More MBA reflections

As I read for my MBA courses, I occasionally reflect on how this data might apply to the JBU context. Here's an example.

For my MMBA work, we've been reading about balanced scorecards, which is sort of what cabinet has been doing as it determines its top 10 indicators to benchmark on an executive "dashboard." One of the benchmarks that all businesses use is market share because it doesn't really mean much if the numbers are going up when everyone else you're competing with is going up even faster. Is there any way that we can address that issue in our various assessments JBU?

For example, I've seen charts from you in the past that show how many high school grads there are by state. If we picked our top 5-6 six states and then tracked how many students we were getting from those states, would that give us a meaningful number for market share for our TUG recruiting? Could we get even more precise than that by tracking types of grads (ACT, types of schools, etc.)?

Similarly, Dick once sent me a comparison to LeTourneau in terms of their total potential recruiting population and how many students they actually have vs. what JBU was doing. That type of analysis could probably be expanded to include a few other key competitors to see how well we're doing in terms of market share in G&PS.

On a related note, I've been reading a bit about regression analyses applied to possible customers to determine which customers should get what sort of preferential treatment. I immediately thought of financial aid as an example. I know that Rich Sherry at Bethel does some of this analysis for his institution, and I assume that Noel-Lovitz runs many of their processes along these lines.

Do we and other CCCU institutions have similar regression analyses on past data that we use in determining which applicants are likely to matriculate, at what levels of financial support, and will likely persist and graduate at what sort of final GPA? I've chatted with some stats people in the past about such ideas, and they appeared to believe that such approaches would be doable at JBU (depending on the confidence intervals you're shooting for, of course), so maybe the idea's worth exploring?