NotesFAQContact Us
Search Tips
Peer reviewed Peer reviewed
Direct linkDirect link
ERIC Number: EJ1046311
Record Type: Journal
Publication Date: 2014
Pages: 5
Abstractor: ERIC
ISSN: ISSN-1536-6367
Causal Indicator Models: Unresolved Issues of Construction and Evaluation
West, Stephen G.; Grimm, Kevin J.
Measurement: Interdisciplinary Research and Perspectives, v12 n4 p160-164 2014
These authors agree with Bainter and Bollen that causal effects represents a useful measurement structure in some applications. The structure of the science of the measurement problem should determine the model; the measurement model should not determine the science. They also applaud Bainter and Bollen's important reminder that the full measurement model needs to be fully tested even if each individual component shows adequate fit, an admonition that also applies to approaches that validate traditional measurement models using other constructs (e.g., Lengua, West, & Sandler, 2008). At the same time, they believe that issues of test construction associated with defining, refining, and validating causal effect constructs have not been fully addressed; and they worry that the theoretical advantages of Bainter and Bollen's causal effects model may not be worth the confusion and biased estimates that can occur under misspecification. These authors speculate that the use of the weighted composite model to represent the causal effects structure may be less sensitive to misspecification and serve as a better measurement model in practice.
Psychology Press. Available from: Taylor & Francis, Ltd. 325 Chestnut Street Suite 800, Philadelphia, PA 19106. Tel: 800-354-1420; Fax: 215-625-2940; Web site:
Publication Type: Journal Articles; Opinion Papers; Reports - Evaluative
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Grant or Contract Numbers: N/A