Vol. 2, Issue 7, Part J (2016)

Abstract
Data for one or more of factors levels in factorial experiment analyzed by a usual factorial ANOVA from a population whose distribution violates the assumptions of normality, the results on the original data may provide misleading results or it may not be the most powerful tests. In such cases transforming the data using proper data transformation technique or using non parametric tests may provide better estimation. In this paper, we analyzed two factor factorial experiments which violate assumptions of normality using usual parametric factorial ANOVA, various data transformation techniques applying to original data and non-parametric method (Aligned rank method). The results show significant differences for different tests. So, here we proposed analytical procedure of aligned rank transformation when data violates several assumptions of parametric ANOVA to avoid consequences in the datasets.