Institutional Research Reporting & Analytics About ODA

ODA Home > Institutional Research > Faculty and Staff > Faculty Salaries > Comparisons Over Time

Calculation differences between CU-Boulder and CU System

Comparison with analyses by the CU System. Peer comparisons can be done by different methods. Prior to 2006, due to differences in data available to CU-Boulder (in Office of Data Analytics), we typically calculated unweighted average salary and compensation figures while CU system officials calculate weighted averages. The story was the same in both cases but the precise numbers differed in the two sets of results.

CU Boulder completely revamped its calculation methods for faculty compensation comparison in AY 2014 to allow for more sophisticated comparisons (e.g., limiting the comparison group to institutions without medical schools). By default, the institutions included in the comparison group of AAU public peers should be the same as the group used by CU system officials. However, beginning in AY 2014, CU Boulder changed the default calculations to include only faculty with 9-10 month salaries. This change was implemented because we learned that some institutions may appoint temporary faculty with a professor title (and 11-12 month contract), whereas the practice at CU Boulder is to appoint such faculty with a different title (e.g., Visiting Professor). By comparing CU Boulder faculty salaries with those of faculty with 9-10 month contracts at peer institutions, we are making a more fair comparison.

Although CU Boulder now calculates weighted averages the same as CU system officials, there are other differences in our calculation methods that may explain differences in our numbers over time. Here we call the two sets the "ODA results" (for Office of Data Analytics) and "system results" (for CU system). Here are some of the differences that we are aware of:

  • Peer institutions included
    • In ODA results, 2001-02 and later data consists of all non-union AAU public institutions including Texas A&M and SUNY-Stony Brook. Calculations for 2000-01 and earlier exclude Texas A&M and SUNY-Stony Brook.
    • System results do not include data for Texas A&M and SUNY-Stony Brook for all years prior to 2004-05.
    • Starting in AY 2014, ODA results include data for Georgia Tech going back to 2002.
  • Data differences prior to AY 2006
    • System results include UCB figures whereas CU-Boulder generated graphs exclude CU-Boulder figures.
    • Data precision is different since system calculations use exact averages for each institution, while the CU-Boulder graphs include figures published by AAUP, which are rounded to the nearest $100 by institution.
  • The way the peer averages were calculated prior to AY 2006.
    • Averages are always within rank, within year, in both methods
    • Consider a case with just two peer institutions
      • A: salary $50,000, N of faculty at that rank 100
      • B: salary $80,000, N of faculty at that rank 200
    • ODA calculates a straight average: $50,000+$80,000 / 2 = $130,000/2 = $65,000. This method captures the salary at the "average institution"
    • System uses an average calculated by AAUP, which weights by N of faculty at that rank. ((100*50,000) + (200*80,000)) / 300 = 210,000/3 = $70,000. This method captures the "average of all faculty in peer institutions" in the specified rank.
    • Institutions with higher salaries in each rank tend to have higher proportions of their faculty at higher ranks (e.g., Berkeley is over 65% full profs). Therefore the system/weighted method produces relatively higher peer averages at higher ranks, lower at lower ranks, than the ODA unweighted method.
    • PBA's analyses of salaries by college and discipline also use unweighted averages.

Last revision 05/27/16

ODA Home  |   Institutional Research  |   Reporting & Analytics  |   Contact |  Legal & Trademarks |  Privacy
15 UCB, University of Colorado Boulder, Boulder, CO 80309-0015, (303)492-8631
  © Regents of the University of Colorado