The asymptotic posterior normality of the latent trait in an IRT model

It has long been part of the item response theory (IRT) folklore that under the usual empirical Bayes unidimensional IRT modeling approach, the posterior distribution of examinee ability given test response is approximately normal for a long test. Under very general and nonrestrictive nonparametric assumptions, we make this claim rigorous for a broad class of latent models.This research was partially supported by Office of Naval Research Cognitive and Neural Sciences Grant N0014-J-90-1940, 442-1548, National Science Foundation Mathematics Grant NSF-DMS-91-01436, and the National Center for Supercomputing Applications. We wish to thank Kumar Joag-dev and Zhiliang Ying for enlightening suggestions concerning the proof of the basic result.The authors wish to thank Kumar Joag-Dev, Brian Junker, Bert Green, Paul Holland, Robert Mislevy, and especially Zhiliang Ying for their useful comments and discussions.     

Three forms of stress in Cambodian adolescent refugees

Sixty-nine Cambodian adolescents and young adults were interviewed to determine their experience as children surviving the Pol Pot regime (1975–1979);their first-year experience of resettlement in this country; and their experience of stressful events during the past year. Current DSM-III-R diagnostic status was also determined. A strong relationship between earlier war trauma, resettlement strain, and symptoms of posttraumatic stress disorder (PTSD) was found. In contrast, the strongest relationship with depressive symptoms was found for recent stressful events. These results are discussed in light of current findings from stress and PTSD research.Dr. Clarke is Assistant Professor of Psychiatry in the Division of Child Psychiatry, Oregon Health Sciences University, Portland, Oregon. Dr. Sack is Director of the Division of Child Psychiatry and Principal Investigator of the Khmer Adolescent Project. Brian Goff is a Research Assistant.This project was supported by the National Institutes of Mental Health [NIMH] grant #5-RO1-MH42927-02, Dr. William H. Sack, M.D., Principal Investigator.     

Gender roles,computer attitudes,and dyadic computer interaction performance in college students

Females tend to score lower than males on measures of computer aptitude and attitudes. This study examined the potential effects of several mediating factors, sex of experimenter, sex of experimental partner, sex typing (Bem Sex Role Inventory score), and level of past experience, on a computer interaction task involving a dyad. College students, drawn from a primarily white college population, after completing paper-and-pencil measures assessing computer experience, computer attitudes, and gender roles, participated in a computer task involving drawing up to ten prescribed geometric patterns on a computer screen. Research assistants recorded the number of correctly completed patterns and videotaped participant interaction through a one-way mirror. Males reported experiencing more computer involvement than females. For males, past computer experience and masculinity were correlated with more positive computer attitude scores. For females, only past computer experience related to more positive computer attitude scores. Analyses of the computer interaction variable indicated that participants asked male research assistants significantly more questions than female research assistants. Several interaction effects were also found.A previous version of this paper was presented at the annual meeting of the Southwestern Psychological Association, New Orleans, April 1991.     

Social reactions to depictions of casual and steady acquaintance rape: The impact of AIDS exposure and stereotypic beliefs about women

While much of the early research on the social perception of sexual assault focused on instances of stranger rape, recent work has indicated that acquaintance rape occurs much more frequently. The purpose of the present investigation was to examine a gender role socialization analysis of acquaintance rape. The subjects in the present investigation, predominately Caucasian-American females (i.e., over 90% were nonminority), were first preclassified as either traditional or nontraditional in their gender role views about women. We then examined their perception of rape victims and rape perpetrators as a function of whether the rape occurred between casual dating partners vs. long-term steady daters, and also as a function of whether the assailant exposed the rape victim to the disease AIDS (vs. no exposure to AIDS). The results indicated that individuals who varied in their stereotypic views about women differed in their reactions, with more traditional subjects being more likely to negatively evaluate the rape victim and less likely to evaluate the rapist in an unfavorable manner. In addition, it was found that when rape victims were exposed to AIDS, the social perceptions of others were strongly influenced, with casual-dating rape acquaintances being more likely to bear the onus of a negative reaction from others.     

Reasons for Living Following Success and Failure in Suicidal and Nonsuicidal College Students1

To see if suicide ideators responded differently than nonideators after success or failure experiences, 61 college students who were relatively high or low ideators participated in a computer task on which half got bogus success feedback, the other half got bogus failure feedback. Both groups then completed the Reasons for Living Inventory (RFL). Two (High versus Low Suicide) X 2 (Success or Failure) analyses indicated main effects for suicide on overall RFL and four of the six RFL subscales. Although there were no main effects for success versus failure, an interaction indicated that, in the failure condition, low suicidal subjects were significantly higher than high suicidal subjects in overall reasons for living. Results suggest that suicidal individuals may be vulnerable to failure, and may need interventions to help them avoid potential failure or to bolster their coping mechanisms.     

Neural nets for generalization and classification: comment on Staddon and Reid (1990)

The neural net model of Staddon and Reid (1990) explains exponential and Gaussian generalization gradients in the same way as the diffusion model of Shepard (1958). The "cognitive" generalization theory of Shepard (1987), which also has been implemented as a connectionist network, goes beyond both of these models in accounting for classification learning.     

Analyzing sequential categorical data: Individual variation in markov chains

Markov chains are probabilistic models for sequences of categorical events, with applications throughout scientific psychology. This paper provides a method for anlayzing data consisting of event sequences and covariate observations. It is assumed that each sequence is a Markov process characterized by a distinct transition probability matrix. The objective is to use the covariate data to explain differences between individuals in the transition probability matrices characterizing their sequential data. The elements of the transition probability matrices are written as functions of a vector of latent variables, with variation in the latent variables explained through a multivariate regression on the covariates. The regression is estimated using the EM algorithm, and requires the numerical calculation of a multivariate integral. An example using simulated cognitive developmental data is presented, which shows that the estimation of individual variation in the parameters of a probability model may have substantial theoretical importance, even when individual differences are not the focus of the investigator's concerns.Research contributing to this article was supported by B.R.S. Subgrant 5-35345 from the University of Virginia. I thank the DADA Group, Bill Fabricius, Don Hartmann, William Griffin, Jack McArdle, Ivo Molenaar, Ronald Schoenberg, Simon Tavaré, and several anonymous reviewers for their discussion of these points.     

Gender stereotypes and dimensions of effective leader behavior

Research has consistently found that effective leadership is perceived as characterized by traits similar to those associated with masculine gender roles. These perceptions would appear to be at odds with extensive research indicating that effective leadership requires consideration and structuring behaviors—behaviors that seem to represent both masculine and feminine styles. In two separate studies, the correspondence between gender stereotypes and dimensions of effective leadership were assessed. Results indicate that consideration behaviors are perceived to be feminine, while structuring behaviors are perceived to be masculine. Similarly, qualities that characterize the masculine gender role are perceived to be consistent with structuring, while qualities associated with the feminine gender roles are perceived to be consistent with consideration. It is suggested that an increased awareness of the androgynous nature of effective leadership behaviors might weaken the biases in favor of male leaders.     

Feigned insanity in nineteenth-century America: Tactics,trials, and truth

Feigned insanity in nineteenth-century America is appraised through a review of the medical and legal literature. The authors focus on the explanations for feigning, procedures used in uncovering feigning, and the role of feigning in the courtroom. This discussion of feigned insanity demonstrates the remarkable consistency of approach to this form of malingering over the past 200 years.     

An MMPI-based empirical model of malingering and deception

The empirically-based model of malingering and deception described by Rogers (1984b) was operationalized using MMPI and clinical interview data. Subjects (N = 159) were patients committed to an inpatient forensic hospital as ‘Incompetent to Stand Trial’ or ‘Not Guilty by Reason of Insanity’. Inter-rater reliability was acceptable for the 11 criteria used to define response styles. Each subject was categorized into one of five response style groups: ‘Reliable’, ‘Malingering’, ‘Defensive’, ‘Irrelevant’, or ‘Unclassifiable’. Factor analysis of the rating criteria yielded four factors, three of which are comparable to the assigned groups, providing some validation for the constructs underlying these groups.