Using the most recent institutional data reported in IPEDS, we examined four-year public and not-for-profit institutions with complete data for 2004 to 2014. The data available for first-year retention and six-year graduation always refers to the first-time, full-time, degree-seeking freshmen cohort.
We developed a linear regression model using the most recent year of data (2014). In this model, the dependent variable was a Student Success Score. As the sum of first-year retention and six-year graduation rates, this combined index could range between 0 and 200.
The linear regression was developed by examining the influence of demographic, academic, social, and financial variables available in IPEDS on the overall Student Success Score.
From the model we developed two metrics. The first is the absolute difference between the predicted and actual Student Success Scores. This metric penalizes institutions that do not have much room for improvement and gives a scoring advantage to those that have more room for improvement. To adjust for this, we developed the second metric, the percentage of an institution’s over- or under-performance given the potential for improvement.
We also looked at two very straightforward metrics to measure institutional trends:
- Absolute change in first-year retention rate between 2004 and 2014
- Absolute change in six-year graduation rate between 2004 and 2014
To ensure that our analysis compared similar institutions, the normalized all metrics solely within the context of common institutional categories. These categories are based on institutional control and Carnegie Basic Classifications:
- Public Research/Doctoral
- Public Master’s
- Public Baccalaureate
- Private Research/Doctoral
- Private Master’s
- Private Baccalaureate
Within each of the six institutional categories, each of the four scores (predicted vs. actual, performance vs. room for improvement, retention trend, and graduation trend) were converted to normalized scales (0 to 100). The final score weighted each of the four metrics equally to also produce a normalized score that ranges between 0and 100.
Because Eduventures’ Student Success Score only compares reasonably alike institutions, a private master’s-level institution’s score is not comparable to a public master’s-level institution’s score.
Don’t See Your Institution?
If you do not see your institution in the rankings, it may be for one of two reasons:
- It had incomplete information throughout the time frame
- It is identified in a specialist Carnegie Basic Classification (e.g., health professions, special focus, tribal college)
If you are an Eduventures member, we can run the analysis for you if you provide the complete information. Contact us to get started!