1. Using the dataset "Assumptions met," examine the effects
of varying the range on each of the following: |
- slope
- standard error of the estimate
- Pearson's r
- the corrected value of Pearson's r
|
2. Using the dataset strength, |
- Find the correlation if only subjects in the upper-half of the grip
strength distribution. (Use the top 74 subjects).
- Is it higher or lower than the correlation in the full sample?
- How does the corrected correlation compare with the correlation in
the full sample?
- Find the correlation for the remaining 73 subjects.
- Find the correlation for those subjects scoring below 80 or above
140 on grip strength. Compare this correlation to the full-sample correlation
and the corrected correlation.
|
3. For the SAT data, |
- Compare the standard error of the estimate for subjects scoring below
500 on the SAT Verbal with the standard error of the estimate for subjects
scoring above 500.
- Compare the slopes for these two portions of the sample.
- How do the answers to (a) and (b) above affect the corrected r for
the two portions of the sample. Explain why.
|