ERIC Number: ED226294
Record Type: RIE
Publication Date: 1982-Aug
Reference Count: 0
Problems in Assessing Adverse Impact: 80% Rule and Statistical Significance.
Dalessio, Anthony; And Others
An essential first step in the investigation of employment discrimination is establishing that the selection procedure has had an adverse impact. The 80% rule and statistical significance (as determined with a chi-square test), two methods of assessing adverse impact, were compared considering the variables: number of applicants (N), proportion of applicant pool who are minorities, majority group selection ratio, and minority group selection ratio. Data analyses indicated relatively few conditions where the two methods led to the same conclusion. Problems found with the 80% rule were: (1) under common conditions adverse impact may be concluded when chance differences between majority and minority selection ratios have occurred; and (2) the 80% rule is sensitive solely to variations in majority group selection ratio. Problems found with the chi-square were that with lower values of N, majority selection ratio and proportion minority, statistically significant chi-squares cannot be obtained; at high values of N, majority selection ratio and proportion minority, trivial differences in selection ratios may result in statistical significance. The findings suggest that more attention needs to be focused on choosing an appropriate assessment of adverse impact. (Twelve tables of statistical data are included.) (Author/JAC)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A
Identifiers: Adverse Impact; 80 Percent Rule
Note: Paper presented at the Annual Convention of the American Psychological Association (90th, Washington, DC, August 23-27, 1982).