Evaluating Michigan Schools MME Results with Regression Analysis

In my last post, I looked at factors impacting Michigan primary school test performance.   In this post I look at high school performance using the Michigan Merit Examination (MME)  11th grade  results to determine what is the  role of race, location, poverty, and school type (public, not for profit, and for profit charter).   In all of the exam subjects (Math, Reading, Science, Social Studies, and Writing) for profit charters out perform not for profit charters and public schools at a 95%, or higher, confidence level.  Not for profit charter schools typically outperform public schools at lower confidence level than their for profit comparisons because their are so few in the data (only 16 as opposed to 79).  The higher the ratio of Asian and female students, the better the school performs on the exams.

One of the advantages of a regression analysis is that it can be used to forecast what the result should be based on the factors involved.   Some schools will vary more (above and below) the forecast.  I hope to illuminate that distinction using an online map in a future post.

Continue reading