Accountability Report

Appendices

Appendix D
Defining the Graduation Efficiency Index (GEI)

The Graduation Efficiency Index (GEI), is intended to be a single, easily computable index of the efficiency with which students move through the state’s public higher education system.

The GEI is computed retrospectively for each graduate as follows:

This index avoids many of the problems that are associated with calendar years as a measure. Furthermore, it is a much more defensible measure of the efficiency with which an institution is producing graduates in that it takes the following five variables into account:

The total number of credits that have been earned. Earning more credits than the minimum required for the degree demands reduces efficiency.

The number of credits for courses that have been dropped. Dropped courses, leaving empty seats, adds to inefficiency.

The number of credits for courses that have been repeated, including failures. Students who fill the same seat twice add to inefficiency.

The minimum number of credits required by the major for graduation. Degree programs that legitimately require more credits do not necessarily add to inefficiency.

The number of credits that have been transferred. Given that students move from institution to institution, prior work needs to be recognized but not credited to the degree granting institutions.

The GEI incorporates these five variables in a straight-forward way. It has a number of advantages:

1. The GEI varies from 0 to 100% and is readily interpretable. For example, an efficiency rate of 90% has familiar meaning.

2. It can be used at all degree levels and all institutions where there is a standard and acceptable minimum number of required credits for a degree. (We have chosen to restrict our attention to undergraduate degrees here, this restriction is not necessary.)

3. It is equally applicable to full-time and part-time students.

4. It can be averaged for subsets of students; e.g., for each degree type, for each department, and for transfer vs. non-transfer students. In other words, it is an index that is very useful as a dependent or correlative variable for research and for illuminating problem areas, as will be demonstrated below.

5. It can be easily tracked over time to assess effects of interventions and policy changes to improve efficiency, either locally or globally.

Calculating the Graduate Efficiency Index

Enrollment Census Day Credits

The denominator of the GEI for each student is the sum of the Enrollment Census credits in which the student has enrolled over the course of his/her career at the given institution. At UW this is the tenth day of each quarter. We have chosen to sum enrollment census day credits for two reasons. First, we wanted to choose a time after which, if the students dropped a class, his/her place was less likely to be taken by another student. The tenth day at UW seems to be a good compromise - if one waits longer the efficiencies caused by dropped courses will be missed. If one chooses an earlier date, vacated spaces may be reassigned to other students. At the University of Washington (UW) it is possible for students to enroll after the tenth day with special permission - our index misses there. It is also possible for students to drop before the enrollment census day and have their space not filled. In this regard, the index is flawed.

The second reason is that student record systems make counting at the enrollment census day relatively easier than counting at any other time because it is the official day for reporting enrollments to the state. Practical circumstances such as this might require other institutions to define this variable somewhat differently. This redefinition should not be a major problem if the institution is internally consistent and is not comparing itself with others that use alternative definitions.

Total Credits Required For Graduation

This variable is intended to be represented by a value for each possible undergraduate degree type and major. It is not empirically determined but is based on what the official institutional catalog indicates. At UW it equals 180 quarter credits for most programs but can range up to 225 credits.

Transfer Credits

We recognized early on that transfer credits had to be an element of the index and that the index for transfer students should be comparable to though not necessarily equal to, that of non-transfer students. However, our database is such that we can determine the number of transferable credits for each student, but we cannot determine the efficiency behind these credits. In particular, when a transfer student's total credits are in excess of minimum credits for graduation it is not possible to attribute the excess credits to one or the other institution.

There appear to be two ways to handle transfer credits that lead to somewhat different results. The GEI, as it is defined above, subtracts transfer credits from the numerator. The implication of this choice is that the efficiency at the sending institution is estimated by the efficiency at the graduating institution and is weighted by the relative number of credits taken at both places.

An alternative formulation adds transfer credits to the denominator rather than subtracting them from the numerator. Efficiency as computed by this alternative formula is always equal to or larger than by the formula above. It implicitly assumes that all of the transfer work was done with 100% efficiency, that is, all course work taken contributed to graduation - no courses were dropped after the tenth day, no courses were repeated, no courses were nor needed for fulfilling graduation requirements, etc. By direct implication, it assumes that all inefficiency is manifest by the transfer student at the graduating institution.

Empirically, we found that the average efficiency index of transfer students was almost exactly equal to the average efficiency index of non-transfer students under the alternative version, yet we are confident that the transferring process adds some degree of inefficiency when one views the entire undergraduate career of a student, not just that at the degree-granting institution. Thus, we prefer the initial GEI formulation because of the theoretical rational explained above and because our empirical results suggest that the alternative formulation does indeed over-estimate the efficiency of transfer students. Furthermore, while the GEI is not necessarily strictly correct for every transfer student, we feel it is correct on average.

Normally, remedial courses, that is, courses deemed prerequisite to required courses but not counting toward graduation, would be counted as part of the enrollment credits. It might be tempting for institutions to suggest that remedial course credit not be counted as part of enrollment credits, thus assuring more comparability among institutions with varying commitments in that regard. However, not counting these credits denies the reality that some institutions will be less efficient than others by reason of their respective missions (see Trow, 1996, p. 5).

We include extension credits with transfer credits.

To see how these formulae differ, consider a student who transfers 90 quarter credits and who graduates with an additional 100 tenth day quarter credits in a degree program that requires 180. The proposed GEI would calculate the student index as [(180-90)/100] or 90%. The alternative version would calculate his index as (180/190) or 95%. In other words, each additional credit over the minimum subtracts more from the proposed GEI formula than from the alternative.

Accountability Report Appendices

Provost's Committee on Accountability
provost@u.washington.edu
June 1997