Hypothesis Testing and
Tests of Differences
Learning OutcomesAfter reading this chapter you will be able to:
• Choose an appropriate test of significance—monadic (single sample) or
comparative (two sample), Z-test (large sample) or f-test (small sample).
• Calculate the value of the test statistic. • Decide on the significance of the test performed. • Draw a conclusion and make information-based marketing
This chapter begins with the considerations we have to take into account in conducting a test of significance when determining the differences in means or proportions between two parameters. It then deals with a monadic test of significance for a single large sample and a comparative test for two large samples. The concept of degrees of freedom is briefly explained. This is followed by a f-test of significance for a single small sample. The f-test for the significance of sample values for two independent samples is demonstrated next. The importance of using a comparative f-test for estimates using the same respondents (paired f-test) is demonstrated. You will find Figure 9.1 useful showing how you can choose an appropriate test.
In attempting to choose a particular significance test, at least four questions should be considered:
1. Has the selection of the sample been done using probabilistic methods? It is necessary that respondents are chosen with a probabilistic method for conducting a test of significance as the test uses the standard errors of the estimates.