Our Statistics Mistake

It's always exciting to hear about the number of campuses where fraternity/sorority members exceed the average GPA at the regional leadership conferences every year. But something struck me this time: we have an embarrassing statistics problem. Here is an illustration:

  • Imagine the all fraternity/sorority GPA is a 2.8.
  • There is a minimum standard of 2.0 to join and remain a member.
  • Our population therefore includes only those students above a 2.0.
  • The all campus GPA is a 2.75, so we celebrate the fact that our average is higher.
  • But the average all campus GPA is an average of every student, including those below a 2.0.

 

Academics Diagram.png

Do you see the problem? The populations don't match! Comparing the all student GPA to the all fraternity/sorority GPA is a horrible mistake. If you randomly select any group from those students above a 2.0, you naturally and automatically get a higher average GPA!

This raises a few important questions. Does your fraternity/sorority community exceed the GPA by a great enough margin to account for the flawed comparison? After factoring out our standards, do fraternity/sorority members perform better in their academic career than other students who meet the same standard? And if so, does involvement in a fraternity/sorority improve your academic ability, or is some other factor responsible?

We need a more telling statistic:

  • Weight the all-campus GPA comparison to include only those students who meet or exceed our standard.
  • Establish a system that measures GPA before and after joining. On average, how much does academic performance change for fraternity/sorority members versus the average student?
  • Compare the proportion of fraternity/sorority students on the Dean’s list to the proportion of all students on the Dean's list. And there are many more to chose from.

We say we take our commitment to scholastic achievement and our high academic standard seriously, but in this case, the current standard is embarrassingly lower than average. People rise to the expectation you set, and it’s time for a new one.