Refined item digraph analysis (RIDA) is a way of studying diagnostic and prescriptive testing. It permits assessment of a test item's diagnostic value by examining the extent to which the item has properties of ideal items. RIDA is illustrated with the Orange Juice Test, which assesses the proportionality concept. (TJH)

Mathematical analysis and computer simulation methods are used to show that interactive models of context effects can exhibit classical context effects if there is variability in the input to the network or the network itself. Interactive models represent hypotheses about information-processing dynamics leading to the global asymptotic behaviors…

This paper proposes that patterns of proximity data that have been characterized in terms of asymmetric similarity may be alternatively characterized in terms of differential bias. An additive similarity and bias model is reviewed, and it is proposed that biases can be stimulus based as well as response based. (SLD)

A theoretical framework is developed in which the effects of some common forms of violation of assumptions of linearity of regression and homoscedasticity can be investigated. Simple expressions are derived for the restricted and corrected correlations in terms of the target (unrestricted) correlation in these situations. (SLD)

Argues that most currently used information retrieval models are unsuitable to describe recent techniques such as semantic based retrieval. A more general model is presented in which the basis for information retrieval is viewed as logical implication and the estimation of the correspondence between queries and documents is described in terms of…

New methods of analysis--equations and graphs for iso-r(sup 2) contours--were introduced and used to illustrate location effects for pooled data sets. The "r(sup 2)" is the coefficient of determination. Results are used to highlight imprecise statements in the literature about the behavior of the correlation coefficient for pooled data…

Based on the Gaussian model, methods for using measurements that depend on the true attribute to compute rankings are proposed and compared. Measurements based on an empirical Bayes model produce estimates that differ from ranking observed data. Ranking methods are illustrated with school achievement data. (TJH)

A method is proposed for detecting influential observations in iterative principal factor analysis. Theoretical influence functions are derived for two components of the common variance decomposition. The major mathematical tool is the influence function derived by Tanaka (1988). (SLD)

Linear equating methods for the common-item non-equivalent populations design were compared when true-score correlation between the test and anchor was less than unity. Scores from two groups of approximately 300 examinees illustrated three methods: (1) the Tucker equating method; (2) the Angoff-Levine method; and (3) the Congeneric-Levine method.…

An approach is provided for calculating maximum values of the Kappa statistic of J. Cohen (1960) as a function of observed agreement proportions between evaluators. Separate calculations are required for different matrix sizes and observed agreement levels. (SLD)

The average response method (ARM) of scaling nonbinary data was developed to scale data from the assessments of writing conducted by the National Assessment of Educational Progress (NAEP). The method is described and illustrated with data from the 1983-84 NAEP. (SLD)

Confidence envelopes for one-parameter, two-parameter, and three-parameter logistic item response models are illustrated. M-line plots showing the genesis of the envelope and the density of lines in the confidence region are described and illustrated. (Author/SLD)

Issues arising from difference scores with zero reliability that nevertheless allow a powerful test of change are discussed. Issues include the appropriateness of underlying statistical models for psychological data and the relationship between difference scores and power. Increases in reliability always increase power for a fixed effect size.…

A technique to estimate variances of weighted estimates in program evaluation is discussed. BRR estimates of variance are constructed from orthogonally weighted subsamples and can provide better estimates from complex samples than can other methods. Use of BRR is illustrated in a government evaluation of a food service program. (SLD)

Two forms, pure-rank and mixed-rank, of a nonparametric, general, linear model-based statistic that can be used to test several hypotheses are presented. A Monte Carlo study was used to investigate the distributional properties of these forms, and their use is discussed. (SLD)