Data transformation is typically used to?

• A. It increases sample size
• B. It changes endpoints
• C. It helps meet model assumptions for statistical tests
• D. It replaces missing data

Answer: (C)

Data transformation is used to satisfy the assumptions behind many statistical analyses, especially parametric tests. When data are skewed or variances differ across groups, the residuals may not be normally distributed or the relationship may not be linear. Transforming the data (for example, using a log or square-root) can stabilize variance, normalize distributions, and linearize relationships, making the results from these tests more valid. After analysis on the transformed scale, you can interpret or back-transform the results as needed. This approach does not increase sample size, change the study endpoints, or replace missing data.