Conclusions one would make from looking at average faculty salaries in a specific department were compared with those conclusions evolving from a regression analysis of the same salaries. The standard salary study does little more than report the average salary associated with each faculty rank within each department, while a regression study takes into account the qualifications of all the involved faculty. The comparison of the two approaches reveals contradictions, an indication that faculty are not homogenous in background, that such differences do not average out across a department, and that the results of the regression analysis are in some instances drastically different from results derived from average salaries. The results suggest that the comparison of average salaries may lead to completely erroneous conclusions about the adequacy of an individual's or even an entire department's salary level. The results also indicate a need for further study, particularly one that would use a larger and possibly more homogeneous department such as economics or mathematics. Given the current emphasis placed on salary equity and the penalties for unexplainable differences, the results suggest that perhaps the additional effort required by a regression analysis is not unwarranted. A simple exchange among institutions listing only four pieces of data (rank, department, academic year equated salary, and time in rank for each unidentified faculty member) could support a regression analysis that would be far more accurate than the use of average salaries. (SW)