Evidence that conceptual physics leads to strong understanding of college physics

This is the first year I've taught AP physics to students who have been through our recently revised conceptual physics program.  The results after a marking period are illuminating -- those who took our conceptual physics are definitively outperforming those who have taken a traditional calculational-style previous physics course, whether that course was here or at another school.

This result is independent of teacher -- the students that my colleague Alex taught in 9th grade are doing just as well as the students I taught.  And our former conceptual students aren't merely remembering content, either, as we didn't even discuss much of this AP material three years ago.  Furthermore, this correlation of top performers to veterans of 9th grade physics did not exist in the days when we taught calculational physics to 9th graders.

No, the difference is the discipline we instill of answering every physics question with a fact from our sheet, or with a carefully-taught methodology to make a calculation.  The top half of the class doesn't have to learn how to learn physics.  

My conclusion: a conceptual approach to first-year physics is better than a calculational approach, even for those who go on to study physics at the university level.  I take this as an article of faith, but many physics teachers, parents, administrators, and university folk would not agree.  Certainly more and more are on coming on board with a conceptual approach.  But lots of people -- professionals, even -- would say how we don't teach our students the math skills, or that we deny our students an understanding of "real" physics, or that we're too fluffy, or that we don't prepare our students for college.  These concerns aren't unique to 9th grade conceptual physics: they're some of the same critiques I've heard of the AP Physics 1 program.

The naysayers are wrong.  The evidence is clear to me.  Observationally, those who had conceptual physics behave in a more confident manner, are more comfortable discussing and explaining physics, and can adapt far more quickly to new and challenging physics concepts.

What about some non-observational evidence?*

* Obviously not scientific evidence, just evidence.  Teaching is a craft, not a science.  Others may have different evidence.  If so, please describe it in the comments.

I ranked the 20 students in my AP Physics 1 section by grade.  Numbers 1-8, 11, and 13 took 9th grade conceptual physics.

Numbers 9, 12, and 14-20 took a previous calculational physics class: two took our Regents-style calculational physics as juniors, and the rest had calculational physics at another school.  In all cases that I've asked about, that previous calculational class covered more topics than our conceptual class does.  Teaching physics isn't about topics covered, it's about the style of approach to the subject.) 

They've all seen physics before.  But the ones who have seen rigorous conceptual physics are doing best.