No CrossRef data available.
Article contents
Stratification, covariance and weight gain in experiments with repeated observations
Published online by Cambridge University Press: 27 March 2009
Summary
The use of weight gain in long-term experiments can produce serious bias in treatment comparisons unless the initial weights of treatment groups are equal or nearly so. When observations are repeated on the same experimental unit, covariance analysis of the data for each occasion produces treatment comparisons adjusted by the regression on the covariate at that time; polynomial trends in these adjusted comparisons with time can be easily calculated even though the regression on the covariate varies with time.
- Type
- Research Article
- Information
- Copyright
- Copyright © Cambridge University Press 1980