summarize Two trained raters independently scored each behavioral measure from eight separate viewings of the videotape. On the first viewing, the rater used a stopwatch to score the five duration-based instructional behaviors. On the second viewing, the rater scored the frequency-based instructional behaviors. On the third viewing, the rater scored the student's puzzle-solving performance. On the fourth viewing, the rater scored the five aspects of the student's engagement. After these four viewings, the rater then repeated all four ratings at a later date to double-check the original ratings. One rater scored all dependent measures for all 72 pairs. A second rater independently scored a random sample of one third of the sessions so that interrater reliabilities could be estimated