ERIC Number: ED447149
Record Type: Non-Journal
Publication Date: 1999-Apr-19
Reference Count: N/A
Evaluating Faculty Salary Equity Using Hierarchical Linear Modeling.
Stapleton, Laura M.; Lissitz, Robert W.
This paper presents results from a comparison of the multiple regression (MR) approach to examining faculty salary equity (with clusters for the various disciplines) and hierarchical linear modeling (HLM) for the same problem. The comparison was done in two steps. First, a practical example of applying both techniques, using empirical data, is provided, and then a simulation study examined varying data conditions to show the differences resulting from these two statistical approaches. For the empirical example, data were used for 1,216 tenured and tenure-track faculty grouped into 7 discipline clusters. Variables were: (1) discipline; (2) rank; (3) years in rank; (4) refereed articles; (5) presentations; (6) sponsored research; and (7) gender. For the simulation, 5 discipline clusters, each of 10 departments, were generated for 20 faculty observations for each, a total of 1,000 observations. The empirical study showed that both approaches provided similar results, but that the HLM model could provide more information, although with a more complicated picture of the salary process. Simulation results confirm that if departments within discipline clusters are not homogenous with regard to the percentage of males and salary levels (controlling for all other variables), the analysis may be biased when using MR. An appendix shows cluster averages and the range of department averages. (Contains 14 tables, 4 figures, and 18 references.) (SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A