Suppose that a relationship between presidential voting ...

[sammy]

Suppose that a relationship between presidential voting preference and gender has a statistical significance level of .001 and the P-R-E measure of association for this relationship is .35. Which of the following statements is correct? (Answer: B)
 
  A. By knowing the dependent variable gender, we can reduce the number of errors in predicting the independent variable presidential voting preference by 35.
  B. By knowing the independent variable gender, we can reduce the number of errors in predicting the dependent variable presidential voting preference by 35.
  C. The relationship is not statistically significant; accept the null hypothesis.
  D. There is a statistically significant relationship, but the relationship is not strong enough to be of any theoretical importance.

Question 2

Which of the following does a P-R-E measure of association tell us?
 
  A. the degree of relationship between the dependent variable and the independent variable
  B. how well we can predict the dependent variable by knowing the independent variable
  C. how well the independent variable explains the dependent variable
  D. All of the above are correct; they all mean basically the same thing.
  E. None of the above is correct.

[janeli]

Answer to Question 1

B.

Answer to Question 2

D.