Module 9: Testing the difference b/w 2 Sample Means

Question 1

What is a t-test for independent samples?

Answer 1

Test between 2 different populations

Question 2

What is a t-test for paired (or dependent) samples?

Answer 2

Test for the Same Population @ different times

Question 3

What are the 3 Designed Assumptions for a t-test for independent samples?

Answer 3

1. The dependent variable is measured @ interval/ratio level
2. The independent variable is measured at the nominal or ordinal level, contains only two categories, and those categories are independent from each other
3. The samples were randomly drawn from their respective population

Question 4

What are the 3 Statistical Assumptions for a t-test for independent samples?

Answer 4

4. The dependent variable is normally distributed across the two levels of the independent variable
5. The variances are homogeneous (homogeneity of variances)
6. There are no extreme outliers

Question 5

Statistical Assumptions

How do we determine whether the data is normally distributed or not?

Answer 5

Question 6

Statistical Assumptions

How do we determine whether the data variances are homogeneous or not?

Answer 6

Question 7

Statistical Assumptions

How do we determine whether there is extreme outliers or not?

Answer 7

Kurtosis - Heavy Tails = More likely outliers

Boxplot

Question 8

What is the Hartley’s Fmax test?

Answer 8

Used to determine whether the data variances are homogeneous or not

Question 9

Hartley’s Fmax test Example

Answer 9

Question 10

What are the 5 Assumptions for the t-test for paired samples?

Answer 10

1. The dependent variable is measured @ interva/ratio level
2. The two groups are related, dependent or paired (within subjects design)
3. The samples was randomly drawn from the population
4. The differences between the two groups (time points) are normally distributed
5. There are no extreme outliers