Abstract
Nonsense coding systems can be constructed that retain outcomes regarding R^2 values, F values and multiple comparison tests. Nonsense coding highlights the flexibility of coding ANOVA problems to be analyzed by multiple linear regression procedures; however, no additional analytic power appears to be gained from their use.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Copyright (c) 1987 John D. Williams (Author)
Downloads
Download data is not yet available.