive correlated with test performance.”

Hough et al. (2001) found that differential item functioning generally had a weak impact on overall test scores where it existed.

“DIF, whereby two individuals possessing identical levels of a latent trait but belonging to different groups have different chances of getting a given question correct, was generally found to have a weak impact on overall test scores where it existed (Hough et al., 2001).”

Hough et al. (2001) concluded that cultural differences have no more than a modest impact on test score differences.

“They concluded that cultural differences have no more than a modest impact on test score differences.”