Teacher Training Puzzle

Today's puzzle, which uses a sample of over 500 rookie teachers: Which rookie teachers end up doing better* at the end of the school year?

*Better as measured by all possible ways -- observers, student survey. 

1) Those with high "selection scores."

Obviously all 500 were ultimately selected for the job out of thousands of applicants. But some of the 500 seemed really impressive during interviews, had higher GPAs, better writing samples -- those are the ones with higher "selection scores."

Or

2) Those with high practice scores.

During summer training, all 500 teachers had to practice techniques, like Strong Voice. They practiced with other rookie teachers. Each time they did a practice, if even for a few minutes, a judge gave them a score. This is least like real teaching.

Or

3) Those with high ratings on summer "Sample" teaching. Unlike #2, here each novice is teaching a real lesson, in front of actual students. Again, a judge scored each effort.

*

You are searching for correlation.

Which one thing -- the screening process, a short practice on a single teaching move, or a sample teaching lesson in front of real kids -- best predicts who is actually good a year later...that is, after a full year doing real life teaching?

Or maybe two things correlate highly. Or maybe none. What do you think?

Decide and then...

...The inestimable Doug Lemov has the answer here.

Congrats to our friends in that teacher development organization for taking on such thoughtful research!