Psychological Bulletin

Mehrabian, Albert; Ksionzky, Sheldon

doi: 10.1037/h0029603pmid: N/A

Reviews and integrates findings and measures available in the area of affiliation and the related field of conformity. Affiliative characteristics are represented in terms of a 2-dimensional scheme, based on generalized expectation of the positively and negatively reinforcing quality of interpersonal relationships. Treating each of these dimensions dichotomously, 4 categories of affiliator are obtained. These 4 categories are related to existing measures of affiliative tendency. Implications of available research are used to elaborate a model in which affiliative behavior, the dependent variable, is a function of affiliative characteristics, and attributes of the target of affiliation or the situation in which affiliation occurs. Conformity behavior, in a related model, is then expressed as a function of affiliative characteristics and target attributes. (2 p. ref.)

Bailey, Kent G.; Sowder, W. Thomas

doi: 10.1037/h0029633pmid: 5485506

Reviews the research literature on self-confrontation in psychotherapy with emphasis on audio- and videotape playback techniques. Playback is lauded by most writers and numerous benefits are ascribed, but experimental support is lacking. The few controlled studies extant are criticized, and suggestions are offered for designing and interpreting future experiments in the area. Self-confrontation is presented as a rapidly developing approach which deserves a thorough investigation of its practical and heuristic possibilities. (64 ref.)

D'Agostino, Ralph B.

doi: 10.1037/h0029499pmid: N/A

R. C. Geary's test of normality, based solely on the ratio of the sample mean deviation to the sample standard deviation is suggested to the research psychologist. This test is computationally easy and has good sensitivity to nonnormality. To further facilitate the test and to broaden its applicability a normal approximation to the test is given and investigated.

Rabinowitz, F. Michael

doi: 10.1037/h0029551pmid: N/A

Briefly reviews the literature relevant to the response-response sequential dependencies exhibited by humans in multiple-choice situations (k > 2). It is concluded that human behavior is characterized by 3 types of sequential dependencies that appear in a wide variety of experimental situations. These include the tendency to: (a) not repeat responses, (b) respond to adjacent loci on successive trials (respond in a series), and (c) use all k possible responses on k successive trials. It is noted that the tendency to respond in a series decreases as a function of increasing age and increasing interresponse interval, while the tendency not to repeat responses decreases as a function of increasing interresponse interval. It is also concluded that characteristic sequential dependencies are not usually the result of either attempts to optimize stimulus input or minimize reactive inhibition. (30 ref.)

Humphreys, Lloyd G.

doi: 10.1037/h0029500pmid: N/A

It has been shown that the resemblance coefficient between related pairs, e.g., twins, on a composite measure is a function of 3 kinds of correlations involving the component measures: within-person (homogeneity), cross-pair-same-measures, and cross-pair-different-measures correlations. Present data suggest that there is substantial variation among these from 1 related group to another. As a result, the size of the resemblance coefficient is an arbitrary function of the components used by the test constructer. The research person need not accept current tests as given, but instead can control the weights assigned to components in accordance with a rational plan of analysis.