This outline is also available in PDF.
Held: Wednesday, 30 April 2008
We explore a similar and different kind of two-quantiative-variable
test, one in which each observational unit from the first population
has a natural pair in the second population.
- Due Friday: 26-1, 26-2, 26-3, 26-4 (not to be turned in).
- I'd like you to do the first half of topic 26 on your own. (We won't do the second half.)
- Don't forget the poster session tomorrow!
- I have optional permission forms for you to sign.
- EC for attending the student art show.
- Two more exam-lettes to finish off the semester: Friday and Wednesday.
- I will be leaving class today at 1:55 to pick up my son from school. Katherine and Cassie can handle the last ten minutes.
- There is an error in activity 23-4 a. Similar data were not provided in Activity 22-3. Nonetheless, you should be able to reflect on the meaning of the problem.
- What are paired data?
- How should we analyze paired data?
- The poster session runs from 11:00 a.m. to 12:30 p.m.
- Do your best to be there by 11:00 a.m.
- If you have to (want to) leave by noon, and have completed your work,
- Lunch will be available for all participants.
- Remember that you provided your P-card number, so this is your lunch
for the day.
- We will hang your posters in the morning. (At least I hope we will.)
- You will be responsible for reviewing two other posters from our class,
and one poster from outside of our class.
- In class: Assigned
- Outside of class: Your choice
- All of your forms go to me.
- If possible, try to have one person from your group at the poster at
- Groups with one person can't meet this requirement.
- You can keep your posters.
- If you don't want your poster, let me know.
- If you can't agree on who gets to keep your poster, talk to me or Katherine.
- Sometimes we have a two sets of values in which we can naturally pair
one value from the first set with a corresponding value in the second
- Ages of married men and women
- Before and after scores
- Reponses of subjects to two different stimuli
- When analyzing such data, we should not use the two-sample
- Instead, we do a standard t-test, using the
differences between paired samples as our variable.