50 interview questions with answers on Non-Parametric Tests

 Here are 50 interview questions with answers on Non-Parametric Tests in Research Methodology, ideal for academic, teaching, and research-based interviews.

 Basic Conceptual Questions

1.      What is a non-parametric test?
A non-parametric test is a statistical test that does not assume a specific distribution of the data, often used when assumptions of parametric tests are not met.

2.      When should you use a non-parametric test?
When the data is ordinal, not normally distributed, or when sample sizes are small or variances are unequal.

3.      What is the major difference between parametric and non-parametric tests?
Parametric tests assume a known distribution (e.g., normal), while non-parametric tests do not.

4.      Name some common non-parametric tests.
Mann–Whitney U test, Wilcoxon signed-rank test, Kruskal–Wallis test, Friedman test, Chi-square test, Spearman's rank correlation.

5.      Are non-parametric tests less powerful than parametric tests?
Generally, yes—when the parametric assumptions are met, parametric tests tend to be more powerful.

 Test-Specific Questions

6.      What is the Mann–Whitney U test?
It is used to compare differences between two independent groups when the dependent variable is ordinal or not normally distributed.

7.      What is the Wilcoxon signed-rank test?
It compares two related samples or matched pairs to assess whether their population mean ranks differ.

8.      What is the Kruskal–Wallis H test?
A non-parametric alternative to one-way ANOVA used for comparing more than two independent groups.

9.      What is the Friedman test?
A non-parametric test used to detect differences in treatments across multiple test attempts in related samples.

10.  What is the Chi-square test used for?
It tests relationships between categorical variables.

 Correlation and Association

11.  What is Spearman’s rank correlation?
It measures the strength and direction of association between two ranked variables.

12.  How is Spearman’s correlation different from Pearson’s correlation?
Spearman’s uses ranks and does not assume normality; Pearson’s uses raw values and assumes normal distribution.

13.  What is Kendall’s Tau?
A non-parametric statistic used to measure ordinal association between two measured quantities.

14.  What are the assumptions of Spearman's correlation?

·         Variables are ordinal, interval, or ratio

·         Monotonic relationship

·         Independent observations

15.  What does a Spearman’s rho value of 0.9 indicate?
A strong positive monotonic relationship between the two variables.

📐 Application-Based Questions

16.  Which test is used to compare two medians from independent groups?
Mann–Whitney U test.

17.  What test is used for before-and-after studies with non-normal data?
Wilcoxon signed-rank test.

18.  Which test compares more than two related samples?
Friedman test.

19.  Which test compares proportions between categorical variables?
Chi-square test of independence.

20.  What non-parametric test is suitable for testing one-sample median?
One-sample Wilcoxon signed-rank test.

📋 Interpretation and Reporting

21.  What does a significant result in a Kruskal–Wallis test mean?
At least one group median is different from the others.

22.  How do you interpret a Chi-square test result?
If p-value < α, there is a significant association between the variables.

23.  Can effect size be calculated for non-parametric tests?
Yes, using measures like r or eta-squared.

24.  Can you use box plots to visualize non-parametric test results?
Yes, especially useful for medians and spread comparisons.

25.  How do you report Mann–Whitney U test results in APA?
Example: U = 142.5, z = −2.31, p = .021, r = .34

🧠 Conceptual Deep-Dive

26.  Why are non-parametric tests called distribution-free tests?
Because they do not rely on assumptions about the underlying distribution.

27.  Do non-parametric tests require homogeneity of variances?
No, most do not require this assumption.

28.  What are the advantages of non-parametric tests?

·         Fewer assumptions

·         Suitable for small samples

·         Handle outliers better

29.  What are the disadvantages of non-parametric tests?

·         Less power

·         Cannot estimate parameters like mean

·         Harder to model complex relationships

30.  Can non-parametric tests handle ordinal data?
Yes, they are ideal for ordinal data.

 Practical/Software Use

31.  Which software is used to perform non-parametric tests?
SPSS, R, Python (scipy.stats), SAS, Stata

32.  How do you run the Mann–Whitney U test in SPSS?
Analyze → Nonparametric Tests → Legacy Dialogs → 2 Independent Samples

33.  How do you perform the Wilcoxon test in R?
wilcox.test(x, y, paired=TRUE)

34.  Which Python library supports non-parametric tests?
scipy.stats

35.  How do you visualize ordinal data for non-parametric testing?
With bar charts, boxplots, or rank plots.

📚 Theory & Design

36.  What is the null hypothesis in the Wilcoxon signed-rank test?
The median difference between the paired observations is zero.

37.  What is the null hypothesis in Kruskal–Wallis test?
The distributions of all groups are equal.

38.  How do you handle tied ranks in non-parametric tests?
Average the ranks of the tied observations.

39.  Why is ranking used in non-parametric tests?
It minimizes the influence of outliers and skewed distributions.

40.  What does it mean if a Chi-square test yields a p-value > 0.05?
The variables are not significantly associated.

🧩 Comparative & Decision-Making Questions

41.  Which is better: t-test or Mann–Whitney U test?
t-test is better if assumptions are met; otherwise, Mann–Whitney is safer.

42.  What to use if the data is ordinal and paired?
Wilcoxon signed-rank test.

43.  Which test is a non-parametric alternative to repeated measures ANOVA?
Friedman test.

44.  Can non-parametric tests be used in large samples?
Yes, especially when data distribution is unknown or violated.

45.  What is a limitation of Chi-square test?
It cannot be used when expected frequencies are too low (less than 5 in over 20% of cells).

🧾 Miscellaneous

46.  Are non-parametric tests valid for nominal data?
Yes, especially Chi-square tests.

47.  What is the minimum sample size for Mann–Whitney U test?
Typically, n ≥ 5 per group is recommended.

48.  How do you deal with outliers in non-parametric tests?
They are less affected, so no strict treatment is necessary.

49.  What is the advantage of using ranks?
It simplifies calculations and reduces the effect of extreme values.

50.  Can you use non-parametric tests in mixed-methods research?
Yes, for quantitative ordinal or non-normal data within the quantitative component.