The **Box-Cox transformation** is a statistical method used to transform non-normally distributed data into a normal distribution. Many statistical procedures (like ANOVA or linear regression) assume that the underlying data is normally distributed.

By finding an appropriate power parameter λ, the transformation:

• Reduces skewness to make the distribution symmetric.
• Stabilizes variance across groups or ranges (reduces heteroscedasticity).
• Improves the validity of parametric statistical tests.

Depending on the optimal λ value, the transformation corresponds to standard math functions:

• λ = 1.0: No transformation (linear shift only).
• λ = 0.5: Square root transformation (useful for Poisson-distributed count data).
• λ = 0.0: Natural logarithm transformation (extremely common for exponential growth or right-skewed data).
• λ = -1.0: Reciprocal transformation.

Note: Box-Cox can only be applied to positive values (x > 0). For datasets containing zero or negative values, the Yeo-Johnson transformation is typically used instead.