ERIC Number: ED232576
Record Type: RIE
Publication Date: 1983-May
Reference Count: 0
Multivariate Alternatives to Regression Analysis in the Evaluation of Salary Equity-Parity. AIR 1983 Annual Forum Paper.
Carter, Richard D.; And Others
The use of canonical analysis and multiple discriminant analysis to analyze equity-parity in colleges and universities is assessed and distinguished from multiple regression analysis. Multiple regression analysis forces the variable weights throughout the salary structure to be uniform, permits only one criterion or dependent variable to be examined at a time, restricts the coordinates to those provided by the variables as measured, and treats qualitative or discrete variables as if they were continuous, assuming ordinality of the numbers used. Canonical analysis and multiple discriminant analysis, both of which define new coordinate systems evaluation of dimensions underlying salary decisions, are multivariate alternatives to multiple regression analysis. These alternative techniques identify new coordinates in terms of the eigenvalues and eigenvectors of the appropriate correlation-covariance or dispersion matrices. Canonical analysis and multiple discriminant analysis were used in an annual salary adjustment process at the University of Wisconsin-Superior. Data for all full-time ranked faculty with a 100 percent appointment were analyzed for 2 successive years. Statistics for the 113 faculty members in the 1981-1982 study and the 111 faculty members in the 1982-1983 study are examined. (SW)
Publication Type: Numerical/Quantitative Data; Speeches/Meeting Papers; Reports - Research
Education Level: N/A
Authoring Institution: N/A
Identifiers: AIR Forum
Note: Paper presented at the Annual Forum of the Association for Institutional Research (23rd, Toronto, Ontario, May 23-26, 1983).