The goals of the study were: (1) to determine the relationships among selected sociological, health, and behavioral variables and third-grade word analysis test scores, and (2) to test a causal model employing part analysis. Subjects were 314 Head Start students in 1968. Data on race, sex, perinatal complications, number of children in the family, father absence, and subjects' behavior were obtained from medical records. Data indicate that for the most disadvantaged children (those selected for Title I programs) the absence of the father may be a possible cause of low word analysis test scores. Data tables are included. (CS)