Related Experiment Videos
The importance of considering alternative structural equation models in evaluation research
Evaluation & the Health Professions
|February 9, 1987
Summary
Structural equation modeling (SEM) software like LISREL aids program evaluation. Reanalyzing data shows alternative SEM models are crucial for robust evaluation research findings.
Area of Science:
- Social Sciences
- Program Evaluation
- Psychometrics
Background:
- Structural Equation Modeling (SEM) software, including LISREL, has increased accessibility for program evaluation.
- A previous study (Sheehan et al., 1985) utilized LISREL to examine the link between residents' moral reasoning and patient care.
- The availability of advanced statistical software facilitates complex data analysis in research.
Purpose of the Study:
- To reanalyze existing data from a study on family medicine residents.
- To demonstrate the significance of exploring alternative structural equation models.
- To highlight best practices in evaluation research methodology.
Main Methods:
- Reanalysis of a dataset previously used in a 1985 study.
- Application of structural equation modeling techniques.
- Comparison of results from alternative model specifications.
Main Results:
- Alternative structural equation models can yield different interpretations of the same data.
- The choice of model significantly impacts the understanding of relationships between variables.
- Initial findings suggest the importance of model fit and theoretical grounding.
Conclusions:
- Researchers should consider and test multiple structural equation models.
- Methodological choices in SEM can influence evaluation outcomes.
- Emphasizes the need for critical assessment of model selection in program evaluation.