Suchergebnisse

Replik auf Jack Halperns Artikel über die Polarisierung in Rhodesien (The World Today, 27 (1971) 1. S. 1-8). (DÜI-Ker)

A nat'l survey of 1,200 adults & another 6,000 adults in 6 central cities in the North were used to identify pop subgroups that differed greatly from each other on att's toward race relations utilizing a technique known as Polarized Subgroup Analysis (PSA). Indices of race relations, the pace & effort of integration, causes of riots, & ways to prevent riots were developed based on inter-item r's. The most important independent variables were selected based on stepwise multiple regression results. These included the amount of the R's past pol'al behavior, his educ, race, sex, age, region of the country, & the city's past riot history. Independent variables were then divided into categories & subgroups were defined by the combination of several variables. Although whites & nonwhites were in approximate agreement on some general statements about integration, specific programs, particularly in housing, lead to widely diff viewpoints even when multiple attributes were used to define racial subgroups. On indices based on combining several items, neither survey found white subgroups whose average scores overlapped in a meaningful way with those of nonwhite subgroups. Further, while younger & better-educated whites were closer to blacks than other white subgroups, younger & better-educated blacks tended to be further from whites than other black subgroups. AA.

Two experiments examined the group polarization hypothesis. In Experiment I group discussion polarized the evaluations of six hypothetical faculty members, three described positively and three negatively. 'Good' faculty were rated and paid even more favorably after group interaction and contrariwise for 'bad' faculty. Experiment II separated subjects into groups which were conservative or liberal in attitudes regarding women. Subsequent discussion of statements regarding the role of women yielded an increase in the attitude gap between the conservative and liberal communities.

A periodic decision process, analogous to a repeated election, in a student-living group was studied to test the polarization model proposed by Berelson, Lazarsfeld & McPhee in VOTING (See SA 3242). Sociometric & attitude questions were askec in 3 panel waves: at the height of controversy as to whether the group should have a `beer bust'; 2 months later, in a period of calm; & 4 months after the initial wave, when a 2nd collective decision was taken. Results for reciprocated choices supported the polarization model, but those for unreciprocated choices were in the reverse direction. A steady trend toward attitude homogeneity in choices was also superimposed on the pulsating group polarization. Further studies of 'middle-sized' groups are recommended to illuminate the processes of attitude change & structural change in groups undergoing internal conflict. IPSA.