Lee Sechrest

Incremental Validity : a Recommendation

distinction is to be made between convergent and discriminant validity. It is necessary to demonstrate not only that a measure covaries with certain other connotatively similar variables, but that its covariance with other connotatively dissimilar variables is limited. Campbell (1960) has suggested several possible additions to recommended validity indicators, all of which focus on the problem of discriminant validity, i.e., the demonstration that a test construct is not completely or even largely redundant with other better established or more parsimonious constructs. He has suggested, for example, that correlations with intelligence, social desirability and self-ratings should be reported since these variables are likely to be

Deviant Response Tendencies: Their Measurement and Interpretation

IN recent years a wide variety of systematic biases in the response tendencies of individuals have been identified. Few of these, however, are of greater potential theoretical or practical importance than the one described by Berg (1955), accounted for in terms of his &dquo;deviation hypothesis,&dquo; and further elaborated in a number of research studies (Adams & Berg, 1961; Barnes, 1955; Berg, 1957,1959, 1961 ; Berg & Collier, 1953; Hesterly & Berg, 1958; Roitzsch & Berg, 1959). The initial statement of the deviation hypothesis was: &dquo;Deviant response patterns tend to be general; hence those deviant behavior patterns which are significant for abnormality and thus regarded as symptoms are associated with other deviant response patterns which are in noncritical areas of behavior and which are not regarded as symptoms of personality aberration&dquo; (Berg, 1955, p. 62). It is our purpose to attempt a clarification and explication of the measurement and interpretation of deviant response tendencies. We have gained the impression from a study of published reports and from our own research on deviant response tendencies that certain

Graffiti in four types of institutions of higher education

Of gTeat interest are those responses which are freely emitted without apparent regard for the possibility that they may ultimately come under the purview of science (Webb et al., 1966). They have the advantage of being less subject to initial censoring tendencies and of providing opportunities for repeated measurement without the measurement operations affecting the responses. Some years ago Kinsey and his associates (Kinsey et al., 1953) made some initial observations and quantification of writings on toi]et walls and noted their possible research value. Recently there has been a great deal of interest in graffiti as an index of social variablesn much of which has been of a popular nature and very little of which has been quantitative in any degree (e.g. Dundes, 1966; Lowas & Welman, l96G; Reisner 19679. However, Sechrest and Flores (1969) compared toilet wall writings in two different cultures, and Landy and Steele (1967) have reported quantitative findings concerning toilet wall writings in different kinds of university buildings. The present investigation is a study of (male) toilet wall inscriptions in four types of institutions of higher education in a large metropolitan area. Specifically, trade schools, junior collegess four year colIeges, and professional schools are involved. From the standpoint of the present investigation the main findings of the Kinsey study were that most (86%) inscriptions on the walls of male facilities are sexual in nature, and most (75%) of the sexual inscriptions involve homosexuality in some manner arld degree. Sechrest and Flores showed that the occurrence of homosexual inscriptions was culturally biased, the frequency in the U.S. being far higher than in the Philippines. It was surmised that the U.S.-Philippines difference was related to a difference in degree of homosexual conflict in the two cultures. On various grounds it seemed likely that homosexual ccxnflict was stronger in the U.S. than in the Philippines and, therefore that homosexual inscriptions would be more frequent in the American culture. The present study stemmed in part from findings of Kinsey (Kin-

Exercise as an Operant Response for Retarded Children

1 This research was supported by NIMH Grant #M 5749. The writer wishes to thank William Hines for his assistance in the construction of the apparatus, Ruth Wallace for her help in collecting data, and the Glenkirk School for making children available for the study. In the past decade or so, Skinnerian psychology has had a great impact on education and mental health in part because of its promise for bringing under the control of manipulable stimuli many diverse aspects of behavior. Early studies with rats and pigeons have been extended into investigations with primates and recently into a variety of efforts to control human behavior. Nearly all of the human responses shown to be controllable, however, have been responses requiring a low level of energy expenditure. It remains to be seen whether responses requiring a high energy expenditure, i.e., work, are equally responsive to simple reinforcement systems. Moreover, for obvious reasons, most of the human studies have attempted to sustain responses only over very brief periods of time, usually one or two experimental sessions. Again, it remains to be demonstrated whether work responses can be sustained over long periods of time by feasible schedules of reinforcement.