Hypothesis Testing

This chapter discusses the concept of hypothesis, its importance in statistical analysis and the process of testing the hypothesis. Since hypothesis is formulated in the backdrop of the distribution of the sample/data, we first discuss different types of distributions. Depending upon the nature of hypothesis, the distributions could be discrete or continuous. We first discuss the continuous distributions, and the commonest of the continuous distributions, the normal distribution in detail. We also give a quick overview of t- and F-distributions, also continuous in nature, and commonly used. We then explain the concept of binomial distribution, the commonest of the discrete distributions. We then explain by examples how a typical statistical hypothesis is formulated. We then explain the key terminologies associated with hypothesis testing, and how each of these influences the rejection (or otherwise) of the null hypothesis. We end the chapter with an example of hypothesis testing for an independent sample t-test.