Suppose a researcher RW decides to give his nine participants a series of 50 photographs. He asks the participants to rank order (from 1-50) how well-dressed the individuals in the photograph were and to rank order (from 1-50) how intelligent they thought the individuals in the photograph were. One of the photographs was of a well known actor. The well-dressed rankings and the intelligence rankings for this one photo are as follows.

WELL-DRESSED INTELLIGENT

14 12

35 39

16 17

41 43

39 40

25 21

26 27

11 9

24 25

RW performs a simple linear regression (with WELL DRESSED as the Independent Variable and INTELLIGENT as the Dependent Variable) and obtains the following results.

Model Summary

Model

R

R Square

Adjusted R Square

Std. Error of the Estimate

1

.988a

.975

.972

2.10093

a. Predictors: (Constant), welldressed

ANOVAa

Model

Sum of Squares

df

Mean Square

F

Sig.

1

Regression

1215.992

1

1215.992

275.492

.000b

Residual

30.897

7

4.414

Total

1246.889

8

a. Dependent Variable: intelligent

b. Predictors: (Constant), welldressed

Coefficientsa

Model

Unstandardized Coefficients

Standardized Coefficients

t

Sig.

B

Std. Error

Beta

1

(Constant)

-3.180

1.886

-1.686

.136

welldressed

1.133

.068

.988

16.598

.000

a. Dependent Variable: intelligent

From these results, he concludes

1. that being well dressed causes people to view you as intelligent.

2. that the linear relationship between well-dressed ratings and intelligence ranks for the photos is Intelligence=-3.180(Well Dressed) + 1.133.

3. that the null hypothesis (b=0) has a probability =.000

4. that since it is desirable to get an R2=1, it may be necessary to add more Independent Variables to the analysis.

5. that with this type of analysis one doesn’t have to worry about being overly influenced by extreme values.

Your task is to answer the following about the above project.

Do you see any problems with the design of the study, the data analysis run, or the results reported?

Are there analyses that should have been run that weren’t? What would you have done? Why?

Do you see any errors in the conclusions reached? If so, what should have been (or should not have been) concluded?

If you feel that other alternative statistical analyses should have been performed, then do them. If you do alternative analyses include any results you have obtained