An Alternative Numerical Method for Finding the Value of Transformation Parameter in the Box-Cox Transformation

To analyze non normal data, the data should be transformed to a normal distribution. The well-known Box-Cox transformation can be used to get the normality. Since the transformation parameter is usually unknown, many statisticians studied the methods of estimation of the transformation parameter, lambda , appearing in the Box-Cox transformation. Moreover, we can use the numerical methods for single variable function optimization to find the value of the transformation parameter lambda such as bisection method, Newton's method, secant method, and others. The alternative numerical method lambda is applied to find the suitable value of lambda. It is better than the bisection method in the context of the number of function evaluations and the processing time.