Which stats test?

Question 1

Parametric, paired data with 2 groups

Answer 1

Paired T-test

Compares means from two groups

Question 2

Parametric, unpaired data with 2 groups

Answer 2

unpaired (independent) T-test

Compares means from two groups

Question 3

Parametric, paired, with >2 groups

Answer 3

Repeated measures, one way ANOVA
Compares means from more than two samples. Only tells if a difference exists between A, B and C. Won’t tell you which group is different

Question 4

Parametric, unpaired, with >2 groups

Answer 4

One way ANOVA
Compares means from more than two samples. Only tells if a difference exists between A, B and C. Won’t tell you which group is different

Question 5

Non-parametric, paired, with 2 groups

Answer 5

Wilcoxen signed rank

Question 6

Non-parametric, unpaired, with 2 groups

Answer 6

Mann- Whitney U test

Question 7

Non-parametric, paired, >2 groups

Answer 7

Friedman test

Question 8

Non-parametric, unpaired, >2 groups

Answer 8

Kruskal-Wallis test

Question 9

Which test is used for discrete data regardless of whether you’re identifying a correlation or comparing and what does it measure?

Answer 9

Chi-squared
Measures difference between observed and expected frequencies. The higher the X^2 the bigger the difference between the observed and expected frequencies

Question 10

Testing for correlation for parametric data

Answer 10

Pearson’s
Get an r-value between -1< r < +1
Closer to 1= stronger correlation

Question 11

Testing for correlation for non-parametric data

Answer 11

Spearman’s rank

Question 12

Which test is used to test for normality if n<50?

Answer 12

Shapiro-Wilks

Question 13

Which test is used to test for normality if n>50?

Answer 13

Kolmogorov-Smirnov