Graphical regression models for polytomous variables

When modeling the relationship between two nominal categorical variables, it is often desirable to include covariates to understand how individuals differ in their response behavior. Typically, however, not all the relevant covariates are available, with the result that the measured variables cannot fully account for the associations between the nominal variables. Under the assumption that the observed and unobserved variables follow a homogeneous conditional Gaussian distribution, this paper proposesRC(M) regression models to decompose the residual associations between the polytomous variables. Based on Goodman's (1979, 1985)RC(M) association model, a distinctive feature ofRC(M) regression models is that they facilitate the joint estimation of effects due to manifest and omitted (continuous) variables without requiring numerical integration. TheRC(M) regression models are illustrated using data from the High School and Beyond study (Tatsuoka & Lohnes, 1988). This article was accepted for publication, when Willem J. Heiser was the Editor ofPsychometrika. This research was supported by grants from the National Science Foundation (#SBR96-17510 and #SBR94-09531) and the Bureau of Educational Research at the University of Illinois. We thank Jee-Seon Kim for comments and computational assistance.     

Reviews: Books Photography

RELATIONAL PSYCHOANALYSIS: THE EMERGENCE OF A TRADITION. Edited by Stephen A. Mitchell and Lewis Aron. 514 pp. Hillsdale, NJ: The Analytic Press, 1999. $39.95. D.H. LAWRENCE AND THE PARADOXES OF PSYCHIC LIFE. By Barbara Ann Schapiro. 155 pp. Albany: State University of New York Press, 1999. $17.95. TECHNOLOGY AS MAGIC: THE TRIUMPH OF THE IRRATIONAL. By Richard Stivers. 240 pp. New York: Continuum Press. 1999. $24. REFLECTIONS ON A RAVAGED CENTURY. By Robert Conquest. 317 pp. New York: W. W. Norton & Company, 2000, $27.95. THE MYSTERIES WITHIN A SURGEON REFLECTS ON MEDICAL MYTHS. By Sherwin B. Nuland. 274 pp. New York: Simon & Schuster, 2000, $24.00. DISEASE, PAIN, AND SUICIDAL BEHAVIOR. By Elisabeth Stenager and Egon Stenager. 127 pp. New York: The Haworth Medical Press, 1998. $17.95. TRAUMA ASSESSMENTS: A CLINICIAN'S GUIDE. By E. B. Carlson. 307 pp. New York: The Guilford Press, 1997. $30. ON INFANCY AND TODDLERHOOD: AN ELEMENTARY TEXTBOOK. By David A. Freedman, M.D. 247 pp. Madison: International Universities Press, Inc., 1999. $27.95. TUESDAYS WITH MORRIE. By Mitch Albom. 192 pp. New York: Doubleday, 1997. $21. THE SORROWS OF THE QUAKER JESUS; JAMES NAYLER AND THE PURITAN CRACKDOWN ON THE FREE SPIRIT. By Leo Damrosch. 322 pp. Cambridge, MA: Harvard University Press, 1996, $39.95. THE PASTORAL CARE OF DEPRESSION: A GUIDEBOOK. By Binford W. Gilbert. 127 pp. New York: The Haworth Pastoral Press, 1998. $19.95. LEARNING DISABILITIES AND PSYCHIC CONFLICT: A PSYCHOANALYTIC CASEBOOK. By Arden Aibel Rothstein, Ph.D., and Jules Glenn, M.D. 504 pp. Madison: International Universities Press, Inc., 1999. $78. HAUNTED CHILDREN: RETHINKING MEDICATION OF COMMON PSYCHOLOGICAL DISORDERS. By Arthur F. Roemmelt, MD. 196 pp. New York: State University of New York Press, 1998. $17.95. NAVIGATING THE DEEP RIVER: SPIRITUALITY IN AFRICAN-AMERICAN FAMILIES. By Archie Smith, Jr. 181 pp. Cleveland, Ohio: United Church Press, 1997. $15.95. AL-JUNUN: MENTAL ILLNESS IN THE ISLAMIC WORLD. Edited by Ihsan Al-Issa. 382 pp. Madison: International Universities Press, 2000. $48. Witnessing the Lack of Sanctuary     

Measuring Adherence to Medical Treatments in Childhood Chronic Illness: Considering Multiple Methods and Sources of Information

Adherence to medical treatment is a significant problem for children and adolescents with chronic conditions, such as asthma, diabetes, and cystic fibrosis. The consequences of nonadherence can be serious, contributing to increased symptoms, unnecessary hospitalizations, and declines in physical functioning. The quality of data obtained from clinical trials can also be affected by poor adherence, leading to erroneous conclusions concerning the efficacy of drug treatments and the dosages that are needed to achieve those effects. Adherence problems in both clinical research and practice also lead to substantially higher health care costs. In order to further our understanding of the barriers that lead to poor adherence and identify strategies that are effective in addressing them, we need to develop reliable and valid measures of adherence behaviors. Using cystic fibrosis as a model of a serious, chronic disease that requires a difficult and time-consuming medical regimen, three different types of adherence measures are considered: self-report questionnaires, daily diary reports, and electronic monitors. The specific advantages and disadvantages of each type of measurement are reviewed, and specific recommendations are made for future research.     

The Role of Interest in Understanding the Career Choices of Female and Male College Students

Mismatch between college students' work goals and perceived goal affordances of physical/mathematical science careers may help explain gender differences in interest and career choice. In Study 1, the desire for interesting work was cited by most students in the sample (89% White, 6% Asian, 5% other). Compared to men, women reported interpersonal work goals more and high pay and status work goals less frequently. In Study 2, students (79% White, 12% Latino, 5% Asian, 4% other, predominantly middle class) perceived physical/mathematical science careers as less likely to afford interpersonal goals and more likely to afford high pay and status goals compared to other careers. Interpersonal goal affordances predicted greater interestingness for all careers, whereas high pay and status goal affordances predicted greater interestingness only for physical/mathematical sciences. Interestingness positively predicted likelihood of career choice.     

Contextual determinants of retrieval in three-month-old infants

Three-month old infants learned to kick to produce movement in an overhead crib mobile on a conjugate reinforcement (FRI) schedule. All infants were trained in the presence of one of two distinctive crib bumpers. In Experiment 1, infants receiving cued-recall tests in the presence of the original training bumper remembered the conditioned association 1 week later, but those tested in the presence of a different bumper did not. Two weeks later, neither training group evidenced retention of the contingency. In Experiment 2, infants who received a 3-min exposure (“a reactivation treatment”) to the original training bumper 24 h prior to the 2-week retention test exhibited excellent long-term retention, whereas those exposed to a different bumper as a reminder did not, irrepective of whether it was identical to the test bumper or not. These findings demonstrate that the context alone, even though it was never exclusively paired with reinforcement, can act as a retrieval cue.     

Characteristics of life events during adolescence

Two studies of the characteristics of life events during adolescence are reported. In the first, open-ended reports of major life events and daily events were obtained from early, middle, and late adolescents. Analyses of the frequency of past events perceived as positive versus negative revealed effects for type of event (major vs. daily), age, and gender. In contrast, anticipated events were perceived similarly across age and gender. The second study examined judgments by older adolescents of the valence, type, and domain of life events. Consensus regarding these characteristics was achieved on only a small portion of events. The implications of these findings for the measurement of life events during adolescence are highlighted.     

Androgyny in a male-dominated field: The relationship of sex-typed traits to performance and satisfaction in engineering

The relationship of sex-typed traits to performance and to satisfaction in engineering was investigated using a sample of men and women engineers with five years or less of professional work experience. Both men and women engineers high in instrumentality (androgynous and masculine) reported greater levels of supervisory and technical responsibility, salary, involvement in professional activities, and satisfaction than those low in instrumentality (feminine and undifferentiated). Expressiveness was not significantly related to any of the measures of performance or satisfaction. Although a few sex differences were found, the magnitude of the effects were generally smaller than those for instrumentality. Engineers' self-ratings of various abilities were also positively related to instrumentality. Only a small percentage of variance in the performance and satisfaction measures was accounted for by sex and sex-typed traits.