SELF-REPORTS OF INTERACTION AMONG 100 GRADUATE STUDENTS IN TWO SECTIONS OF A FIVE-WEEK SUMMER SESSION COURSE IN ELEMENTARY STATISTICS WERE ANALYZED IN AN ATTEMPT TO INCREASE UNDERSTANDING OF SOCIAL ASSOCIATION AMONG HIGHLY SPECIALIZED PERSONS. SUMMARIES WERE COMPILED FOR 1203 INSTANCES OF INTERACTION--408 OF MUTUAL STUDY-HELP, 388 OF HELP GIVEN TO OTHERS, AND 407 OF HELP RECEIVED FROM OTHERS. MAJOR VARIABLES STUDIED INCLUDED FREQUENCY AND EXTENT OF INTERACTION, SALIENCY AND DIFFICULTY OF LEARNING, ANXIETY, AND SELF-ESTIMATE OF COMPETENCE. STATISTICAL ANALYSIS PROVIDED LIMITED SUPPORT TO THE STUDY'S PRIMARY THESIS, THAT INTERACTION AMONG HIGHLY SPECIALIZED PEERS MAY BE EXPLAINED BY SOCIAL EXCHANGE THEORY, WITH THE EXTENT OF INTERACTION DEPENDENT UPON PERCEIVED COSTS AND REWARDS. THIS PAPER WAS PRESENTED AT THE ANNUAL MEETING OF THE AMERICAN EDUCATIONAL RESEARCH ASSOCIATION (NEW YORK, FEBRUARY 17, 1967). (JK)