학술논문

Higher-Order Factor Analysis of Data from the Problems of Teaching Survey.
Document Type
Reports - Research
Speeches/Meeting Papers
Source
Subject
Correlation
Elementary School Teachers
Elementary Secondary Education
Factor Analysis
Likert Scales
Matrices
Problems
Secondary School Teachers
Surveys
Teacher Attitudes
Language
English
Abstract
Many researchers recognize that factors can be rotated obliquely, yielding factors that are correlated with each other. However, not as many researchers realize that this interfactor correlation matrix can then also be factor analyzed. This "higher-order" factor analysis has many useful applications and, in fact, some researchers suggest that higher-order analyses should always be conducted whenever first-order factors are rotated obliquely. Results of a survey of teachers from Louisiana and Mississippi (n=291) as to self-perceived teacher problems are analyzed with SECONDOR, a FORTRAN computer program that computes various first-order principal components and then computes second-order factor solutions. The results of this analysis provide informative insights into the relationships underlying responses. Six tables, one figure. (Contains 15 references.) (Author)