How to do the test:

- State your Null Hypothesis in the form:

"There is no difference between the populations from which the two samples come" - Choose a Critical Significance Level (α: alpha)

This is typically α = 0.05 - Calculate the statistic

Use a spreadsheet to calculate your results,**or**

Access a helpsheet for SPSS

and check here for data entry into SPSS - Reject or accept your Null Hypothesis

**Either**compare the calculated statistic with Critical Value,

CASE A - using critical values of the H statistic

If H ≥ H_{critical}→ reject H_{0}→ significant result

If H < H_{critical}→ accept H_{0}→ non-significant result

CASE B - comparing H with critical values of the Χ^{2}statistic for large amounts of data

If H ≥ Χ^{2}_{critical}→ reject H_{0}→ significant result

If H < Χ^{2}_{critical}→ accept H_{0}→ non-significant result

**or**

Look at the calculated probability (*P*value)

(SPSS calculates the X^{2}value)

If*P*≤ α → reject H_{0}→ significant result

If*P*> α → accept H_{0}→ non-significant result - Report your results

Plot your data using the appropriate graphs

**Either**

(Kruskall-Wallis: H = …, df = …,*P*= …)

**or**

(Kruskall-Wallis: Χ^{2}= …, df = …,*P*= …)

View an example report