The pbjective of the transformation was to promote symmetry and linearity relationships. It deals with the non-normality of data and the non-homogenous nature of variance.
The power of transformations form the following ladder – x-2, x-1, x-1/2, log x, x1/2, x, x2, x3, and so on where x>1.
For example, when a square root transformation is used on poisson data, the approximate transformation would result in a data that follows a normal distribution.
Related Business Content