Review of Some New Techniques for Studying Social Behavior by Dorothy Swaine Thomas
Kimball Young
Some New Techniques for Studying Social Behavior. By DOROTHY SWАINE THOMAS and ASSOCIATES. New York: Bureau of Publications, Teachers College, Columbia University, 1929. Pp. x+203. $1.50.
In the statistical method of sociology the three most important matters are: decision as to "what are the critical things to examine"; the determination of "our units of measurements"; and the invention of "means for controlling our instrument of measurement (i.e., the observer)." In this series of papers all three matters are given attention with special consideration to the last problem. Miss Thomas writes, "We feel that it is more important in this field to control the observer than to control the experiment." The work was carried on at the Child Development Institute at Teachers College.
Miss Barker's report on the social-material activities of children indicates first of all the difficulty in getting agreement among observers unless the complex activities are broken up into smaller parts than can be conveniently handled. This is especially true of longer activities of children. The reliability coefficients which here ranged from 0.47 to 0.80 reveal just this difficulty. Some of the facts about the behavior of the children are noted: "The higher the mental age, the larger the number of social contacts a child is apt to have." Further, "The greater the number of activities per unit of time, the greater the space covered by the child."
In this chapter there are some suggestive classifications of child activities, although the reviewer does not quite see how the distinction between social and non-social is brought about. A good many of the contacts described under the latter rubric are certainly socially and culturally conditional. Shall "social" be used only for the overt contact of child to child, and not in regard to what might be called self-activities? What has become of the influence of James, Baldwin, Cooley, and Mead when statistical students of social behavior ignore the fact that even so-called self-activities are also social in the sense of early conditioning. It will be a sad day for sociology when the influence of behavioristic stimulus-response psychology constricts the use of the term "social" to mean only overt person-to-person stimulation.
(298)
Chapter ννν by Miss Loomis reports the attempt to measure the physical contacts made by nursery-school children. The principal difficulty here was not noting of number of contacts within a unit of time, but the determination of comparable categories, that is, determination of the kind of contact to record—the behavior units involved in such terms as "hit," "pull," "push," "caress," "accident," "assistance," etc. Moreover, as Miss Loomis points out, the varied nursery-school situations determine certain directions of contact, and not all of these can be controlled for careful observation. While for the most part the number of contacts made to others balanced the number received, there were some interesting divergences indicating possibly some criteria for differentiating between introvertive and extravertive trends in the behavior of some children.
A short analysis is made of spontaneous group formation. Participation in group activities in terms of time was fairly easily measured, but the reliability was measured by the rank order method which is not so satisfactory as the Pearson r. The report indicates, moreover, that under the set-up so much time was consumed in recording names of children and the time that the social activity itself often had to be neglected. The authors suggest that perhaps a qualitative method will have to be used for this aspect of the investigation.
A brief but purely tentative study of laughter situations, as indicative of social responsiveness, is given. The criteria of social responsiveness was "the number of laughs for each individual as compared with the number of times he was in a group where laughter occurred," but there was great variability among three different groups studied, apparently based on the fact that "there was undoubtedly greater reliability in recording the occurrence of a situation than in recording a description of its function." The statistical reliability of the observers in this study is not known.
Chapter vi describes "ganging" in a preschool group. The reviewer would say that the characteristics of the behavior follow more those of a congeniality group growing out of the nursery-school playtime than a gang in the usual sociological sense. There is no quantitative analysis whatsoever but an interesting picture of personality differences and of social interaction among three boys between three and four years of age.
The last five chapters of this book deal with psychological test situations as indices of social behavior. Mrs. Nelson reports an analysis of resistances set up in the children during the testing and shows that age, maturity of mentality, and the nature of the test all produce differences in the reactions to the tests. Praise and blame are shown to alter considerably the responses to the tests. Here we have additional proof of what students
(299) of social behavior have been saying for years, viz., that intelligence tests reveal, in their scores, emotional and social responses quite as well as "purely" intellectual ones. The tone of voice, the facial gestures, the manner toward the child may make much difference in his reaction. Too frequently standardized tests assume that the child operates on an intellectual level without regard to emotional and social factors.
Miss Wise's analysis of the children's statements of "I can't" and "I don't know" brought out the fact that often these statements show genuine appreciation of difficulties or problems beyond the child's capacity. At other times, it means merely a verbal response to orders which the child first meets in this negative manner without there being genuine resistance there. In still other instances they show distinct negativism.
Chapter ix gives a complete stenographic report of one test and extracts from others. No one can read this material without realizing how much place social interaction has in the success or failure of the child in mastering the tests. The final chapter is a brief discussion of the reliability of stenographic as against student reports on the test situation. The former are more complete, accurate, and reliable, especially when the stenographer is given some brief preliminary training in the procedure and significance of testing.
On the whole this book constitutes not so much contributions to our knowledge of child behavior as methodological suggestions for child study. The reviewer cannot agree with Miss Thomas when she says that "our ultimate aim is, of course, statistical." Is it not rather sociological? That is to say, the matter here is one of statistical or other methodology, but after all our aim is a clearer understanding of social behavior no matter whether by one method or another.
KIMBALL YOUNG
UNIVERSITY OF WISCONSIN
