Sample size requirements in structural equation models under standard conditions |
| |
Authors: | Andreas I Nicolaou Michael M Masoner |
| |
Institution: | 1. Bowling Green State University, United States;2. University of Cyprus, visiting 2003–04;3. Emeritus, Southern Illinois University at Carbondale, United States |
| |
Abstract: | When a research of risk considers researchers who wish to utilize structured equation modeling (SEM), all users would ask for their sample size. The paper argues for the formulation of a single barebones minimum standard to be accompanied by a hierarchy of higher quality standards. The paper also offers a specific recommendation for such a barebones standard and ranks quality standards by their sample size cost. In sum, the solution to this problem involves integration: over the multitude of standards, over both single-study and multi-study perspectives, and over a broad array of research environments. Our solution deals with the multitude of solutions. It seeks maximum flexibility to accommodate a multi-study perspective, and it is sensitive to the needs of research settings where sample size is constrained or collection is costly. For these reasons we offer important suggestions and summarize recommendations. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|