Arnold S. Tannenbaum

Structural Versus Individual Effects

Methods for separating the effects of group structure or composition from individual effects have been proposed by Blau and by Davis, Spaeth, and Huson. Both methods are useful, but they do not always hold individual or group characteristics strictly constant as intended, thereby making it possible to obtain spurious group or individual effects. Several modifications of these techniques are proposed, making use of more precise matching and correlational techniques. The advantages and limitations of each are discussed. Several additional problems are considered, including the effects of deviants, overlap of distributions among groups, selection processes, and the conceptual definition of structural or compositional effects.

Social psychology of the work organization

Foreword. Preface. 1. The Work Organization 2. The Human Factor 3. The individual in the Organization 4. Personal Adjustment and Conflict in the Work Organization 5. The Group in Organizations 6. Supervision 7. Participation 8. Applications. References. Index.

Some Social Psychological Effects of Direct and Indirect Participation in Ten Swedish Companies

Sweden has introduced schemes of participation widely in its industrial and business organizations. We present in this article data concerning some of the social psychological effects of two forms of participation, indirect and direct, in ten Swedish factories. Indirect participation entails decision making through representatives of the employees; direct participation entails decision making by the employees them selves. Theories propose that participation has the effect of enhancing member reactions such as sense of commitment, favourable attitude and satisfaction in the company and of reducing dysfunctional reactions such as absenteeism. The data from these companies appear consistent with the above theories insofar as direct participation is concerned but not with respect to indirect participation. The data support the view that participation is likely to have the predicted effects on the reactions of members to the extent that participation is felt as a personal experience. Members are likely to feel committed and satisfied, first and foremost to the extent that they perceive themselves personally to have authority to make decisions; second, and in lesser degree, to the extent that they perceive their immediate work group to make decisions, and hardly at all to the extent that they perceive that decisions are made by representatives. These findings are consistent with conclusions drawn earlier by a number of researchers. They are inconsistent with the expectation underlying the laws of participation in Sweden (and in most other countries of Europe) insofar as these laws mandate indirect participation exclusively and are formulated with the expectation that such participation will have the positive psychological outcomes that many advocates of participation claim.

Ownership and Alienation in Kibbutz Factories

As well as hiring workers who are members of the kibbutz and who, therefore, are owners of the means of production, some kibbutz factories have hired workers who are not kibbutz members. Our hypotheses, drawn from the writings of Marx and others, suggest both “individual” and “contextual effects” of ownership on alienation. At the individual level, hired workers will feel more alienated than kibbutz workers. Similarly, at the contextual level, persons in factories where some workers are hired will feel more alienated than will persons in factories that do not include hired workers whether or not the persons are themselves hired workers. These hypotheses imply intervening variables such as influence by workers that are examined through a path analysis. The analysis indicates only individual effects of ownership on feeling of alienation although ownership does have a contextual effect on aspects of the participativeness of the factor.

The Effect on Organization Members of Discrepancy Between Perceived and Preferred Rewards Implicit in Work

Several current theories predict the psychological effect of a discrepancy between the reward persons receive from their job and the reward they would like to receive. Data from 1, 750 persons in 52 industrialplants of five countries are analyzed to explore some conflicting implications of these theories. For example, researchers dispute the assertion that obtaining more than one wants can be as dissatisfying as obtaining less than one wants. The analyses reported here suggest aformula that helps to explain the conditions under which overcompensation may or may not be dissatisfying.

Reactions of Members of Voluntary Groups: A Logarithmic Function of Size of Group

Size is a variable of considerable importance in the study of groups and organizations. Simmel (1955) was among the first to express an interest in the effects of size of groups, and an increasing concern with this variable has since developed. Empirical studies show size to be relevant to such processes as communication and consensus in groups and to such variables as morale, accident rates, and absenteeism in industrial organizations (Borgatta & Bales, 1953; Hare, 1952; Revans, 1959). In a recent study of a large voluntary organization (the League of Women Voters), the size of each of 104 local branches was correlated with a number of other variables measured through the responses to questionnaire items of approximately 25 members in each branch (Tannenbaum, 1961). Size of branch was also correlated with average scores of 29 raters who judged the effectiveness of the local branches. Size of branch in the sample varies from 30 to 3000 members. The product-moment correlation between the effectiveness ratings and branch size is .55 . However, the relationship is dearly nonlinear, and when size is transformed logarithmically the correlation becomes .74, suggesting that the data conform reasonably well to a logarithmic function. [The era correlarlon yields a value of .78 (Donald, 1958).] This led to the computation of correlations between log size and the other variables which had been correlated previously with linear size. A comparison of the magnitudes of these correlations proves interesting. As the following table indicates, the relationships berween the average responses of members in each branch and branch size is likely to be greater when the logarithmic transformation is used.