The table shows results of fitting a regression model to data on Oklahoma State University salaries (in dollars) of 675 full-time college professorsof different disciplines with at least two years of instructional employment. All of the predictors are categorical (binary), except for years as professor, merit ranking, and market influence. The market factor represents the ratio of the average salary at comparable institutions for the corresponding academic field and rank to the actual salary at OSU. Prepare a summary of the results in a couple of paragraphs, interpreting the effects of the predictors. The levels of ranking for professors are assistant, associate, and full professor from low to high. An instructor ranking is nontenure track. Gender and race predictors were not significant in this study.
Modeling professor salaries

Dependent variable is the logarithm of annual salary. Model summary: Adjusted R2 = 0.94; F- ration = 411.76; N = 675; P-value = 0.001. Source: Some data from New Directions for Institutional Research, no. 140, Winter 2008 (www.interscience.wiley.com).

Transcribed Image Text:

Variable Intercept Nontenure track Instructor Associate professor Full professor Years as professor Average merit rating Business Education Engineering Fine arts Social science Market influence Estimate 17.870 0.010 0.284 0.170 0.407 0.004 0.044 0.395 0.053 0.241 0.000 0.077 7.046 Std. Error 0.272 0.011 0.018 0.013 0.018 0.001 0.005 0.015 0.015 0.014 0.018 0.013 0.268