Thirty-three student teachers were observed in elementary school classrooms to determine if they manifested 16 interactive behaviors identified as desirable by college of education faculty. Teaching assistants used the Georgia Assessment of Teaching Effectiveness (GATE), an instrument which requires the observers to objectively record, but not judge, classroom events occuring during a five-minute period. Test developers selected items from the GATE which reflected each of the 16 identified behaviors. An analysis of variance permitted investigation of the reliability of GATE clusters of items which were descriptors of the identified behaviors. Findings were based upon statistical analysis of individual and group profiles generated by computer across each of the 16 behavior statements. Reliability estimates showed that GATE paralleled 14 of the 16 behaviors, thus implying that this instrument can be used to validate the classroom interactive portion of a student teaching program. The 16 GATE items and 13 behaviors not measured by the GATE are included. (CP)