Module 9

Parametric tests

Statistical tests in which we make assumptions regarding the distribution of the population

Non-parametric tests

Statistical tests in which we do no make any assumption regarding distribution of the parameter under study. We use these types of tests:

• When data do not meet distributional assumptions

• There are outliers

• The data under analysis is ordinal (where the variable have natural, ordered categories and the distances between the categories is not known)

• When the hypothesis being tested does not concern a parameter

Population correlation hypothesis testing

This is when we test for correlation between population and if the population are related to one another or not.

What non-parametric do we use if we want to test the correlation of populations?

The Spearman Rank Correlation Coefficient

Test of independence using contingency table steps

We first calculate our expected values (using the expected frequency formula)

We than calculated the scaled squared deviation and add them all up to give us or X²

We than find our critical value using the contingency table (df = (columns - 1)(rows - 1))

Than based on the critical value, if X² is below critical value we fail to reject, but if its larger we reject the null hypothesis