Related Experiment Videos
Statistical analysis of repeated measures data using SAS procedures
R C Littell1, P R Henry, C B Ammerman
1Department of Statistics, University of Florida, Gainesville 32611-0339, USA.
Journal of Animal Science
|May 15, 1998
Summary
Mixed linear models, initially for animal breeding, now benefit diverse research. PROC MIXED in SAS enables accurate analysis of random effects and covariance structures, crucial for repeated measures data.
Area of Science:
- Statistics
- Animal Breeding
- Data Analysis
Background:
- Mixed linear models originated in animal breeding for genetic potential evaluation.
- Their application has expanded across research fields due to advanced software.
- Traditional methods using fixed-effect approaches limited covariance structure modeling.
Purpose of the Study:
- To highlight the advancements in mixed model analysis with PROC MIXED.
- To explain the importance of modeling covariance structures, especially for repeated measures.
- To demonstrate how PROC MIXED overcomes limitations of older methods like PROC GLM.
Main Methods:
- Utilizing PROC MIXED within the SAS System for statistical analysis.
- Implementing models that incorporate random effects.
- Explicitly modeling the covariance structure of the data.
Main Results:
- PROC MIXED allows for efficient estimation of fixed effects.
- Valid standard errors for these estimates can be computed.
- The procedure effectively handles complex data structures, including repeated measures.
Conclusions:
- PROC MIXED offers a significant improvement over older methods for mixed model analysis.
- Accurate modeling of covariance structures is essential for robust statistical inference.
- This advancement is particularly beneficial for analyzing correlated data, such as longitudinal studies.