Probability and Statistics

7

Tests of Significance

7.1 Introduction

In Chapter 6, we have considered large sample statistical estimation and hypotheses testing techniques which depend on the central limit theorem to justify normality of the estimators and the test statistics. They apply when the samples are large (n ≥ 30).

Situations arise where large sampling is not possible. Consider, e.g., populations such as synthetic diamonds, satellites, aeroplanes, super computers and nuclear reactors which involve heavy expenditure. In such cases, the size of the sample is small (n < 30) and we have to consider statistical techniques that are suitable for them.

7.2 Test for One Mean (Small Sample)

7.2.1 Student's t-Distribution

While discussing sampling distribution, we noted ...