The stepwise regression method of selecting predictors for computer assisted multiple regression analysis was compared with forward, backward, and best subsets regression, using 16 data sets. The results indicated the stepwise method was preferred because of its practical nature, when the models chosen by different selection methods were similar…

University-level mathematics placement examinations might yield summary measures that are differentially accurate due to individual differences in the recall of previously-learned material. In particular, test scores in algebra and trigonometry may be subject to sporadic recall of solution procedures, resulting in aberrant response patterns. The…

Two studies investigated the role that parental attitudes and behavior play in filtering the complexity of the world into manageable forms for children. Cognitive and behavioral patterns used as dependent variables were cognition (reading scores, language use, general information, beliefs, reality-fantasy discrimination); imagination;…

The use of a robust statistical technique to implement a decentralized faculty allocation procedure at a medium size public university is discussed. The least median of squares (LMS), a regression technique introduced by Rousseeuw, was used instead of the usual least sum of squares, or least squares. The LMS technique was used to assist in the…

This paper investigates the reasons why teachers change school districts by examining teacher characteristics and school districts as predictors of district change. The data set traces the careers of Michigan public school teachers during the 1970s. The study is restricted to districts with stable or expanding enrollments to ensure that the job…

Five distribution-free alternatives to parametric analysis of covariance (ANCOVA) are presented and demonstrated using a specific data example. The procedures considered are those suggested by Quade (1967); Puri and Sen (1969); McSweeney and Porter (1971); Burnett and Barr (1978); and Shirley (1981). The results of simulation studies investigating…

The classification problem in educational testing is a decision problem. One must assign subjects to one of several available treatments on the basis of test scores, where the success of each treatment is measured by a different criterion. Examples of classification decisions include individualized instruction, counseling, and clinical settings.…

A prediction equation for teacher morale was developed from the responses of 236 teachers in a large school system in Mississippi. The Supervisory Conference Rating (SCR) was used to measure teachers' perceptions of the supervisory conference following observation by the principal. The Zones of Indifference of these teachers were measured with the…

Least squares fitting process as a method of data reduction is presented. The general strategy is to consider fitting (linear) models as partitioning data into a fit and residuals. The fit can be parsimoniously represented by a summary of the data. A fit is considered adequate if the residuals are small enough so that manipulating their signs and…

The traditional core Masters in Business Administration (MBA) curriculum consists of a broad range of courses that can be considered as a whole, or divided into qualitative and quantitative courses. Regression models were developed with "QualGPA" and "QuantGPA" as response variables, and gender, pre-MBA academic indicators, and…

This paper focuses on the methodology of cross group comparisons of regression coefficients commonly used to identify group differences in status attainment. Using the study of race differences in educational attainment as an example, the paper reviews a number of relevant published studies and demonstrates that evidence for such differences is…

Further analysis of data previously discussed in a journal article ("Episodic Instability in School District Elections," by L. Criswell and D. Mitchell, in the July, 1980 issue of "Urban Education") indicates that the success of a school board candidate in defeating an incumbent increases the probability that other school board…

Three linear models (Fixed, Random, and Provisional) and their inferential counterparts are presented and discussed. The three linear models are considered in relation to the analysis of nonlinearity, and limitations in the use of polynominal and product variables are discussed. Several exemplary research papers are examined and reconsidered in…

This paper's goal is to redefine fiscal health by broadening its predictive function and to determine which fiscal indicators are useful for forecasting fiscal health for one, two, and three years. Results indicate that school district fiscal health forecasts are potentially great planning tools for local for local decision makers. Includes 11…

Comments on Kerwin et al.'s article on covariance structure analysis (LISREL). Cautions that much effort must be made in developing the theory that describes the relationships among constructs and in establishing the measurement of these constructs before covariance structure is used. (ABB)