Why Do We Bother?

For the past two days most of my at-computer time has been spent using my amateur knowledge of statistics to see if the one-credit-hour "helper" course that we offer as an add-on to College Algebra and Precalc actually helps at all.

A higher percentage of students who take the helper coures are successful than of students who don't take the helper course. But the difference is not statistically significant.

However, the students who take the helper course have a lower mean placement test score than the students who don't take the helper course. This difference is statistically significant.

The question that I can't answer, though, is whether this difference in placement test score is large enough that you would expect the students with the lower score to be less successful so the fact that they are just as successful means the course helped. Or does the course have no meaningful effect?

And while we're at it, do the students who pass these courses do any better at calculus than the students who slip through the cracks, skipping College Algebra and/or precalc, and going straight to calculus?