Selecting people randomly

This article considers what justification can be found for selecting randomly and in what circumstances it applies, including that of selecting patients to be treated by a scarce medical procedure. The author demonstrates that balancing the merits of fairness, common good, equal rights, and equal chance as they apply in various situations frequently leads to the conclusion that random selection may not be the most appropriate mode of selection. Broome acknowledges that, in the end, we may be forced to conclude that the only merit of random selection is the political one of guarding against partiality and oppression.     

SEX DIFFERENCES IN THE ATTRIBUTION OF SOCIALLY DESIRABLE AND UNDESIRABLE TRAITS TO OUT-GROUP MEMBERS*

One hundred and forty-four male and female Australian subjects rated an Australian or a Vietnamese target group on the possession of five socially desirable and five socially undesirable personality traits. As predicted, the out-group was perceived less favourably than the in-group, but only by the males, the females not making the distinction.     

Emerging independence in nursing graduates: An analysis of traditional-nontraditional value patterns

The study sought to determine whether current nursing graduates value assertiveness, independence, and achievement to a greater degree than had past graduates, and whether selection or socialization might account for such a change. Scores on the Allport Vernon Lindsey Scale of Values for 55 1974 graduates of a private baccalaureate school of nursing were compared with 66 1980 graduates. Only scores on the political scale were significantly higher for the 1980 class. 1980 graduates displayed a traditional value pattern of high social and low political scores upon entering the educational program. Longitudinal comparison of scores for the 1980 class showed a significant rise in scores on the political scale; graduating scores did not differ significantly from college norms. The results suggest a rise in nontraditional values which is the result of the educational socialization process.     

Sex-role orientation and depressive symptomatology

Sex Roles - The incidence of depressive disorders among women in our society is considerably higher than it is for men. Several investigators have suggested that this is due to traditional sex...     

Variables related to reported learning in brief experiential groups held at professional meetings

Reports of learning from 119 participants of 18 general psychodynamic groups from the 1981 and 1982 annual meetings of the Canadian Group Psychotherapy Association were investigated. Two main learning factors emerged from a factor analysis of participant ratings, learning about oneself in a small group (factor one) and learning how to be an effective leader with professional skills (factor two). Participants reported moderate to considerable learning for both factors even though the stated task of the groups did not emphasize the learning of professional skills. Relationships between several classes of variables (group, leader, demographic) and reported learning were studied. Results that were consistent with several previous investigations were found. The discovery of a participant, sex-difference finding, i.e., greater reported learning by the women, generated a number of questions that may be addressed in future investigations.This article was derived from a paper presented at the Fourth Annual Meeting of the Canadian Group Psychotherapy Association, October 27, 1983, Banff, Canada.     

Scaling of output regularities in free recall: A review

Problems involved in scaling the output regularities observed in free recall are outlined, and assumptions underlying attempted solutions for both categorized and “unrelated” lists are discussed. Several studies that make claims for the superiority of particular measures have indeed demonstrated that it matters which index is selected, but they are shown to provide no justification for preferring one to another. Evidence from simulation studies, however, suggests that a number of measures have “built-in” correlations with theoretically relevant variables such as recall. Some of these biases have proved elusive because their appearance is confined to nonrandom output sequences. The relation between theoretical assumptions and simulation studies is discussed.