NotesFAQContact Us
Collection
Advanced
Search Tips
Peer reviewed Peer reviewed
ERIC Number: EJ540326
Record Type: Journal
Publication Date: 1996
Pages: N/A
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: ISSN-0149-7189
Using Latent Class Analysis To Detect Behavioral Patterns in Systems of Observational Variables.
Wessels, Holger; Von Eye, Alexander
Evaluation and Program Planning, v19 n4 p321-31 Nov 1996
A two-step procedure is proposed for detecting behavioral patterns in observational data in evaluation research. Latent class analysis is used to derive a classification of single observations at the variable level. The second step uses the probabilities that each observation falls into latent classes to derive behavioral preference scores. (SLD)
Publication Type: Reports - Evaluative; Journal Articles
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A