The Journal of Applied Behavioral Science

This paper analyzes the major political problems evaluation research must solve if it is to be effective in an innovative agency. Peace Corps annals supply the illustrations. In solving these problems, the role of the "inside" researcher is detailed. He is both consultative (promotes collaboration between researcher and line people who will use the results) and entrepreneurial (integrative of diverse resources); he must bridge the cross-cultural gaps within the agency and with respect to the client the agency serves, the author explains. In the next section he deals with the role of the "outside" researcher and the relative merits and demerits of contracts and grants for evaluation. There are also political problems inherent in the organization's response to evaluation, especially where "unacceptable findings" and "unintended consequences" result. Finally, the author stresses the constructive impact careful research can have on an innovative agency's program and mission.

Two concepts borrowed from Everett Rogers' Communication of Innovations-heterophily and the role of change agents-are utilized to explain problems which welfare workers have in being influential family planning change agents for welfare mothers. These concepts are viewed as embedded in the social context of welfare department attempts to influence welfare mothers to use family planning and the related criticisms which welfare rights protagonists have recently voiced. Structured interviews with 58 welfare workqrs and 300 welfare mothers in one metropolitan Tennessee county reveal that heterophily characterizes these two groups in social and demographic characteristics, birth control attitudes, family size norms, and abortion attitudes. It is further suggested that this heterophily may partially account for welfare workers' limited conformity to the change agent role in family planning counseling with their clients. A number of innovative solutions are proposed to deal constructively with public welfare problems in family planning, some of which have shown promise in other health programs here and a broad.

The data obtained imply the usefulness of taking "social desirability" (SD) differences into account in interpreting any effects of OD interventions measured by self-reports. Broadly, SD differences can mask measurements of such effects, especially in before vs. after designs. High SD scorers seem to provide preintervention responses that reflect the socially acceptable rather than what exists, and these responses consequently constitute an unreliable benchmark against which to compare postinter-vention responses. The Crowne-Marlowe SD scale is used to assess SD differences between individuals.

This report deals with a measurement issue, and specifically in relation to Likert's Profile of Organizational Characteristics. The results apply strictly to uses of the Profile, which is a prominent measuring instrument in OD work. However presumptively, the present findings are also meant to "cast their shadow" on interpretations of various other self-reports used to measure OD effects.

The specific purpose of this experiment was to test the effectiveness of the Data Survey and Feedback method of Organization Development. A pre-post experimental design was utilized with an experimental group which received the Organization Development treatment, and a control group which did not receive the treatment. Both individual satisfaction and team performance in a complex management simulation game were measured prior to and subsequent to the treatment. Although no differences were found between the control group performance and the experimental group performance, the experimental group was found to be more satisfied than the control group, subsequent to the treatment.

An organization development effort spanning several years in one organization is reexamined in order to assess the differential effects of interventions. The findings illustrate how organizational change efforts can have different effects upon different groups within an organization. Significant differences among hierarchical levels of the organization are found in satisfaction with several aspects of the work environment, the most important finding being the relatively low levels of satisfaction among supervisory and technical workers. Supporting clinical data identify the lack of an integrated change program, the failure to adequately structure supervisors into change activities, and problems in the management of participation as possible sources of the differential effects. Recommendations for the design of interventions are made, and the conclusion is reached that there is a need for the application of more systemic models of change combined with more comprehensive measurement of the effects of interventions.

Sixty managers attending an executive course responded to questionnaires specifying their perceptions of the needs for improvement of themselves and of their bosses. Half of the group, randomly assigned, answered before participating in a 3-day T-Group program and the other half, immediately after. When the responses were compared, it was seen that the perception gap between perceived "self" and "boss" improvement needs in the interpersonal area was wider after training than before. From this finding and an analysis of the literature, it is argued that T-Group programs tend to have a negative influence on perceptions of outsiders and that this effect may be functional as well as dysfunctional in Organization Development.