| Article ID: | iaor20011589 |
| Country: | Cuba |
| Volume: | 21 |
| Issue: | 2 |
| Start Page Number: | 113 |
| End Page Number: | 120 |
| Publication Date: | May 2000 |
| Journal: | Revista de Investigacin Operacional |
| Authors: | Vega Vivian Sistachs, Prez Nidia Hernndez, Barroso Elina Miret |
Additivity, variance, homogeneity and normality of the errors are often violated in the ANOVA of experimental design. Considering the results in the literature, we propose a general strategy in statistical analysis of an experimental design. This strategy may select the adequate transformation and model when the basic model assumptions are violate.