THIS COMPARATIVE ANALYSIS OF PRODUCTIVE THINKING IN RETARDED AND NONRETARDED CHILDREN USED DATA FROM A SAMPLE OF 78 PUBLIC SCHOOL CHILDREN DIVIDED INTO THREE GROUPS--(1) MENTALLY RETARDED CHILDREN IN SPECIAL CLASSES, (2) RETARDED CHILDREN IN REGULAR CLASSES, AND (3) NONRETARDED CHILDREN IN REGULAR CLASSES. THE GROUPS WERE SELECTED SO THAT THERE WAS NO SIGNIFICANT DIFFERENCE WITHIN THE GROUP BETWEEN THEIR MENTAL AGES AS MEASURED BY THE CALIFORNIA TEST OF MENTAL MATURITY. FIRST A BATTERY OF PRODUCTIVE THINKING TESTS WAS ADMINISTERED TO EACH SUBJECT CONSISTING OF--(1) VERBAL STIMULI REQUIRING VERBAL RESPONSES, (2) NONVERBAL STIMULI REQUIRING VERBAL RESPONSES, AND (3) NONVERBAL STIMULI REQUIRING NONVERBAL RESPONSES. THE SECOND PART OF THE STUDY WAS AN ANALYSIS OF THE THOUGHT PROCESSES CONTAINED IN VERBAL INTERACTION IN THE CLASSROOM. DATA WERE GATHERED VIA RECORDINGS IN CLASSROOMS DURING ARITHMETIC, SCIENCE, AND SOCIAL STUDIES LESSONS. THOUGHT PROCESSES WERE THEN CLASSIFIED INTO SUCH CATEGORIES AS ROUTINE, COGNITIVE-MEMORY, CONVERGENT THINKING, EVALUATIVE THINKING, AND DIVERGENT THINKING. THE RESULTS WERE THAT COGNITIVE MEMORY ACCOUNTS FOR APPROXIMATELY ONE-HALF OF THOUGHT PROCESSES CONTAINED IN VERBAL INTERACTION. IRRESPECTIVE OF THE PRESENCE OF MENTAL RETARDATION, WHEN COMBINED WITH ROUTINE, NEARLY 80 PERCENT OF THE THOUGHT PROCESSES IN CLASSROOM VERBAL INTERACTION WERE ACCOUNTED FOR. OCCURRENCE OF EVALUATIVE THINKING AND DIVERGENT THINKING WAS INFREQUENT IN THE OBSERVED DATA. (GC)