COMPLETE EVALUATION DATA WERE GATHERED ON 162 CHILDREN IN 23 CLASSES IN NINE HEAD START CENTERS. FIVE AREAS EXPLORED WERE--(1) ITEM ANALYSIS OF EVALUATION TESTS--STANFORD-BINET AND THE CALDWELL-SOULE PRESCHOOL INVENTORY, (2) COMPARISON OF AGE-MATCHED GROUPS--43 MATCHED PAIRS WITHIN THE HEAD START POPULATION (THERE WERE NO SIGNIFICANT DIFFERENCES ON THE ZIGLER BEHAVIOR INVENTORY EXCEPT THAT THE CONTROL GROUP RECEIVED MORE FAVORABLE RATINGS IN THE AREA OF JEALOUSY), (3) FOLLOWUP OF CHILDREN FORMERLY ENROLLED IN A PREKINDERGARTEN PROGRAM--20 MATCHED PAIRS OF SECOND GRADERS (DATA COLLECTED FROM SCHOOL RECORDS OF ACHIEVEMENT IN READING, WRITING, AND ARITHMETIC SHOWED NO SIGNIFICANT DIFFERENCES BETWEEN CONTROL CHILDREN AND CHILDREN WHO HAD ATTENDED PREKINDERGARTEN), (4) CHILDREN'S CONCEPTUAL DEVELOPMENT AND LANGUAGE COMPREHENSION (RESULTS INDICATED THAT WHEN VERBAL INSTRUCTIONS ARE TO BE FOLLOWED, SYNTAX, TEMPORAL ORDER OF ELEMENTS, AND SEMANTICS GUIDE CHILDREN'S ACTIONS), AND (5) EXPLORATORY WORK ON BLOCKBUILDING AS A COGNITIVE INDICATOR (PROTOCOLS WERE COLLECTED ON THE MANIPULATIVE AND VERBAL BEHAVIOR OF 100 CHILDREN). SEVEN TABLES RELEVANT TO THE ABOVE FIVE AREAS ARE INCLUDED IN THE REPORT, ALTHOUGH PORTIONS OF THE REPORT ARE NOT AVAILABLE FROM EDRS. (MS)