This paper introduces a model for describing outliers (observations which are extreme in some sense or violate the apparent pattern of other observations) in linear regression which can be viewed as a mixture of a quadratic and a linear regression. The maximum likelihood estimators of the parameters in the model are derived and their asymptotic…

The approach of J. D. Sneed is used to argue for a distinction between theoretical and nontheoretical (or empirical) entities. Attention is directed toward this approach as an important facility for looking at statistics and for constructing a framework for the correct use of statistical theory in information retrieval experiments. (MBR)

Recommends that analysts (1) use conventional meta-analytic statistics when testing for moderator variables; (2) perform tests between mean effect sizes; and (3) continue to perform model tests in meta-analyses for which study outcomes are already consistent if they have theoretical expectations about moderators. (RS)

Presents two models for statistically analyzing occupational mobility. The models show how the destination distribution is shifted for each shift of one or more steps on the scale of occupational origins. (Author/AV)

A methodological framework is provided for quantitatively integrating Type I error rates and power values for Monte Carlo studies. An example is given using Monte Carlo studies of a test of equality of variances, and the importance of relating metanalytic results to exact statistical theory is emphasized. (SLD)

The use of statistical methods to combine the results of independent empirical research studies (metanalysis) has a long history, with work mainly divided into tests of the statistical significance of combined results and methods for combining estimates across studies. Methods of metanalysis and their applications are reviewed. (SLD)

A basic methodological approach in developmental studies is the collection of longitudinal data. Behavioral data cen take at least two forms, qualitative (or discrete) and quantitative. Both types are fallible. Measurement errors can occur in quantitative data and measures of these are based on error variance. Qualitative or discrete data can…

Four experiments designed to study attitudes toward United States Presidents are described. The project uses the conceptual framework and methodology of a mathematical theory of information integration to compile information on attitudes and opinion formation based on high school civics and history courses. Results of the research findings of each…

Four different procedures were used for estimating the proportion of persons who would be classified consistently as either passing both of two parallel tests or failing both. These four methods were applied at each of four different mastery level scores for each of three different length tests. Data were based on 50 replications of each procedure…

The planning and management of educational systems has been examined widely in the last decades despite the lack of a standard research method and a unified underlying theory. Research based on quantitative analysis yields data that may be processed using the graphic capabilities of computers. The combination of numerical data and computer…

A MODEL IS CONSTRUCTED TO DETERMINE THE BEST PROPORTION OF FORMAL AND ON-THE-JOB TRAINING IN MILITARY OCCUPATIONS. SPECIAL CONSIDERATION IS GIVEN TO THE UNIQUE SITUATION OF ENLISTED PERSONNEL'S FIXED LENGTH OF SERVICE, THE SMALL PERCENTAGE OF RE-ENLISTMENT, AND THE NECESSITY OF TRAINING ALL ENLISTEES FOR ASSIGNED OCCUPATIONS. THE MODEL FORMULA…

Selection of items for analogy tests according to the Rasch item probability of "goodness of fit" to the model is compared with three commonly used item selection criteria: item discrimination, item difficulty, and item-ability correlation. Word, picture, symbol and number analogies in multiple choice format were administered to several…

Bayesian estimation procedures are summarized and numerically illustrated by means of simulation methods. Procedures of data generation for simulation purposes are also delineated and computationally demonstrated. The logistic model basic to the Bayesian estimation procedures is shown to be explicit with respect to the probability distribution…

The growing complexity of policy decisions in such areas as education and manpower necessitates increased research, which in turn requires an adequate data base and adequate methodology. Recognition that the variables affecting the vocational and educational choices of high school graduates are interrelated and require examination from several…