Institutional peer groups: Incorporating measures of graduate program and research quality into an existing peer institution selection proces

The purpose of the study was to analyze how changes in an established peer institution methodology impacted the institutions selected as peers using two universities as the 'home' institutions, Kansas State University (KSU) and the University of Kansas (KU). Changes in the composition of the peer groups were examined with regard to differences in average faculty salaries.;The Kansas Board of Regents peer selection methodology is well known in the literature. The current study changed the peer selection methodology in a number of ways. The first alteration involved adding a research component to the methodology that was based on (a) a combination of graduate program quality and effectiveness ratings, and (b) per capita research productivity measures. The second change involved removing certain threshold and methodology restrictions that limited the potential pool of peer institutions. Finally, the robustness of IPEDS data often used in peer selection methodologies was determined by comparing peers selected based on a survey conducted one year with peers selected based on multiple years of IPEDS data.;The results of the study found that the research and graduate program measures used in the revised peer selection methodology only impacted the University of Kansas peer group configurations but not the peer group configurations for Kansas State University. These changes in peer group configurations minimally impacted the common institutional benchmark of average faculty salaries. Other adjustments to the Kansas peer selection methodology in combination with the new research measures found a continuation of the pattern of only impacting the University of Kansas peer group configurations and minimal changes in differences in average faculty salaries.;When one, two, and four years of IPEDS data were used as part of the peer selection methodology different peer group configurations were selected for both KU and KSU. Although the change in peer group composition was, at the most, only one university, it is recommended that more than one year of IPEDS data be included in a peer selection methodology.