ABSTRACT
D Uring the four terms of the experiment all but one of the sociometric tests given to the experimental group were given also to the control group. It was hoped that by this means it would be possible not only to study the social development of the experimental group, but also to compare it with that of the control group over a period of time. In January 1948, when the experiment was in its third term, an opportunity arose of comparing the children's choices of work companions with their choices of companions in the school canteen at lunch time. Both groups were given a sociometric test on this new criterion in January; places at canteen tables were allocated on the basis of this test and a retest was given in April. The results of these tests will be discussed and compared with the other tests in the course of this chapter. An attempt was also made to relate the children's choices (on the work criterion) to such factors as home districts, previous schools and special friendships. It was possible to gauge to some extent which of the children were most stable in their friendship and which were able to establish the most satisfactory social relations with others in their form. Certain characteristics of the groups as social units also came to light as time went on. These would not have been so clear had the sociometric tests been limited to one criterion.
