PN 3002 EXAM 6 2026 ACTUAL PAPER
QUESTIONS WITH SOLUTIONS GRADED A+
◉ What are the four non parametric tests?
Answer: 1. Wilcoxon rank sum test/ Mann whitney test
2. Wilcoxon signed rank test
3. Kruskal Wallis test
4. Friedman ANOVA
◉ What is the type of design (independent or repeated) for a
wilcoxon rank sum test/mann whitney ?
How many groups can it test for?
Answer: Independent groups design
It can test for 2 group only (e.g saline vs drug)
◉ What is the type of design (independent or repeated) for a
Wilcoxon signed rank test?
How many groups can it test for?
Answer: Repeated groups design
,It can test for 2 groups only
◉ What is an independent groups design?
Answer: Two groups of individuals with individuals in group 1 being
different then individuals in group 2
◉ What is a repeated measures design?
Answer: Two sets of scores but they are generated from the same
individuals
- Can have many conditions but same individuals must remain in
each condition
◉ What is the type of design (independent or repeated) for a kruskal
wallis test?
How many groups can it test for?
Answer: Independent measures design
It can test for more than 2 groups e.g 3 groups 1,2 and 3 with one set
of scores
,◉ What is the type of design (independent or repeated) for a
Friedman ANOVA test?
How many groups can it test for?
Answer: Repeated measures design
It can test for more than 2 groups
◉ What does parametric tests assume?
Answer: Additivity
Linearity
Normality
Homogeneity of variance/homoscedasticity
Independence
◉ Does non parametric tests make any assumptions?
Answer: They make no assumptions about population parameters
including the shape of the distribution of scores
◉ What is an alternative to non parametric tests?
Answer: Bootstrapping allows parametric tests without the
assumptions of normality
, ◉ What are some disadvantages when using bootstrapping for
parametric tests?
Answer: - Not many analysis can be bootstrapped ( not available)
- It doesn't solve the problem with violating the other assumptions
of parametric tests
- Is not good to use on small sample size and outliers
◉ What is the equation for Wilcoxon rank-sum test?
Answer: Ws = total sum of ranks/2
◉ What does the term " asymptotic" mean in the SPSS table for the
mann whitney u test?
Answer: Theory assuming large sample size
◉ Which parametric test uses the test statistic H ?
Answer: kruskal wallis test
◉ Pairwise comparisons can be generated through () which
calculates the ()?
Answer: Dunn's test
difference between mean ranks
QUESTIONS WITH SOLUTIONS GRADED A+
◉ What are the four non parametric tests?
Answer: 1. Wilcoxon rank sum test/ Mann whitney test
2. Wilcoxon signed rank test
3. Kruskal Wallis test
4. Friedman ANOVA
◉ What is the type of design (independent or repeated) for a
wilcoxon rank sum test/mann whitney ?
How many groups can it test for?
Answer: Independent groups design
It can test for 2 group only (e.g saline vs drug)
◉ What is the type of design (independent or repeated) for a
Wilcoxon signed rank test?
How many groups can it test for?
Answer: Repeated groups design
,It can test for 2 groups only
◉ What is an independent groups design?
Answer: Two groups of individuals with individuals in group 1 being
different then individuals in group 2
◉ What is a repeated measures design?
Answer: Two sets of scores but they are generated from the same
individuals
- Can have many conditions but same individuals must remain in
each condition
◉ What is the type of design (independent or repeated) for a kruskal
wallis test?
How many groups can it test for?
Answer: Independent measures design
It can test for more than 2 groups e.g 3 groups 1,2 and 3 with one set
of scores
,◉ What is the type of design (independent or repeated) for a
Friedman ANOVA test?
How many groups can it test for?
Answer: Repeated measures design
It can test for more than 2 groups
◉ What does parametric tests assume?
Answer: Additivity
Linearity
Normality
Homogeneity of variance/homoscedasticity
Independence
◉ Does non parametric tests make any assumptions?
Answer: They make no assumptions about population parameters
including the shape of the distribution of scores
◉ What is an alternative to non parametric tests?
Answer: Bootstrapping allows parametric tests without the
assumptions of normality
, ◉ What are some disadvantages when using bootstrapping for
parametric tests?
Answer: - Not many analysis can be bootstrapped ( not available)
- It doesn't solve the problem with violating the other assumptions
of parametric tests
- Is not good to use on small sample size and outliers
◉ What is the equation for Wilcoxon rank-sum test?
Answer: Ws = total sum of ranks/2
◉ What does the term " asymptotic" mean in the SPSS table for the
mann whitney u test?
Answer: Theory assuming large sample size
◉ Which parametric test uses the test statistic H ?
Answer: kruskal wallis test
◉ Pairwise comparisons can be generated through () which
calculates the ()?
Answer: Dunn's test
difference between mean ranks
