Related Experiment Videos
Choice of structural model via parsimony: a rationale based on precision
Psychological Bulletin
|September 1, 1989
Summary
In large samples, choosing the simpler of two nested models improves estimation precision. Parsimony in model selection directly enhances the accuracy of estimated parameters.
Area of Science:
- Statistical modeling
- Econometrics
- Model selection
Background:
- Nested models are frequently compared in statistical analysis.
- Model complexity can influence the precision of parameter estimation.
Purpose of the Study:
- To demonstrate that parsimony in nested models enhances parameter estimation precision.
- To provide a statistical justification for using parsimony as a model selection criterion.
Main Methods:
- Theoretical analysis of nested statistical models.
- Examination of sampling variance for common parameters under competing models.
- Large sample theory application.
Main Results:
- The more parsimonious nested model yields an estimator with smaller sampling variance.
- Parsimony is directly linked to increased precision in estimating common parameters.
Conclusions:
- Parsimony serves as a valid criterion for selecting between statistically acceptable nested models.
- Choosing a parsimonious model leads to more precise estimation of shared parameters.