Lecture 5

Question 1

Parametric Statistical Tests

Answer 1

Tests that compare means of 2 groups
(1 sample t test, paired samples t test etc.)
Tests that compare means of more than 2 groups

Question 2

Parametric assumptions

Answer 2

Data are numeric
Random sample drawn from a: Population where the variables of interest have Gaussian distrib.)
Homogeneity of variances in the groups studied (intervention groups)

Question 3

Two-way ANOVA

Answer 3

• Difference in means according to 2 or more factors (between subjects)
• Difference in means among 2 or more groups when measures are repeated (within subjects)

Question 4

Purpose of Two-way ANOVA

Answer 4

Question effects of 2 factors on a numeric outcome

• Main effects (A,B)
• Interaction (A*B)

Question 5

Non-parametric Tests for Numeric Data

Answer 5

Parametric assumptions are not met

-Distribution is non-gaussian

Question 6

Correlation Tests

Answer 6

Two or more numeric variables

• Parametric (pearson correlation, simple linear regression)
• Non-parametric (spearman correlation)

Question 7

Assessment of linear correlation

Answer 7

Correlation (p value)
Strength of correlation (r)

.75:Excellent or strong correlation

Question 8

Spearman Rank correlation

Answer 8

Non-parametric correlation

• Data are ordinal
• Indicated by Spearman Rho (looks like a p rho)