In investigating mediating processes, researchers usually use randomized experiments and linear regression or structural equation modeling to determine if the treatment affects the hypothesized mediator and if the mediator affects the targeted outcome. However, randomizing the treatment will not yield accurate causal path estimates unless certain assumptions are satisfied.
Download count: 0
- Partial requirement for: Ph. D., Arizona State University, 2013Note typethesis
- Includes bibliographical references (p. 123-129)Note typebibliography
- Field of study: Psychology