Question

For each of the examples below, decide whether to use a parametric or a nonparametric test and explain your choice.

1. Levy investigated sex differences in jealousy. Men and women were asked which they thought was worse: sexual infidelity or emotional infidelity. Levy found that of the 50 men interviewed, 35 thought that sexual infidelity was worse, and 15 thought that emotional infidelity was worse. Thirty of the 65 women thought that sexual infidelity was worse, and 35 thought that emotional infidelity was worse.

2. Handley investigated the persuasiveness of arguments. Arguments were rated on a scale from 1 (not at all persuasive) to 7 (extremely persuasive). Handley also counted how many facts were presented in each argument. She wants to know if the persuasiveness of arguments is related to the number of facts presented.

Answer 1

1. We can use parametric Z test for two proportions in this example because the conditions to apply the Z test for two proportions are satisfied that is:

1. Each sample contains at least 10 success and 10 failures.

2. The samples are independent

2. Since the variable Arguments is measured on an ordinal scale hence, we will have to use a nonparametric test to know if persuasiveness of arguments is related to the number of facts presented or not because the parametric methods require the assumption of normality which is not possible for an ordinal variable.

Good luck with your studies!!!