NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED449224
Record Type: Non-Journal
Publication Date: 2000-Nov
Pages: 19
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
Toward Reflective Judgment in Exploratory Factor Analysis Decisions: Determining the Extraction Method and Number of Factors To Retain.
Knight, Jennifer L.
This paper considers some decisions that must be made by the researcher conducting an exploratory factor analysis. The primary purpose is to aid the researcher in making informed decisions during the factor analysis instead of relying on defaults in statistical programs or traditions of previous researchers. Three decision areas are addressed. First, the importance of determining the number of participants and variables is discussed. Second, the similarities and differences in the two most common factor extraction methods, principal components analysis (PCA) and principle factor analysis (PFA), are addressed. Finally, the paper compares different methods for retaining factors (e.g., K1 rule, scree plot, parallel analysis, and others). Overall, although PCA and PFA results converge with an increasing number of variables and communality estimates, PFA may be the better choice for attempting to determine underlying factors in a data set. Furthermore, parallel analysis shows great promise for facilitating precision in factor retention, but a researcher should take into consideration a combination of methods in order to be as accurate as possible in developing the factor model. Empirical studies indicate that defaults in statistical packages are frequently used and that superior methods, such as parallel analysis, are grossly underused. (Contains 18 references.) (Author/SLD)
Publication Type: Reports - Descriptive; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A