Logistic latent trait models with linear constraints

Two linearly constrained logistic models which are based on the well-known dichotomous Rasch model, the ‘linear logistic test model’ (LLTM) and the ‘linear logistic model with relaxed assumptions’ (LLRA), are discussed. Necessary and sufficient conditions for the existence of unique conditional maximum likelihood estimates of the structural model parameters are derived. Methods for testing composite hypotheses within the framework of these models and a number of typical applications to real data are mentioned.     

Pastoral Care: Myth and Reality

The increase in pastoral-care systems in secondary schools over the last fifteen years is examined. It is suggested that pastoral care is more effective in allaying the anxieties of teachers and their organisations than it is in meeting the social and personal needs of pupils. It is also suggested that pastoral-care systems assume responsibilities which are more properly those of the family, and that the best way schools can offset the effects of poor home environments is via good subject teaching and increased pupil involvement in schools.     

Perceptions of date and stranger rape: A difference in sex role expectations and rape-supportive beliefs

The sex role expectations and rape-supportive beliefs which are differentially applied to date and stranger rape were assessed. A sample of 62 college females and 33 males read one of three rape scenarios which varied according to the victim-perpetrator relationship (steady dating partners/acquaintances on a first date/strangers). Then participants rated the extent to which several sex role expectations and rape-supportive attitudes were applicable to the assault. MANOVAS and univariate ANOVAS showed that females' and males' perceptions of date rape, more than stranger rape, incorporated sex role expectations and that rape-supportive beliefs, primarily of males, were stronger in relation to steady date than first date or stranger rape. These findings are discussed in the context of the sex role analysis of rape.This research was supported by the University of Connecticut Research Foundation Grant #1171-000-22-00215-35-760. The author thanks Laurin Hafner for his help with the data analysis.     

The job climate for women in traditionally male blue-collar occupations

Rapid growth has been observed in recent years in the number of women entering traditionally male blue-collar occupations, yet researchers have paid little attention to this pioneering group. The present study undertook a comparison of the job conditions of two groups of nontraditional women workers, tradeswomen (N=71), and transit workers (N=151); a third group, school secretaries (N=389), was studied as a comparison group of traditionally female workers. Results from a 16-page mail questionnaire revealed that women in traditionally male occupations encountered significantly more adverse working conditions than did their traditional counterparts, and, in addition, reported significantly less satisfaction and more stress at work. Tradeswomen were the most likely to experience sexual harassment and sex discrimination, and black tradeswomen to experience race discrimination. The degree of job satisfaction expressed by the secretaries was unexpected, since most evidence suggests that clerical workers lack autonomy, and encounter boredom and routinization on the job. These findings are discussed in terms of gender segregation and the need to focus future research efforts on specific occupational groups in order to make appropriate policy recommendations as well as to provide help for women in these jobs.This research was supported by grants from the National Institute for Occupational Safety and Health (Grant No. 1 RO1 OH02162-01A1) and the College of Health and Human Development, The Pennsylvania State University. The authors wish to acknowledge the invaluable contributions of Patty Lambert, Ming Qi Wang, Mary Harmon, Joseph Vasey, Gina Jay, Darrell Harvey, and Haleh Rastegary.     

Age and depression: unique and shared effects

The degree to which psychosocial variables associated with depression were also associated with age was examined in 3 samples of community residents 50 years of age or older (N = 4,617). Most of the expected concomitants of depression were found. With only a few exceptions, age was not correlated with depression-related psychosocial variables. Rather, age was most strongly associated with levels in neuropsychological and psychophysiological functioning. In addition, the magnitudes of the correlations in women compared with men and in young-old age groups compared with old-age groups were examined. A number of significant differences emerged, and their implications for theories of depression are noted.     

Planning in the preschool years: The emergence of plans from general event knowledge

A framework for conceptualizing the relationship between event knowledge and planning is proposed, and two experiments are reported that examined children's ability to draw on event knowledge in planning. Preschool children were asked to plan and execute shopping trips to a pretend grocery store. Half of the children planned for two events on successive trials (Level 1, single-goal planning) and half of the children planned for two events simultaneously (Level 2, multiple-event planning). The amount of external support for planning was also manipulated. In Experiment 1, 3-, 4-, and 5-year-olds were presented with either a clustered or interleaved display. In Experiment 2, 3- and 4-year-olds were given adult assistance in plan construction. Results indicated that children's planning becomes more complex and flexible with age. Older children also rely less on external supports for planning. However, when external support was provided, 3- and 4-year-olds displayed higher-level planning abilities. Results are discussed in terms of the roles of event knowledge and external support in the early development of planning skills.     

Perceived depth vs. structural relevance in the object-superiority effect

When subjects must identify a barely visible line in a briefly flashed display, their accuracy depends on the configuration of the context in which the target line appears. Weisstein and Harris (1974) found that accuracy is highest when the target is part of a pattern that resembles a unified, three-dimensional object, and lowest in a flat-looking pattern composed of disconnected lines; they labeled this phenomenon the object-superiority effect. In the three experiments reported here, identification accuracy was found to correlate highly and significantly (r =.78) with the judged depth of the patterns. Judged structural relevance of the target line to the pattern (McClelland & Miller, 1979) was uncorrelated with accuracy (r=?.28). Even when the target line appeared as an isolated fragment within the context pattern, a pattern perceived as three-dimensional yielded higher identification accuracy than one perceived as flat.     

Facial and vocal cues of deception and honesty

Facial expressions and vocal cues (filtered speech) of honest and deceptive messages were examined in posed and spontaneous situations. The question of interest was the degree to which nonverbal cues transmit information about deception. Results indicated that (a) for both the facial and vocal channels, posing (as compared to spontaneous behavior) produced a higher level of communication accuracy; (b) facial expressions of deceptive (as compared to honest) messages were rated as less pleasant, while vocal expressions of deception were rated as less honest, less assertive, and less dominant, particularly in the posed condition; (c) the sender's ability to convey honesty was negatively correlated with his/her ability to convey deception, suggesting the existence of a demeanor bias—individual senders tend to appear and sound consistently honest (or dishonest) regardless of whether they deliver an honest or a deceptive message; (d) in the posing condition, the sender's abilities to convey honesty/deception via facial and vocal cues were positively and significantly correlated, whereas in the spontaneous condition they were not; and (e) senders whose full (unfiltered) speech indicated more involvement with their responses were judged as more honest from both their vocal (filtered speech) and facial cues, in both the honest and deceptive conditions.