This page compares the 2009 and 2020 results. Because the 2009 survey offered limited options to endorse combinations of views, this page uses the "exclusive" answer counts, which do not include combinations. You can view longitudinal comparisons of three types:
- Comparable departments compares the answers of faculty members at departments in the 2009 survey (99 departments) with those of faculty at comparable 2020 departments (100 departments), where departments are selected based largely on rankings in both cases.
- Same departments compares the 2009 and 2020 answers of faculty members at the departments we had elected for surveying in 2009.
- Same people compares the 2009 and 2020 answers of the target 2009 respondents.
- Percentages: shows the percentage of respondents who accept or lean toward a given answer exclusively.
- Weights: for each answer, a respondent counts for 100 if they accept it (exclusively), 75 if they lean toward it, 50 if they selected a neutral answer ("agnostic", "too unclear to answer", and "no fact of the matter"), 25 if they lean toward another option, and 0 if they reject it. Weights are averaged over all respondents who give an exclusive answer or a neutral answer.
- Change is the 2020 percentage (or weight) minus the 2009 percentage (or weight).
- Swing toward an answer is defined as the relative strength of that answer in 2020 minus its relative strength in 2009. The relative strength of an answer is defined as the percentage (or weight) of respondents associated with that answer minus the average of the percentages (or weights) associated with all answers to the same question. For comparative purposes, swings are more meaningful than changes, as constant upward or downward changes across all options cancel out, and swings are guaranteed to sum to zero.