Teaching Hypothesis Test in an Easy Way

Abstract

This paper introduced an easy way for selecting null (H0) and alternative hypotheses (H1), finding p-values and taking decisions. Should a research hypothesis or a claim be H0 or H1? This paper proposed a new method to select H0 and H1. Considering two sets of hypotheses (a) H0: θ ≤ θ0 against H1: θ > θ0 and (b) H0: θ ≥ θ0 against H1: θ < θ0 for testing a location parameter θ if the absolute value of the test statistic obtained from a sample is less than the absolute critical value then the two decisions corresponding to two sets of hypotheses (a) and (b) are contradictory. This paper proposed new methods to overcome these issues and making consistent decision. This paper also introduced a general formula for finding P-value. These new methods and formula have been applied in an undergraduate class. It is observed that the performances of proposed methods are significantly better than that of the traditional methods.