Evaluation of Head Start programs is approached through identification of interactions with families and communities to learn what effects the Head Start experiences have had on the children involved. As a result of earlier studies, measures were developed for social-emotional and family factors, as well as program characteristics. In 1966-67, the newly-established Evaluation and Research (E&R) Centers carried out the first full-year Head Start evaluation, using class samples representing extremes on important dimensions. In addition, for greater depth of measurement of cognition, social-emotional behavior, program, and family, the E&R Centers "clustered" to collect data for the 1967-68 study. To more clearly identify the factors relevant to child development, the 1968-69 evaluation used a different approach, that of "making things happen," in contrast to previous reliance on natural variation. Common core data were collected on at least 120 children 3 to 5 years of age, attending Head Start classes for the first time. The tests used to collect common core data, cluster data, and follow-up data for the 1967-68 and 1068-69 studies are identified and described in this paper. Also individually examined are evaluation studies by nine E&R Centers, designed to identify relevant dimensions of Head Start programs and to develop adequate measures of these. Data will be provided on what optimal child development programs can be, in contrast to other 1968-69 Head Start evaluation studies. (Author/NH)