Related Experiment Videos
A Macintosh BASIC program for fitting linear additive models to data by weighted least squares methods, with
1Biology Department, McGill University, Montreal, Quebec, Canada.
Computer Methods and Programs in Biomedicine
|April 1, 1993
Summary
This study introduces a BASIC program for fitting linear additive models using weighted or unweighted least squares. The program automates model refinement by removing redundant parameters, enhancing statistical analysis efficiency.
Area of Science:
- Computational statistics
- Statistical modeling
Background:
- Linear additive models are fundamental in statistical analysis.
- Efficient fitting methods are crucial for complex datasets.
Purpose of the Study:
- To describe a BASIC program for fitting linear additive models.
- To provide a tool for weighted/unweighted least squares analysis with matrix procedures.
Main Methods:
- The program utilizes matrix procedures for model fitting.
- It supports user-defined design matrices and orthogonal parameter options.
- Automated removal of redundant parameters is employed for weighted fits.
Main Results:
- The program calculates predicted values, regression, and residual terms.
- It successfully fits models for various analyses including ANOVA, regression, covariance, and genetic analyses.
- Sums of squares for parameters and orthogonal comparisons are computed.
Conclusions:
- The developed BASIC program offers a versatile and automated approach to linear additive model fitting.
- Its application extends to diverse statistical analyses, improving efficiency and accuracy.