PARAMETRIC TESTS

Question 1

Significance or chance?

Answer 1

If 2 samples have large overlaps in a histogram then the difference was not significant, due to chance.
A small overlap shows the difference was significant.

Question 2

What is the aim of a t-test?

Answer 2

They compare averages and SD’s of 2 samples to see if there is a significant difference.

Question 3

Features of dependent t-test:

Answer 3

Matched pairs, paired samples, related, repeated measures. Same subjects all conditions, within subjects design.

Question 4

Features of independent t-test:

Answer 4

Unrelated t-test. Different pps do each condition. Between groups design.

Question 5

Name the parametric tests:

Answer 5

Independent means t-test
One way independent measures analysis of variance, ANOVA
Dependent means t-test
One way repeated measures analysis of variance, ANOVA

Question 6

Assumptions of parametric test:

Answer 6

Data must be interval or ratio scale
Random sampling from defined population, meant to be biased representative of sample
Data is normally distributed
Population variances are equal. 
There should be no significant outliers.