Work

Interpretational Confounding or Confounded Interpretations of Causal Indicators?

Year: 2014

Type: article

Abstract: In measurement theory causal indicators are controversial and little-understood. Methodological disagreement concerning causal indicators has centered on the question of whether causal indicators are inherently sensitive to interpretational confounding, which occurs when the empirical meaning of a latent construct departs from the meaning intended by a researcher. This article questions the validity of evidence used to claim that causal indicators are inherently susceptible to interpretational confounding. Further, a simulation study demonstrates that causal indicator coefficients are stable across correctly-specified models. Determining the suitability of causal indicators has implications for the way we conceptualize measurement and build and evaluate measurement models.

Source: Measurement Interdisciplinary Research and Perspectives

DOI: 10.1080/15366367.2014.968503

Authors: Sierra A. Bainter, Kenneth A. Bollen

Institution: University of North Carolina at Chapel Hill

Language: English

Topic: Psychometric Methodologies and Testing

Subfield: Management Science and Operations Research

Field: Decision Sciences

Domain: Social Sciences

Open Access status: green

Funder: National Institute on Drug Abuse

Awards: A novel application of Bayesian methods for modeling substance use trajectories, Predoctoral Training in Addiction Science