Question

Answer the following questions about the entire module.

A.

How do you calculate degrees of freedom for goodness-of-fit chi-square tests versus chi-square tests for independence and homogeneity

B. How are the expected counts calculated manually for goodness-of-fit chi-square tests versus chi-square tests for independence and homogeneity?

C. What conditions must be checked for any of the three types of chi-square tests?

Answer 1

A. For Chi square goodness of fit test:

df = (no of rows - 1)

For Chi square test of Independence, test for homogeneity:

df = (no of rows-1)*(no of columns-1)

B. For goodness of fit test we have a set of observed values and the expected percentages that we consider the distribution to follow. So expected value: Total observed value* percentage for each row

For Chi square test of Independence or homogeneity, we have : Expected values =

(row total of observed value)*(coulmn total)/table total

C. Conditions that must be checked:

1.In each cell of the tables we have frequency values. And each cell frequency must be > 5

2. Each of the cell frequency must be independent of the others

3. All the events in the cross tabulation of the tables must be mutually exclusive.