Analysis showed that a word problem is difficult to solve if: it is of a different type from the problem that preceded it; its solution requires a large number of different operations; its surface structure is complex; it has a large number of words; or it requires a conversion of units. (Authors)

Article is concerned with a model for school system evaluation. The usefulness of the indices from this model depend on their stability, and this study presents evidence related to their stability when pupils and factors related to time are considered as sources of error. (Author/RK)

This study attempts to explain the observed distribution of students among university departments. A general model describing the student's initial choice of department is constructed. The decisions made by a sample of students were then explained by multiple regression analysis, where students' socio-economic characteristics are the explanatory…

Describes the presence of high correlations between pooled scores on two test formats when administered to 170 students in an introductory physics course. Concludes the gradability of machine scorers is comparable to that of manual grading in this repeated-measures design. (CC)

The hypothesis that it is unwise to change answers to multiple choice questions was tested using multiple regression analysis. The hypothesis was rejected as results showed that there are gains to be made by changing responses. (Author/CK)

Similarities between regression analysis and analysis of variance are noted and it is argued that the former has advantages over the latter. It is also argued that canonical correlation analysis is more suitable than factor analysis in certain cases. The argument is illustrated with four recent pieces of educational research. (DG)

Organizational context, organizational complexity, and bureaucratic control in 50 Japanese industrial manufacturing plants are analyzed. Highly dependent organizations rely on centralization and formalization to maintain administrative control. The significant relationship between formalization and centralization indicates that they are not…

The authors isolate seven patterns of cultural choice using a cluster analysis of a multidimensional scale of items from a 1977 survey of leisure activities in the South. They then use regression analysis to isolate the social demographic correlates of the clusters. (Author/RM)

Fifty-six students from two Minnesota technical institutes were administered the Minnesota Reading Assessment and the Gates-MacGinitie Reading Tests. Statistical analysis indicates that, while the two tests are not identical measures of vocabulary and comprehension, their measures of reading rate and speed are highly correlated. (SK)

Estimates of the criterion-related validity for the National Teacher Examinations (NTE) in mathematics and science were determined empirically. Results suggest that the science tests are not as useful as the mathematics tests as indicators of subject knowledge. (GK)

Examines the credibility of inferences based on cross-group comparisons of regression coefficients using Black-White differences in the educational attainment process as an example. Shows that reanalyses of data from previous studies fail to produce consistent results even when major differences in samples, models, and methods are held constant.…

Sociological and demographic research often uses variables computed as ratios. When the denominators are highly correlated, and the ratios are used in correlation or regression analysis, a statistical dependency is formed. This article investigates this problem, particularly with respect to partial correlation, multiple regression, and the…

Two methods of analyzing nonorthogonal (uneven cell sizes) cross-classified data sets are compared. The methods are direct standardization and the general linear model. The authors illustrate when direct standardization may be a desirable method of analysis. (JKS)