With many massive instructional interventions not even registering a blip on student performance measures, there's a new study write-up in Science indicating that just 20 hours of in-service training in which teachers essentially watch a video of themselves teaching was able to change teacher practice sufficiently to significantly impact student achievement one year later.

Teachers in the study sent video recordings of their own classroom instruction to specially trained coaches, who then analyzed their interaction with students, focusing largely on instructional support, emotional support and classroom organization. Then twice a month, each expert posted snippets of video illustrating areas for growth or positive interactions to a special secure website for the teacher to review. After the teacher reviewed the video, both expert and teacher would talk to strategize about what specific actions the teacher could take next.

The improved performance of students of teachers who received the intervention training was about a fourth of a standard deviation, on the order of the "big gun" interventions, such as using a more effective curriculum or preschool programs. The results are all the more amazing in that it was gold standard research, using a randomized experimental design with a control group.

Because the experiment was conducted on a small group of relatively experienced Virginia teachers, we wonder if some experience is a prerequisite for the effects. If not, could this model be borrowed for use in the student teaching phase of traditional teacher preparation, a phase that our recent report on student teaching indicated is in critical need of improvement?