A logistic regression model for characterizing differential item functioning (DIF) between two groups is presented. A distinction is drawn between uniform and nonuniform DIF in terms of model parameters. A statistic for testing the hypotheses of no DIF is developed, and simulation studies compare it with the Mantel-Haenszel procedure. (Author/TJH)

Methods and variables used to analyze Idaho State University faculty salaries are described, and results of salary analyses are presented. Practical issues confronting the salary study committee, procedures used to resolve them, and limitations of regression analysis are also examined. (MSE)

This study of 218 White and 132 Black students (mean age=10.3) referred for possible learning disability services found that Black subjects were more likely to exhibit a severe discrepancy between expected and actual achievement levels when using a regression procedure than when using a standard score comparison procedure. (Author/JDD)

Critiques William Becker, William Greene, and Sherwin Rosen's review of the research regarding the teaching and learning of high school economics. Calls for sharper definitions of research models to improve focus. Explains a distinction between stock and flow models, and asks for elaboration on topics covered. Suggests examining textbook…

Used Holland's Vocational Preference Inventory (VPI) and a self-monitoring scale to examine 237 undergraduates' association between self-monitoring and occupational preferences. Regression analyses revealed correlations between gender, self-monitoring propensity, and preferred occupational types (i.e., social, enterprising, or artistic). (TE)

A model is presented for the growth of knowledge reflected by 24 progress tests completed by approximately 600 students at the University of Limburg (Netherlands) Medical School. Based on the Rasch model, this model treats both the person's ability and the difficulty of the question as random variables. (SLD)

The diagnostic decision-making standards used by practicing clinicians to determine language impairment were investigated. Results showed significant interrater agreement among the 27 clinicians' decisions and moderate intrarater reliability within clinician's decisions. Most of the clinicians' diagnostic decision-making standards could be modeled…

For a representative national sample (n=1,200) of children from 5 to 17 years old, ability constructs were assessed through the Differential Abilities Scale. Adjustment constructs were assessed through teacher ratings with the Adjustment Scales for Children and Adolescents. Partialled canonical redundancy and regression analyses revealed that…

Evaluation of faculty by executive program participants provides a useful teaching-effectiveness measure. Ratings of individual teaching characteristics can help identify the relative weights of certain characteristics, thereby enabling faculty to concentrate improvement efforts on characteristics yielding the biggest payoff. Faculty effectiveness…

Investigates the job search strategies of high school students (n=261). Respondents differentiated between self-effort and external strategies in seeking employment. Girls were found to be significantly more supportive of both these strategies, and marked differences were noted for males and females. It is concluded that different processes may…

The effects of scheduled intern rotations on the cost and quality of inpatient care at a teaching hospital were studied for 1,705 rotation patients and 3,141 no-rotation patients. Systematic discontinuity of scheduled rotations may be a source of increased health care costs at teaching hospitals. (SLD)

Areas in which human judgment must be used to conduct a fair statistical analysis of college faculty salary equity are discussed, and guidelines for making those judgments are offered. Four decision elements and four fields of study are considered as the basis for establishing criteria for selecting the appropriate regression model. (Author/MSE)

Reports the preliminary results of a survey of the environmental scanning behavior of 207 chief executive officers (CEOs) in the Canadian publishing and telecommunications industries. Topics addressed include information sources, perceived source accessibility, perceived source quality, and frequency of source use. (16 references) (LRW)

Regression to the mean (RTM) is often misunderstood. It is demonstrated that artifactual RTM depends fundamentally on the magnitude of measurement error at pretest. Adjustment usually involves estimating the measurement error and determining consequences, but even without adjustment, effects of RTM can be ameliorated. (SLD)

Analysis of hypothetical data with dependent responses demonstrates how information loss caused by the overt response scale has an unknown influence on effect sizes in moderated regression analysis. The number of scale steps measuring the dependent variable results in a form of systematic error that alters interaction effect sizes. (SLD)