## Instructions

**General instructions:** You will have **an hour (60 min) **to complete this exam. Once you have started the clock with start. In other words, it will not save for you. I am expecting that you have spent the normal time preparing for this exam as you would have for an in-class exam.

**Multiple choice**: You should *not* treat this as an open book exam as looking up the answers will leave you without enough time to complete it.

### Attempt History

Attempt | Time | Score | |
---|---|---|---|

LATEST | Attempt 1 | 51 minutes | 40 out of 46 |

Data from 1995 from over 1,000 colleges were used to predict the percent of alumni who donate to a college from the average SAT score of students attending that college. The resulting regression equation was Y = –29.29 + .05(X). This regression indicates that: |

Which of the following statistics quantifies the improvement in our ability to predict a person’s score when using the regression line rather than the mean? Or more specifically, what statistic tells us how much variance in the “DV” is *explained *by the “IV.”

To *predict* a single outcome variable (a “DV”) from *more than one* predictor variables (“IVs”), which statistical technique would you use?

If a researcher does not have a good theory to determine which of her three variables (“IVs”) is the best predictor of the outcome variable (“DV”), which method of multiple regression should she use?

It is more difficult to reject the null hypothesis when: |

*t*tests when the independent variable has more than two levels?

*F*statistic increases when:

__group__means is:

Suppose I was interested in if people per physician predicts life expectancy. To do so I collect the life expediencies and data on people per physician from Bangladesh, Kenya, North Korea, the United States, and Italy. I find the following means and standard deviations:

Life expectancy: *M = *67.70*, **SD *= 10.35

People per physician: *M = *2957.60*, **SD *= 3627.16

I then compute a regression line by doing the following calculations.

**Step 1: To find the intercept I set x = 0**

Zx = (0-2957.60)/3627.16 = -.8154

Zy = (-.8154)(-.71) = .5789

Y = (.5789)(10.35) + 67.7 = 73.6916

**a = 73.6916
**

**Step 2: To find the slope I set x = 1**

Zx = (1-2957.60)/3627.16 = -.8151

Zy = (-.8151)(-.71) = .5787

Y = (.5787)(10.35) + 67.7 = 73.6895

**b** = 73.6916-73.6916 = **-.002**

**Given my calculations above, what is the regression equation?**

Imagine that your younger brother is convinced that he can *predict* a person’s age based on the number of movies he or she sees on an annual basis. To do so he measures the number of movies people watch and asks them their age.

What type of analysis should your brother perform to test his idea?

Diana wants to see if eating Valentine’s Day candy causes happiness. She asks 100 participants to come to the laboratory on Valentine’s Day and randomly assigns them to eat a piece of candy or some celery. Participants then spend about 5 minutes rating their levels of general happiness.

What type of analysis should Diana preform in order to test her idea?

Margo is interested in what reptiles people are most afraid of. To test this she has participants come into the lab and presents them with a snake, a lizard, and a crocodile. After participants see each reptile they rate their level of fear.

What type of analysis should Margo preform in order to test her idea?