THIS investigation was designed to test the hypothesis that there will he a qualitative difTerence between the controlled projection responses of hoys who, on the basis of their sociometric choice status, are classified as popular, middle and unpopular in their social relations with their classmates.
METHOD
Procedure for determining sociometric status ' •Sociometric tests were administered to 11-yr-old boys in 12 schools in the same area of Manchester. A total of 420 boys participated. They were asked to choose three friends, from amongst their classmates, to each of the following three criteria:(a) Someone you will like to sit next to, (b) Someone you will like to play with, (c) Someone you will like to visit at home. Eight weeks later pupils were given the same sociometric test again. A test-retest reliability coefficient of 0-94 was obtained. This compares favourably with a reliability coefficient of 0-96 obtained after 4 day (Jennings, 1943), and that of 0-93 to 0-94 obtained after 2 day interval (Zeleny, 1939).
Determining popular and unpopular childrenBronfenbrenner (1945) suggested a useful classification of choices received on sociometric tests according to their probability of chance occurrence. This classification was demonstrated to hold provided the number of subjects (in each form) is not more than 50 and the number of criteria is between 2 and 5. Thus scores above 15 are said to be significantly above chance, indicating popularity, a score of 9 is exactly on chance and scores below 3 are regarded as indicating unpopularity. In the present study, the largest of the 12 forms had 35 pupils and the number of criteria was 3. It was considered therefore acceptable to use Bronfenbrenner's classification to determine popular and unpopular children as well as the middle group of children, all of whom then completed the Raven's Controlled Projection Test.
Obtaining controlled projection responsesAs Raven (1951) puts it, "the aim of Controlled Projection is to record for comparative study, the meaning and significance a child attaches to situations of