FURTHER ANALYSES OF DATA COLLECTED IN AN EARLIER STUDY (ED 002 992) WERE CONDUCTED TO DETERMINE EFFECTS OF SELECTED GROUP COUNSELING FACTORS INCLUDING (1) RELATIONSHIPS BETWEEN CLIENT GROWTH AND TOPICS DISCUSSED IN GROUP COUNSELING SESSIONS, (2) EFFECTS OF GROUP COMPOSITION ON COUNSELING GROUP INTERPERSONAL BEHAVIORS, AND (3) WAYS OF MEASURING CHANGES IN CLIENTS' BEHAVIOR. INTERACTIONS AMONG NINTH GRADERS AND PROSPECTIVE COUNSELORS WERE CLASSIFIED IN TERMS OF EFFECT, REFERENT, AND TOPIC. FOUR PSYCHOLOGISTS (NOT PREVIOUSLY CONNECTED WITH THE STUDY) ARRANGED PROTOCOLS OBTAINED FROM A PICTURE-STORY TEST IN ORDER FROM PRETEST TO FOLLOWUP TESTS FOR THE NINTH GRADERS AND THEIR PROSPECTIVE COUNSELORS. THEN THE EFFECT OF CHANGE ON PROTOCOL ARRANGEMENT WAS DETERMINED. INTERPERSONAL BEHAVIORS OF CLIENTS AND PROSPECTIVE COUNSELORS WERE THEN GROUPED UNDER THE CATEGORIES OF APPROACH, ATTACK, AND WITHDRAWAL. ANALYSIS OF VARIANCE, CHI SQUARE, AND RANK ORDER CORRELATIONS WERE THEN PERFORMED. MAJOR OBJECTIVES OF THE STUDY WERE NOT ACHIEVED. CUES USED TO ARRANGE PROTOCOLS, HOWEVER, SUGGESTED MEANINGFUL QUESTIONS FOR SUBSEQUENT ANALYSIS. (WN)