Term
| How do independent measures and correlated groups t test differ? |
|
Definition
Independent Measures:
- 2 separate samples
- presumed to come from 2 separate populations
Correlated Groups
- single sample from single population used twice on the same dependent variable
- 2 samples from same population |
|
|
Term
| Provide examples for a correlated groups t test. |
|
Definition
- Before vs. After
- Cond. 1 vs. Cond. 2,
- Score at Time 1 vs. Score at Time 2 |
|
|
Term
| What advantages are there to running a correlated groups t test? |
|
Definition
- more variables ascertained and controlled
- more realistic design than 1 sample t and z
- don’t need parameters like μ and σ
- SS can be own control group (in repeated measures) |
|
|
Term
| What is the difference between repeated measures and matched pairs design? |
|
Definition
repeated measures:
- single sample randomly selected and generates 2 sets of data
- don't need a control group because the sample is own control
matched pairs:
- 2 heavily matched samples taken from single population and measured one time
- matched on as many variables as possible |
|
|
Term
| What are the mechanics of a correlated groups t test? |
|
Definition
- collect 2 sets of data in Correlated Groups t test
- but use only 1 set of data called Difference Scores (D) --> all calculations use sample Difference scores in Correlated Groups t test.
- makes this a 2 sample t test that looks like a 1 sample t test |
|
|
Term
|
Definition
|
|
Term
| What is represented by the denominator of this hypothesis test? |
|
Definition
| Smd (standard error of sample means) = estimated standard error for the Difference scores.It is how much we can expect between MD and μD to differ based on chance alone |
|
|
Term
| The sampling distribution for a correlated groups t test is what? |
|
Definition
| all the possible values of differences you can get between 2 scores on the x axis and the probability of getting those scores (y axis |
|
|
Term
| Denominator of an independent measures t test is... |
|
Definition
| sM = estimated standard error of the mean (estimated σM) where σ is unknown and we have to rely on the sample s to know how much difference we can expect between M and μ based on chance alone |
|
|
