Week 12 Non-Parametric Tests

Question 1

What are parametric tests?

Answer 1

T-test, ANOVA

• estimation of one or more parameters of the distribution of scores in the population
• rely on assumptions concerning the shape of the distribution
• usually include a normality assumption

Question 2

What are the assumptions of One-Way ANOVA?

Answer 2

• Independence
• Normality
• Homogeneity of Variance

Question 3

What are the assumptions of Within Subjects ANOVA

Answer 3

• Each subject each level
• Normally distributed treatment populations
• homogeneity of variance
• Interval scale of measurement
• Sphereicity assumed
• ID same for each S

Question 4

What are non-parametric tests?

Answer 4

• don’t rely on parameter estimation or distribution assumptions
• Major advantage since they do not rely on an restrictive assumptions concerning shape of sampled pop
• validity of test not affected by variable is normally distributed in population or not

Question 5

What is the disadvantage to non-parametric tests?

Answer 5

Lower power relative to corresponding parametric tests

Question 6

What are some non-parametric tests?

Answer 6

• Wilcoxon’s Rank-Sum Test (two independent samples)
• Mann-Whitney U test
• Wilcoxon’s Matched-Pairs Signed Ranks Test (related samples t test)
• Kruskal- Wallis One Way ANOVA (one way anova)
• Friedman’s Rank Test for k Correlated Samples (repeated measures ANOVA)

Question 7

What are analytic comparisons?

Answer 7

Statistical analysis of difference between means.

Question 8

What is a pairwise comparison?

Answer 8

Analysis of difference between two means

Question 9

What are complex comparisons?

Answer 9

Where a average of two or more groups means are compared with either a single group mean or average of two or more groups means.