The purpose of the study was to develop a linear programming model to be used for prediction, test the accuracy of the predictions, and compare the accuracy with that produced by curvilinear multiple regression analysis. (Author)

Contents of this book include: an introductory preface by C. Loring Brace; "Introduction to Jensenism," C. Loring Brace; "Can we and should we study race differences?" Arthur R. Jensen; "Intelligence in Black and White," Alexander Alland, Jr.; "Whose is the failure?" Vera John; "The influence of…

The Answer-Until-Correct (AUC) procedure has been proposed in order to increase the reliability of multiple-choice items. A model for examinees' behavior when they must respond to each item until they answer it correctly is presented. An expression for the reliability of AUC items, as a function of the characteristics of the item and the scoring…

This study was undertaken for the Center for Applied Linguistics to survey and analyze its information system program for the language sciences. The study identifies and defines the generalized sets of parameters required for subsequent quantitative analysis of proposed alternative Language Information Network and Clearinghouse Systems by means of…

Previous Studies have yielded mixed results concerning the relationships between student learning, instructor leniency, and student evaluations of instructors. A three-equation model is presented that disentangles learning from leniency and relates the results to evaluations. It indicates no significant relationship between either leniency or…

The Rasch Model was applied to the results of an academic self-concept test administered to junior high school students, and some items were eliminated. The resulting scale possessed several properties permitting its use in the measurement of school-induced change in self-concept. The questionnaire is appended. (Author/MV)

An application of confirmatory factor analysis, using the LISREL program, to retrospective analysis of rating data common in health professions education was illustrated. A causal model was specified and tested. Although most measures clustered around latent traits, there was more method variance in the data set than trait variance. (Author/JAZ)

Develops procedures for the automatic generation and incorporation of graphics (both statistical and analytical) into mathematically oriented multiple choice examinations. Uses an integrated spreadsheet to create and associate tabular data, diagrams, questions, and answers. Illustrates the process of constructing multiple choice questions and…

Argues that "appropriate practical work" in mathematics aids understanding. Presents ideas and activities for teaching mechanics. Discusses modelling, problem formulation, and interpretation. (PK)

Fifth graders performed better than third graders on discovering the function rule relating two number pairs. Task difficulty and mode of presentation (graphic or symbolic) varied by function. The children seemed ready to relinquish a hypothesis in the face of disconfirming evidence. (MNS)

Using data from High School and Beyond, this study empirically investigated the extent of within-variable, between-occasion error covariances among variables included in educational achievement models. Little evidence was found to support the statement that reliability estimates for social background variables are inflated because of correlated…

Multidimensional item response data were created. The strength of a general factor, the number of common factors, the distribution of items loadingon common factors, and the number of items in simulated tests were manipulated. LOGIST effectively recovered both item and trait parameters in nearly all of the experimental conditions. (Author/JAZ)

This paper presents parallel estimates of structural equation models to explain children's educational attainment over first grade. One set of estimates is based on a 3-school sample drawn in 1971-77 and another on a 20-school sample drawn in 1982-83. The evidence suggests that, within reasonable limits, results are comparable. (Author/JAZ)

The Lisrel-model with structured means was used to study similarities and differences in the development of mathematical ability between two student groups as measured by two tests on three successive occasions. Procedures for testing models to assess the contribution on an individual parameter to the goodness of fit are described. (Author/BS)

High School and Beyond data set and path analytic techniques were used to compare Black and Hispanic high school seniors' achievement in public and in Catholic schools. When better ability measures were added to the causal models the apparent effect of Catholic schooling on minority achievement was greatly reduced. (Author/BS)