This is the Cato Daily Podcast for Thursday, September 6, 2018.

I'm Caleb Brown.

When research comes forth proclaiming that some schools are better than others or as in a recent study that some schools are definitely not better than others

What is left out? What is essential for that research to prove its claims?

Corey DeAngelis is a policy analyst at the Cato Institute.

We discussed education research and what alert readers should look for

when trying to evaluate it.

There is new research from the University of Virginia and the result as it's been interpreted in public media is you get no better results from private schools over public schools for low-income students.

So what do you make of that?

This study published by Pianta and Sari is his colleague simply looked at a thousand students that were born in

around 1991 over until they were about 15 years old.

And they simply did a observational study where they controlled for some characteristics

of students in public and private schools around the country and they found overall

that there were no results that there was no statistically significant correlation

between the private school outcomes and the public school outcomes.

Okay, and what should we, when we look at a study like that, what should we think about it?

Well, we should look at it just for how it is. It's a

correlational study we can't say these effects are causal. It's a you know

researchers use the term a bronze standard evaluation not a gold standard

evaluation.

So on the other hand, we have 17 experimental studies

in the United States that use random assignment.

...