What is the purpose of applying a transformation to data?

• A. To correct distributional abnormality such as skew or kurtosis.
• B. To increase sample size
• C. To change the unit of measurement
• D. To test the null hypothesis more easily

Answer: (A)

Applying a transformation to data is mainly about making the data behave more nicely for analysis by addressing distributional issues. Transformations such as the log, square root, or Box-Cox can reduce skew and excess kurtosis and stabilize variance across values. This helps meet the common assumptions of many parametric tests, like normality and homoscedasticity, and can also help linearize relationships in regression. That’s why the primary purpose is to correct distributional abnormality such as skew or kurtosis. Transformations don’t increase sample size, and while they may change the scale of measurement, the core aim is not changing units but making the data suitable for valid statistical inference.