Are individual subject high school grades better predictors of college subject performance than the ACT?

Well, not in Kentucky!

In case you missed my earlier blog about high school grade point averages (GPAs) versus the ACT college entrance test as predictors of college performance, there is an effort underway across the country and in Kentucky to reduce attention to the fairly disappointing scores from the ACT college entrance test. While the ACT certainly isn’t perfect, I think the attempts to completely ignore this college entrance test are misguided. And, I have data to back up my opinion.

In my earlier blog, I analyzed data from the Kentucky Center for Statistics for the Kentucky public high school graduating classes of 2014 and 2015 to explore how freshman year college GPAs in Kentucky’s 2- and 4-year public postsecondary schools compare to the overall high school GPAs and the ACT Composite Scores. In an earlier blog I also explored evidence of massive variations in grading standards that are found in Kentucky’s public high school system. Taken together, these two blogs provide considerable, data-based evidence that grading in Kentucky’s high schools is highly variable, which makes it no surprise that when I examined if the ACT Composite Score or high school GPA was a better predictor of how the graduates of each Kentucky high school performed in their freshman year in college, the ACT came out as the clearly stronger (though far from perfect) indicator.

But, there is more to the story.

In discussing these findings with one of the state’s education data professionals, a comment was made that while the overall high school GPA might not correspond well to college freshman year performance, this was not true for grades in individual subjects. The claim was that individual high school subject grades did do a better job than the ACT’s individual subject grades at predicting freshman year results (ACT separately reports scores for English, math, reading and science in addition to an overall ACT Composite Score).Fortunately, the same Kentucky Center for Statistics custom spreadsheet that I used for the earlier blog about ACT Composite versus high school overall GPA also contains individual student grades for specific high school subjects and individual freshman college grades in addition to the ACT scores for individual subjects.

So, I was able to look at the correlations between high school and college grades for individual subjects and the ACT individual subject scores. Again, I looked at results for the public high school classes of 2014 and 2015.I assembled the results into a new spreadsheet you can access here. Table 1 summarizes how the various correlations between high school grading and the ACT individual subject scores related to college grading.

Table 1

Correlations for Individual ACT Subjects and HS and Freshman College Subject Grades

As you can see, for both the Class of 2014 and 2015, across all common ACT subjects of English, math and science, in every case the ACT was a better predictor of the freshman year individual course grades than were the high school senior year grades in that subject.

So, the evidence continues to mount. While the ACT certainly isn’t perfect as a college performance indicator, it is definitely better, on average, than the grades Kentucky students are getting in their high schools.

Parents, students, and those involved with college programs need to take due notice.

If you want more details about this analysis, just click the “Read more” link.

Here are some specific details about the spreadsheet I used for my analysis.

Each of the worksheets in my analysis (except the “Correlation Summary” tab shows results for all Kentucky regular public high schools that had relevant data. Note that due to student privacy laws, some schools are not listed due to small student groups of less than 10 students.

All of the data consistently considers only students who graduated from high school in the year listed and then went on within the year to a Kentucky public 2- or 4-year postsecondary school.

Each worksheet except for the summary lists three sets of scores. Using English as an example, the scores include:

• “Avg English: senior year”

This is the overall average numerical grade for senior English for each high school’s graduating class in the year listed. For the Class of 2014, for example, this is the senior class average for the 2013-2014 school year.

• “Avg English: first year of college”

This is the overall average numerical grade for all the college freshmen from each high school for English. For the high school class of 2014, this is for college freshmen in the 2014-2015 school year.

• “Mean Junior-Year ACT English”

The ACT scores are from the administration of ACT to Kentucky public school juniors for the appropriate year. For the high school graduating class of 2014, this is each high school’s ACT English Score from the 2012-2013 11th grade testing.

The other worksheets have similar titles but cover the other subjects of math and science.

Scroll to the bottom of each individual subject worksheet page and several sets of summary data are presented. These include the simple averages across all listed Kentucky high schools for the senior year high school subject grade, the freshman year college subject grade and the overall ACT average. Recall that the ACT averages for each school and overall average are only for those high school graduates who went on to a Kentucky 2- or 4-year public postsecondary school. These numbers generally will not agree with the reported ACT data for the same years from the Kentucky Department of Education, which include all 11th grade students.

Also, at the bottom of each subject worksheet are the correlations between ACT scores and college freshman year subject grades and the high school subject grade and the college freshman year subject grades. All of these are summarized on the first tab in the spreadsheet.