QM

Question 1

T-test (for independent samples)

Answer 1

Parametric

Does than mean differ between independant samples

Assumptions

• Normal data distribution
• Adequate sample size
• Equality of variance
• Data collected from a randomly selected representative sample of population
• independant

Question 2

Paired T-test (repeated measures data )

Answer 2

For paired values (before and after study)

Difference between before and after values rather than mean

assumptions
- Normality
Independence
Homogeneity og variances

Question 3

ANOVA

Answer 3

Independent samples

Difference between two or more sample means (better than multiple means, decreased chance of false positive

Lervenes test - checks homogeneity of variences

Tukey post hoc test - show which varieties were different

Question 4

2-way ANOVA

Answer 4

ANOVA for studies with two independent variables

Has 3 Null hypothesis

• The population means of the first factor are equal
• Population means of second factor are equal
• No interaction between the two factors

Question 5

Repeated measures ANOVA

Answer 5

Paired T-test for 2 or more repeated measures

Sphericity = equal variances across all time points - use Mauchy’s W test for this

Question 6

ANCOVA

Answer 6

ANOVA with covariates

Covariate a factor that cannot be controlled

Take factor into account
Don’t take factor into account

is there a difference between pops?

Question 7

Linear regression

Answer 7

Assumes causation - prediction

uses independant variable to predict dependant variable

small slope - large residuals - non-significant

large slope - small residuals - significant

Assumptions

• Residuals are normally distributed
• Equal variance
• Relationship is linear
• No relationship between residuals and x or y

Question 8

Pearson Product moment and correlation

Answer 8

Doesn’t assume causation - doesn’t allow prediction

Determines wether the changing of one variable changes the other

Question 9

Mann-Whitney U test

Answer 9

Non para equivalent of T-test

DO the means differ between two independent samples

Normal distribution not assumed

Question 10

Wilcoxon signed ranks test

Answer 10

Non para paired T test

Difference between paired sample values

Normal distribution not assumed

Question 11

Kruskal-Wallis test

Answer 11

Non para one way ANOVA

Difference between two or more sample means

Question 12

Spearmans rank

Answer 12

Tests statistical dependance between rankings of two variables

avoids assumptions of normality and homogeneity of variance