Answers to the Test 3 Review Sheet

1. D. anonymity--it must be confidential, but it doesn’t have to be anonymous

2. C

3. You shouldn’t agree with the conclusion. It’s not an experiment, so you can’t say anything about causes. It may be the flunking out that led people to have low self-esteem rather than the other way around.

4. C

5. Remember to reverse score and make it so that higher scores mean more of the construct (here, hiking enjoyment). Then add them up to get the score.

It’s a Likert scale.

You evaluate using coefficient of correlation or item-total reliability (same thing).

.90 or above is very good

It’s a psychometric scale.

The scenario:

IV = confederate ethnicity, participant ethnicity, game outcome

DV = ratings of how much of own performance was due to luck and skill and ratings of how much opponent’s performance was due to luck and skill

Control variables could be the type of game, the amount of time spent playing, etc.

Levels of the IV:

Participant ethnicity: Asian-American, European-American--between

Confederate ethnicity: Asian-American, European-American--between

Game outcome: Win, Lose, Tie--between

Yes, it’s a factorial: 2 x 2 x 3 12 cells

To have 20 per condition, you need 20 x 12 = 240

3 main effects (one for each IV) are possible, 4 interactions (Part. ethnic by confed ethnic, part ethnic by outcome, confed ethnic by outcome, part ethnic by confed ethnic by outcome) are possible