This study described the relationship of 38 variates to achievement in high school biology and chemistry classes. Forty-four teachers prepared audio tape recordings of discussions throughout a nine-month period. Equal-sized groups received treatment consisting of wait time feedback and/or supportive intervention, a form of peer coaching. Other variables were initial measures, which included the variates of class size and previous science grade, pre- and posttest scores on student attitudes and perceptions, and prediscourse and discourse analyses, which included the variates of wait times, actions, and response durations. While previous research has shown that wait time feedback and supportive intervention are effective means of changing teacher behavior, the results of the present study revealed that previous student achievement and the attitudes of students accounted for 70% of the variation in the final examination scores (New York State Regents) on the basis of only the first three weeks of data collection on the variables assessed. Generally, little emphasis on higher-level thinking, wait time, or problem solving was found at the beginning of the school year, and only minimal changes resulted from the application of the treatment variables. Pressures for content coverage and preparation for externally imposed statewide examinations that emphasize memory-level learning augured against the changes that the independent variables were designed to produce.