At a university of about 800 faculty members offering baccalaureate, masters, and doctorate programs a study was conducted to determine the relationships between recommended salary increases and evaluation of performance. Salary increase proposals were submitted for faculty by department chairpersons. Among other items of information, the proposals contained three categories of data treated in this report: (1) recommended salary increase; (2) workload assignment; and (3) chairperson's evaluation of performance. Correlation analysis was applied to the variables for total faculty and for faculty clustered by discipline categories. Correlations between merit salary increase for teaching and evaluation of teaching and evaluation of overall quality were about the same--moderate (positive). Correlation between salary increases for activities other than teaching and evaluations of those activities differed somewhat from teaching. Merit increase for research correlated (positively) with evaluation of research to a considerably higher degree than was the case for teaching. In general, there was higher correlation between the different variables with respect to evaluation than between salary increases and evaluations of the merit increase categories. The conclusions is that correlations are weak enough to cast doubt on the value of a chairperson's evaluation in granting salary increases. (Author)