Multivariate generalizability techniques are used to bridge the gap between psychometric constraints and the tables of specifications needed in test development. Techniques are illustrated with results from the American College Testing Assessment Program. (Author/PN)

Several randomized response models for gathering self-report data when persons are asked sensitive questions are reviewed. A new general model for reducing error by asking questions conditional upon earlier answers is introduced. Hypothesis testing and confidence-interval procedures are demonstrated for data collected by the randomized response…

Suggests a method for analyzing structure and process in prose using algorithms based on graph theory and rules based in syntax. Describes application of the method to essays written by 52 fourteen year old pupils on a newly learned topic in chemistry. (Author/GC)

Observational data were collected over a period of 120 school days from 11 elementary-level, self-contained, learning disabilities classrooms (N=105), using the Student Level Observation of Beginning Reading system. Findings indicate that reliable and unbiased observations of student behavior were obtained through the use of a single observer.…

Implications of model evaluation (using traditional chi square goodness of fit statistics, incremental fit indices for covariance structure models, and latent variable coefficients of determination) on substantive conclusions are illustrated with an example examining the effects of participation in a compensatory education program on posttreatment…

Formulas for estimating the population measure of effect strength are based on the assumption that sample sizes are proportional to the sizes of their respective treatment populations. Because this assumption is frequently violated, a general method of estimating effect strength for the one-factor, fixed-effects design is presented. (Author/BW)

Browne's definitive but complex formulas for the cross-validational accuracy of an OSL-estimated regression equation in the random-effects sampling model are here reworked to achieve greater perspicuity and extended to include the fixed-effects sampling model. (Author)

Presents a mathematical model utilizing three-letter codes of personality patterns determined from the Self Directed Search. This model compares personality types over time or determines relationships between personality types and person-environment interactions. This approach is consistent with Holland's theory yet more comprehensive than one- or…

The development of probability judgment is explained in terms of working memory, composed of four types of storage operations and three types of processing operations. Age changes in probability judgment were related to changes in frequency retrieval, which stem from changes in constraints on work-space capacity. (Author/RD)

Examines data from California from the 1850 opening of the prison system to 1970. Finds no evidence to support the hypothesis that, for a given society over time and in the absence of major societal upheaval, the proportion of people subjected to punishment by the state closely approximates a constant. (Author/GC)

Examines early retirement plans (ERPs) for public school teachers, compares them with private-sector ERPs, and notes legal considerations involving ERPs' impact on regular retirement systems. Discusses methods for cost analyses of ERPs and proposes a model for assessing an ERP's fiscal impact on a school district. (RW)

Six statistical methods for identifying possible sources of bias in standardized test items are presented. The relationship between chi-squared methods and item-response theory methods are also discussed. (Author/BW)

The jackknife by groups and modifications of the jackknife by groups are used to estimate standard errors of rotated factor loadings for selected populations in common factor model maximum likelihood factor analysis. Simulations are performed in which t-statistics based upon these jackknife estimates of the standard errors are computed.…

An integrated study of Shine's actualized and pure single-subject behavior functions can produce more information than studying only one function. After summarizing the mathematics behind Shine's viewpoint, an ordinary regression analysis approach to the actualized behavior function is integrated with a time-series analysis approach to the pure…

Four measures of influence in intercorporate networks are tested, using data from a sample of large American corporations in 1904. Comparison with historical accounts of the period indicate that modifications of the Bonacich index of a point's centrality in a network yield the most reliable results. (Author/RW)