Comparison of Data Obtained From Mothers and Fathers on Childrearing Practices and Their Relation to Child Aggression1

In reviewing the literature having to do with behavior of children as it is related to behavior and attitudes of parents, one is struck by two methodological biases which do not permit a clear-cut evaluation of the substantive findings of a majority of these studies. The first has to do with the disregard of the father both as an important socializing influence and as an informant. A search of the literature between 1929 and 1956 revealed 16o publications dealing with mother-child relationships but only ii with father-child relationships (io). Since that time there have been a few more studies which have considered the importance of father-centered variables, e.g., Bronfenbrenner (2), Miller and Swanson (9), Sears, Maccoby, and Levin (11i), Sears, Pintler, and Sears (12), but information about fathers in these studies usually has been obtained second-hand, either from the mother or from the child himself. This introduces the second methodological bias which is the contamination that is likely when assessments of both parental and child behavior (presumed antecedent and consequent) are made by the same individual. A study in identification by Levin and Sears (8), as well as the earlier Sears' studies (12, I3), are not subject to this error, although they are to the first. When both the predictor and criterion are obtained from the same subject, then response set cannot be eliminated from the possible interpretations of the results. In addition to contributing important information which has usually been lacking in childrearing studies, the use of both mother and father as informants could serve as one check on