Interpreting representativeness of reported average test scores when only a subset of admitted students take them?

Perhaps you are making it too complicated? The ranges represent the scores of those students who submitted and were admitted. If your daughter submits, that is what she is up against. No adjustment necessary.

In other words, with regard to your daughter’s application, it doesn’t really matter what the range “would have been if we were back in the good old days and everyone was submitting scores.” If her score doesn’t compare well to other similarly situated students (similar school type, region, transcript, ECs, essays, etc.) who are currently being admitted, then her odds are longer.


This is somewhat misleading. It is not as if MIT experimented with test optional, and then decided it it was a failure based on the results of the experiment. MIT was never truly “test optional.” Essentially MIT made a temporary accommodation for those who could not submit test scores because of Covid, but told students who could safely take the test should submit the scores. After more normalcy returned, MIT resumed requiring tests from everyone.

Here is how @MITChris put it in a different thread:

And in response the observation that “many have the misconception that MIT experimented with TO and that it was a failure, but that doesn’t seem to be the case at all,” @MITChris responded:

In short, MIT didn’t “reverse course” on Test Optional. It was never truly Test Optional.

4 Likes