Standardized Tests: Causal Or Correlational?

by Steve Wyckoff on January 25, 2010

I’ve given a lot of thought to the standardized test phenomenon. How is it that so many well-intentioned and highly intelligent people can have so much faith in such a detrimental process? I think I might have at least part of the puzzle figured out. We’ve been using standardized tests for decades, and educators through those decades have tried to convince policymakers and the public alike what a great job we’re doing by sharing our standardized test scores.

And we’ve convinced the public and policymakers how important they are. The problem is we liked our test scores when we could pick and choose which scores to report. Now that we have to report the scores of all kids they’re not, in our minds, nearly as representative of our efforts and success as they were when we only reported the best scores.

But I think the bigger issue is we’ve confused what standardized test scores represent to us. Over the years we became aware that bright, well-educated, highly successful students, did very well on standardized tests. We interpreted that to mean that if you have high standardized test scores you will be bright, well-educated, and highly successful. Our reaction as educators, and the reaction of policymakers was to say that if we had more and more kids obtaining high scorers on standardized test they two would be bright, well-educated, highly successful.

But the relationship is not causal, it’s correlational. And on top of that we’ve reversed the relationship:

Highly successful =  high test scores

THEREFORE …

High test scores =  well-educated

That appears to be the logic that all of us, educators and policymakers alike, seem to be following. Yet when I looked at the issue in these terms I didn’t believe it at all. We’re proving that we can raise test scores but there is a growing sense that our kids are less well educated.  And that doesn’t even begin to address the issue of what “well-educated” means!

We’ve approached the evidence produced by standardized tests as if the relationship between the test and student success is causal. But in reality there is simply a correlational relationship between kids who are successful and their scores on standardized tests.

Previous post:

Next post: