Frances K. Stage

Reporting Structural Equation Modeling and Confirmatory Factor Analysis Results: A Review

ABSTRACT The authors provide a basic set of guidelines and recommendations for information that should be included in any manuscript that has confirmatory factor analysis or structural equation modeling as the primary statistical analysis technique. The authors provide an introduction to both techniques, along with sample analyses, recommendations for reporting, evaluation of articles in The Journal of Educational Research using these techniques, and concluding remarks.

Family and High School Experience Influences on the Postsecondary Educational Plans of Ninth-Grade Students:

The objectives of this study were to review the current literature on status attainment and student college choice and to develop and test a structural model of predisposition to attend college. Family and student background characteristics, parents’ educational expectations for students, level of student involvement in school, and student achievement were cited as influences on students’ predisposition toward postsecondary education and were the chief components of the model. Data from 2,497 ninth-grade students and their parents were used to test the model using LISREL. Parents’ expectations exerted the strongest influence throughout the model. Parents’ education, student gender, high school GPA, and high school experiences also contributed significantly in explaining students’ aspirations.

Influences on the Choice of Math/Science Major by Gender and Ethnicity

Women and minorities continue to be underrepresented in quantitatively based fields of study. Additionally, selection by all students of such majors is declining. Math/Science major choice is of concern in light of the occupational demands created by advancing technology as well as the potential gaps in occupational and economic attainment of women and minorities. This article reports the analysis of a longitudinal model of math/science major choice upon entrance to college for black and white, female and male students. The model was tested using a sample drawn from the “High School and Beyond” data base. The model included background characteristics of students, ability, and an array of high school experience factors to explain choice of quantitative major. Significant predictors of major choice for the subgroups included sophomore choice of major, mathematics attitudes, math and science completed by senior year, and various parental factors. However, there were differences across groups and the model explained nearly twice as much variance for the black male, black female, and white male subgroups compared with the white female subgroup. Recommendations include broadening our ways of researching migration into and out of the mathematics/science pipeline. Argument is made for a focus on success of students enrolled in low level college mathematics classes as a way of augmenting the mathematics/scientific pipeline.

THE USE AND INTERPRETATION OF LOGISTIC REGRESSION IN HIGHER EDUCATION JOURNALS: 1988-1999

This article examines the use and interpretation of logistic regression in three leading higher education research journals from 1988 to 1999. The journals were selected because of their emphasis on research, relevance to higher education issues, broad coverage of research topics, and reputable editorial policies. The term “logistic regression” encompasses logit modeling, probit modeling, and tobit modeling and the significance tests of their estimates. A total of 52 articles were identified as using logistic regression. Our review uncovered an increasingly sophisticated use of logistic regression for a wide range of topics. At the same time, there continues to be confusion over terminology. The sample sizes used did not always achieve a desired level of stability in the parameters estimated. Discussion of results in terms of delta-Ps and marginal probabilities was not always cautionary, according to definitions. The review is concluded with recommendations for journal editors and researchers in formulating appropriate editorial policies and practice for applying the versatile logistic regression technique and in communicating its results with readers of higher education research.

Path Analysis: An Introduction and Analysis of a Decade of Research

The authors review the use and interpretations of path analyses in articles published in The Journal of Educational Research from 1992 to 2002 and discuss related issues. This article provides (a) a brief introduction to path analysis, (b) suggested guidelines and recommendations for reporting results, (c) a sample of a model path analysis, (d) evaluation of the JER path analysis articles, and (e) concluding remarks.

Motivation, Academic and Social Integration, and the Early Dropout

This study employed a measure of Motivational Orientation as a blocking factor in an analysis of college withdrawal within the Tinto framework. To identify early dropouts, college records and surveys completed by first-year university students at the beginning of and 10 weeks into, the semester were used to operationalize key constructs within the Tinto model. The sample of 316 students was divided into three Motivational Orientation subgroups. The three largest, Certification, Cognitive, and Community Service, were used for analysis. Significant influential factors within the final empirical models differed among the three subgroups providing “persistencepatterns” that varied by motivational type.

Science Achievement Growth Trajectories: Understanding Factors Related to Gender and Racial–Ethnic Differences in Precollege Science Achievement

Using hierarchical linear modeling (HLM) and longitudinal data from the first three waves of the National Education Longitudinal Study (NELS:88), we examined achievement and growth rates in precollege science by racial–ethnic and gender subgroups. We found socioeconomic status and previous grades strongly and positively related to students’ eighth-grade achievement across all racial–ethnic by gender subgroups. We also found locus-of-control to be strongly related to eighth-grade science achievement for all subgroups except Asian American males. In modeling the growth rate, we found that the quantity of science units completed in high school was the only consistent predictor of science growth rates across all racial–ethnic by gender subgroups. The relationships between individual-level factors and science growth rates differed greatly for the remaining individual-level variables, highlighting the need for further research that both disaggregates data by race–ethnicity and gender.

University attrition: LISREL with logistic regression for the persistence criterion

This study demonstrates the use of logistic regression in conjunction with LISREL for a university attrition study. College records and surveys completed by first-year university students at the beginning of the semester and in the third month of the semester were used to operationalize key constructs within Tintos attrition model. The sample of 313 students was divided into male and female subgroups for analysis. Significant influential factors within the final empirical models varied between the two groups. LISREL was used to test hypothesized differences between empirical models. Because of the skewed distribution of the final criterion—persistence—logistic regression was used to identify significant predictors.

Reciprocal effects between the academic and social integration of college students

The academic and social integration of college students has been the focus of much recent research. The Tinto model of college student withdrawal focused on the constructs as explanation for commitment to goals and persistence. Research results suggested that perhaps reciprocal relationships existed between academic and social integration. Significant influences from academic integration to social integration and vice-versa indicated differing reciprocal effects for males and females.

Incompatible Goals: Narratives of Graduate Women in the Mathematics Pipeline

Despite 2 decades of research and attention directed at enhancing their participation, the number of women earning a PhD in science and mathematics remains low. This article presents the results of an interpretive study of the educational and career paths of female students who have successfully completed a bachelor’s or master’s degree in mathematics and have chosen to leave the mathematics/science pipeline to research scientist in order to pursue a doctorate in education. The women’s narratives portray the difference between their views of themselves and the nature of their chosen discipline. The emergent themes are examined against the conceptual framework of women’s participation in the mathematics/science pipeline and women and their relationship to mathematics.