UJI KRUSKAL WALLIS | NON PARAMETRIK MATERI | OLAH SPSS | INTEPRETASI

Summary

In this video, the speaker demonstrates how to conduct the Kruskal-Wallis test in SPSS, a non-parametric alternative to ANOVA. The video covers essential steps, including testing for normality using Shapiro-Wilk and Kolmogorov-Smirnov tests, checking for homogeneity of variances, and interpreting results. A practical case study is provided, examining the effects of three chemical treatments on plant growth. The video guides users through data entry in SPSS, assumption testing, and performing the Kruskal-Wallis test, offering clear instructions and interpretations for analyzing the impact of different treatments on plant growth.

Takeaways

• 😀 Kruskal-Wallis test is used when normality and homogeneity of variances assumptions are not met, allowing comparison between multiple groups.
• 😀 The normality test can be conducted using the Shapiro-Wilk or Kolmogorov-Smirnov tests, where a p-value less than 0.05 indicates the data is not normally distributed.
• 😀 Homogeneity of variance test checks if the variance between groups is equal. If the p-value is less than 0.05, variances are considered unequal.
• 😀 If normality or homogeneity assumptions are violated, the Kruskal-Wallis test is a suitable non-parametric alternative to ANOVA.
• 😀 The Kruskal-Wallis test assesses if different treatments or conditions result in significantly different outcomes in the data.
• 😀 In SPSS, data for the Kruskal-Wallis test should be entered in a format with categorical variables (e.g., treatment type) and scale variables (e.g., growth data).
• 😀 To perform a normality test in SPSS, use 'Analyze' → 'Descriptive Statistics' → 'Explore', and select 'Normality Plot with Tests' for a clear test of normality.
• 😀 If the p-value for the normality test is less than 0.05, the data is considered not normally distributed, and further non-parametric tests like Kruskal-Wallis are needed.
• 😀 In the homogeneity of variance test in SPSS, the significance value (Sig.) less than 0.05 indicates that variances between the groups are unequal.
• 😀 The Kruskal-Wallis test results in a p-value that, if less than 0.05, suggests that different treatments (e.g., chemicals) significantly affect the outcome, such as plant growth.

Q & A

What is the Kruskal-Wallis test used for?

The Kruskal-Wallis test is used to determine whether there are statistically significant differences between three or more independent groups on a non-normally distributed dependent variable.

When should you use the Kruskal-Wallis test instead of ANOVA?

You should use the Kruskal-Wallis test when the assumptions for ANOVA, such as normality of data and homogeneity of variances, are not met. If the data is not normally distributed or the variances across groups are unequal, Kruskal-Wallis is the appropriate alternative.

What are the first two assumptions you need to check before performing a Kruskal-Wallis test?

The first two assumptions to check are: 1) Normality of the data (using tests like Shapiro-Wilk or Kolmogorov-Smirnov), and 2) Homogeneity of variances across groups (using tests like Levene's test).

What does a p-value of less than 0.05 indicate in normality tests?

A p-value of less than 0.05 in normality tests, such as Shapiro-Wilk or Kolmogorov-Smirnov, indicates that the data is not normally distributed.

How do you check the homogeneity of variances in SPSS?

In SPSS, you can check homogeneity of variances by conducting a test in the General Linear Model (GLM) analysis. After selecting your dependent and factor variables, you can click on 'Options' and check the 'Homogeneity of Variance Test' box.

What do you do if the assumption of homogeneity of variances is violated?

If the assumption of homogeneity of variances is violated, you can use non-parametric tests like the Kruskal-Wallis test instead of ANOVA.

What does the Kruskal-Wallis test assess in the context of plant growth experiments?

In the context of plant growth experiments, the Kruskal-Wallis test assesses whether there are significant differences in the growth of plants under different chemical treatments (A, B, C).

How can you interpret the output of a Kruskal-Wallis test in SPSS?

In SPSS, the output of the Kruskal-Wallis test includes a test statistic (H) and a p-value. If the p-value is less than 0.05, it indicates a significant difference in the treatment effects. If the p-value is greater than 0.05, there is no significant difference.

What should you do if the Kruskal-Wallis test shows a significant result?

If the Kruskal-Wallis test shows a significant result (p-value < 0.05), you should perform post-hoc tests to identify which specific treatment groups differ from each other.

Can you perform a Kruskal-Wallis test if the data is normally distributed?

While the Kruskal-Wallis test is useful when data is not normally distributed, you can perform it even if the data is normally distributed, but it is generally more appropriate when the assumptions of ANOVA are violated.