The isotonic ordinal probabilistic model (ISOP) is introduced as a common nonparametric theoretical structure for unidimensional models for quantitative, ordinal, and dichotomous variables. Fundamental theorems on dichotomous and polytomous weakly independent ordered systems are derived, and testing at the observed empirical level is discussed.…

Notes that presence of heterogeneous regression slopes in an analysis of covariance design is equivalent to presence of interaction in randomized-blocks design. Describes modification of the usual graphical representation of heterogeneous regressions that can aid in interpreting significant regions for regression surfaces. (NB)

A globally optimal solution is presented for a class of functions composed of a linear regression function and a penalty function for the sums of squared regression weights. A completing-the-squares approach is used, rather than calculus, because it yields global minimality easily in two of three cases examined. (SLD)

Presents a spreadsheet model that is useful in introducing students to regression analysis and the computation of regression coefficients. Includes spreadsheet layouts and formulas so that the spreadsheet can be implemented. (Author)

A methodology is proposed to optimize the training classification performance of any suboptimal model. The method, referred to as univariate optimal discriminant analysis (UniODA), is illustrated through application to a two-group logistic regression analysis with 12 empirical examples. Maximizing percentage accuracy in classification is…

J.H. Kranzler and A.R. Jensen report on a reanalysis of previous data that is again inconsistent with the hypothesis that psychometric "g" is unitary. J.B. Carroll reiterates objections to the evidence that Kranzler and Jensen offer in support of a composite "g." (SLD)

Used multiple regression analysis to develop demographically corrected norms for letter and category fluency in 768 normal adults. Together, age, education, and ethnicity were significant predictors of letter and category fluency performance. Presents tables for converting raw fluency scores to these norms. (SLD)

Discusses the use of Microsoft Excel software and provides examples of its use in an online statistics course at Golden Gate University in the areas of randomness and probability, sampling distributions, confidence intervals, and regression analysis. (LRW)

Current Population Survey data from 1979 and 1991 were used to decompose changes in the gender wage gap into three components: skill distribution, wage structure, and improvements in women's position. Relative wage gains by women may have been a source of increasing wage inequality among men. (SK)

An alternative strategy for computing factor scores was introduced and compared to a popular scoring procedure. The new strategy, which involves unit-weighted composites of standardized items with salient factor score coefficients, is shown superior to the common method. Implications of findings are discussed. (SLD)

Although data is often used to estimate parameters for models in calculus and differential equations, the models themselves are seldom justified. Uses the data itself to motivate mathematical models in introductory mathematics courses. Illustrates various regression and optimization techniques. (Author/ASK)

Examined the regressions of occupational status and income on psychometric "g" factor scores in large samples of white (n=3,484) and black (n=493) U.S. armed forces veterans in their late 30s. Results indicate that, for both job status and income, whites were relatively more disadvantaged when the level of "g" is taken into…

Rasch's multiplicative Poisson model is extended so that parameters for individuals in the prior gamma distribution have continuous covariates. Parameters for individuals are integrated out, and hyperparameters in the prior distribution are estimated by a numerical method separately from difficulty parameters that are treated as fixed parameters…

The Gibbs sampling algorithms presented by M. H. Seltzer (1993) are fully generalized to a broad range of settings in which vectors of random regression parameters in the hierarchical model are assumed multivariate normally or multivariate "t" distributed across groups. The use of a fully Bayesian approach is discussed. (SLD)

Empirically estimates a multiple-product fixed total cost function and output relationship for 295 U.S. private liberal-arts colleges. Statistical results suggest that there are economies of scale and scope in higher education. Research at such colleges is not cost-effective. (Contains 31 references.) (MLH)