More on the grading issues in Kentucky’s schools

I blogged several days ago about “Which Kentucky school districts grade too hard or too easy?,” discussing evidence in the 2018 High School Feedback Workbook assembled by the Kentucky Center for Statistics. The data, which is for the graduating class of 2015-16, includes ACT scores and high school Grade Point Averages (GPA) for each school district in the state that has at least one high school. The results indicate grading standards across Kentucky’s school districts are nowhere close to consistent.

Today, I discuss another analysis derived from the Workbook data. I took a look at how well the Kentucky High School Class of 2016’s ACT performance in each district correlated to those students’ first year college GPAs. I also looked at the correlation between each district’s high school GPAs for that graduating class and the first year college GPAs those students earned.

I was curious to see if something we often hear – that high school GPAs are more closely related to college GPAs than the ACT – was shown by this data.

Quite simply, that often-heard claim isn’t supported by the Kentucky Center for Statistics data for the Kentucky High School Class of 2016.The correlation between the ACT and college GPA was 0.44 and the correlation between the high school GPA and college GPA was very slightly lower at 0.41. Based on this data, the high school GPAs are certainly no better than the ACT as predictor of college performance.

I need to stress that there are some important caveats to this simple analysis. Not all students in each district go on to college, so there is some “apples to oranges” in the comparisons here. Also, the college data only covers students who attended a 2- or 4-year college in Kentucky. GPAs for high school graduates going to out-of-state schools are not included. If the Kentucky Center for Statistics were to do a student-by-student tracking of the data only for those students who go on from high school to college, the results might look different.

Still, the best data currently available to the public indicates that in Kentucky – thanks to what our previous blog shows is uneven grading standards across the state’s public school systems – the common claim that ACT isn’t as good a predictor of college performance as the high school GPA just might not be true. And, folks who want to do away with such testing might need to hold off.

You can see my spreadsheet with the correlation calculations by clicking here.

Added comments about caveats on October 13, 2018