Lee M. Wolfle

The Three-Stratum Theory of Cognitive Abilities: Test of the Structure of Intelligence across the Life Span.

Abstract Recently a three-stratum theory of intelligence that combines the major aspects of Spearmans (1927) theory of general intelligence (g) and Horn and Cattells (1966) theory of crystallized and fluid intelligence (Gf-Gc) has been proposed (Carroll, 1993a). The purpose of this study was to test the three-stratum theory using hierarchical confirmatory factor analysis with the LISREL computer program. Developmental changes in the structure of intelligence were also investigated. Results provided support for the three-stratum theory and suggested the possibility of intermediate factors between the second and third strata but did not support developmental changes in the organization of cognitive abilities over the life span.

A Structural Model of Mathematics Achievement for Men and Women

An extensive body of research indicates that men on the average achieve higher scores in mathematics than women. This paper addresses the issue of how this difference develops by estimating a latent-construct causal model of the process of mathematics achievement. When the model was compared between men and women, we found that the process of mathematics achievement differs. In particular, we found that mathematics ability and attitudes toward mathematics had stronger effects on mathematics achievement for men than for women. The interactions between sex and the variables in the model indicate that the process for men and women is not simply additive, and may be more complicated than previous researchers have assumed.

Strategies of Path Analysis

Path analysis is a method for explicitly formulating theory, and attaching quantitative estimates to causal effects thought to exist on a priori grounds. There are four basic kinds of path models: recursive, block, block-recursive, and nonrecursive. Because some questions can be answered only under certain path analytic structures, the theoretical questions and research goals guide the selection of a particular model. Knowing what sorts of questions each model addresses, educational researchers will be in a better position to formulate hypotheses in model-specific terms. Four illustrations are provided, presenting strategies of analysis typical of each kind of model.

Postsecondary Educational Attainment Among Whites and Blacks

Previous research has indicated that social background variables are more important determinants of educational attainment among whites than among blacks. The present study, based on more recent data and methods that control for estimated measurement error structures, finds that social background plays a similar role for whites and blacks. Increments in background social status variables lead to similar increases in educational attainment for whites and blacks. Moreover, the effects of personal characteristic variables (ability, curriculum, grades) of whites and blacks as they influence educational attainment are also similar for both groups.

Testing a Model of Teacher Satisfaction for Blacks and Whites

The purpose of this study was to construct and test a model of the influences affecting early-career teachers’ job satisfaction. The path model is presented and tested first with a sample of black teachers and then with a sample of white teachers. Although the model provides a plausible explanation for the relationships among the variables for both groups, differences in the structural coefficients indicate the process leading to teacher job satisfaction is different for blacks and whites. The implications of this study suggest it may be critical to take into account factors such as race and sex in teacher induction and training.

The Introduction of Path Analysis to the Social Sciences, and Some Emergent Themes: An Annotated Bibliography

This is the basic theorem (Duncan, 1966, p. 5) or first law (Kenny, 1979, p. 28) of path analysis developed by Sewall Wright over 80 years ago (see Wolfle, 1999, for an annotated review of Wright’s work). Indeed, the basic theorem applies to any set of structural equations, and when there are more equations than parameters to be estimated, the zero-order measures of covariation can be reproduced approximately. In carefully controlled, plausibly defended sets of equations, the various components of the reconstruction of zero-order associations can be considered direct causal effects, indirect causal effects, and noncausal or spurious components. Sewall Wright called this “path analysis.” Years later, primarily in the field of sociology, Blalock (1962, 1964) struggled with methods of analyzing causal relations in social data, approaching the subject with reference to vanishing partial correlations as tests of causal assumptions. Taking up Blalock’s work, Duncan (see Duncan, 1998) remembered a lecture by William Ogburn that had made reference to Burks’s (1928) examination of intelligence. Burks’s work relied heavily on that of Wright, and Duncan retrieved those

Enduring Cognitive Effects of Public and Private Schools

Recent research by Coleman and his associates on the effectiveness of public and private schools has evoked a heated debate on the role of private schooling in U.S. education. All of the existing studies, however, have examined the short-term effects of school type–at most over a 2-year period. This study extends the debate on the effects of private schools to a longer term perspective. It appears from the present analysis that the enduring influence of school type on vocabulary achievement is nearly nonexistent, whereas the enduring influence on mathematics favors public school students. These results conflict with those of previous studies, and may be due to the longer term perspective of the research reported here. Examination of the enduring effects of private education should therefore be given a high priority in subsequent follow-ups of the High School and Beyond (HSB) study. Until that happens, it seems clear that any policy decisions based on HSB analyses to date are premature.

Sewall Wright on the Method of Path Coefficients: An Annotated Bibliography.

When Dudley Duncan and others became interested in linear causal models, they discovered a genetic biologist had laid down the groundwork for them over nearly a 50‐year period. This annotated bibliography covers Sewall Wrights development of the method of path coefficients.

A latent-variable causal model of faculty reputational ratings

Saunier (1985), in an attempt to explain sources of variation in the NRC reputational ratings of university faculty, conducted a stepwise regression analysis using twelve predictor variables. Due to problems with multicollinearity and because of the atheoretical nature of stepwise regression, Sauniers conclusions were only speculative. By using LISREL the present reanalysis demonstrates the value of regressing reputational ratings on three latent variables: size, faculty research productivity, and the quality of program graduates. The model was tested using NRC data for each of six disciplines: English, French, philosophy, geography, political science, and sociology. The relative magnitude of the contributions made by the three latent variables depended upon the discipline of interest, but generally size and research productivity were found to be more important than the quality of the program graduates.

Effects of higher education on achievement for blacks and whites

This study assesses the impact of college on knowledge of vocabulary and mathematics. Unlike many previous studies, this one is able to control for precollege abilities in these academic areas, which presumably tend to select people into or out of college. The results indicate that college attendance has a strong impact on improved performance on these tests, and that these effects are similar for whites and blacks.