Homoscedasticity - Reasons for Screening Data

Homoscedasticity is the assumption that the variability in scores for a continuous variable is roughly the same at all values of another continuous variable.

1. In the bivariate case, this is referred to as homogeneity of variances. Usually the Leven's test is the tool to assess the homogeneity of variances. This test is used to assess the hypothesis that assumes samples of observations come from populations from the same variances. Therefore rejecting it would imply heterogeneity of variances.

2. In multivariate analysis this is referred to Homoscedasticity. Homoscedasticity is related to the assumption of multivariate normality. Therefore bivariate scatterplots could be used to detect heteroscedasticity. Heteroscedastic relationship could also mean that one of the variables in the group of variables to be analyzed has a relationship with the transformation of the other variable.