Final answer:
You would not choose to transform the data in situations where the data set already has a normal distribution or when meaningful outliers should not be removed.
Step-by-step explanation:
A situation where you would not choose to transform the data is when the data set already has a normal distribution or when the outliers are meaningful and should not be removed.
If the data set has a symmetrical distribution and there are no extreme values that would affect the analysis, there would be no need for a data transformation.
Furthermore, if the outliers in the data set are important for the analysis or represent unique cases, it would not be appropriate to transform the data and remove those outliers.