To study the teaching-learning process in classrooms, a small sample of carefully selected teachers at the second- and fifth-grade levels were identified for one phase of the Beginning Teacher Evaluation Study. A unique characteristic of this sample was that it included twenty teachers identified as relatively more effective in influencing achievement gains in their classes and twenty teachers identified as relatively less effective in their ability to teach certain objectives. Procedures and data related to the identification of these two disparate groups of teachers from a larger sample are described in this report. The teachers in the larger sample taught two-week, experimental units in reading and mathematics, and students completed pre- and post-tests. The smaller sample was selected based on information obtained from the student tests and classroom observation. (Authors)