Data Transformations Notes.docx - Data Transformations Natural Logarithm Transformations When lack of linearity is the only condition that is violated

Data Transformations Natural Logarithm Transformations  When lack of linearity is the only condition that is violated, transforming the x values is appropriate  Transforming y values appropriate when non-normality and/or unequal variances are the problems – transforming y can also help smoothen out a curved relationship  Transform both x and y values when everything seems wrong – the data is not linear, not normal, and does not have equal variance  If the primary problem with the model is non-linearity, look at a scatterplot of the data to help