An exploratory factorial study of behavior interactions in autonomous patient groups

Two instruments, the Group Behavior Questionnaire (GBQ) and the Group Member Evaluation (GME), are the primary focus of this study. The first of these is an 18 item nominations-style sociometric questionnaire; the latter consists of a series of 14 direct rating scales. Within the structure provided by these instruments, 402 consecutively admitted participants in the Patients1 Training Laboratory of the Houston Veterans Administration Hospital responded to the actions of fellow participants in small groups. Ratings were made weekly for four weeks in the case of the GBQ and on a bi-weekly basis with the GME. The primary exploratory procedure used in the study was the derivation of a series of Varimax rotated factors from GBQ and GME item scores. Unit weighted factor scores were generated and related to case history and admissions testing data. These relationships then became the basis for step-wise regression procedures, and the resulting multiple predictors were cross-validated using a random sampling of groups reserved from the gross sample of subject data for this purpose. The results seem to warrant the following tentative conclusions: a. The two group observation instruments proved to be capable of yielding principal factor structures which are highly stable over time. These findings were obtained over test-retest intervals of one, two, and three weeks in the case of the Group Behavior Questionnaire and at two week intervals with the Group Member Evaluation. b. On the basis of item content, the principal factors of the GBQ were named Leadership, Contravention, and Social Impotence, while the two GME factors were labeled Group Oriented Activity and Personal Alienation. In terms of the group member behaviors to be inferred from item content and from appropriate summarizing labels, the Group Behavior Questionnaire analysis shows little agreement with comparable analyses found in the literature. On the same basis, however, the two factor structure provided by the Group Member Evaluation seems a better fit for the ...