Scenario:

A multiple regression analysis was conducted to determine if four different measures (test1-test4) were good predictors of a measure of proficiency on the job (jobscore). The following results were obtained.

Please answer all of the questions that follow the output

Model Summary:

Model: 1

R: .918^a R Squared: .843 Adjusted R squared: .811 Std. Error of Estimate: 27.618

a. Predictors: (Constant), test4, test3, test1, test2

Model:

1 Constant: Unstadardized B: 28.150 Std Error: 42.787 Coefficients Beta: Nothing t: .658 Sig: .518

test 1 Unstadardized B: 2.404 Std Error: .288 Coefficients Beta: .804 t: 8.336Sig: .000

test 2 Unstadardized B: .308 Std Error: .216 Coefficients Beta: .138 t: 1.425 Sig: .169

test 3 Unstadardized B: -.874 Std Error: .292 Coefficients Beta: -.277 t: -2.995 Sig: .007

test 4 Unstadardized B: -.778 Std Error: .351 Coefficients Beta: -.209 t: -2.218 Sig: .038

A) What is the r squared value?

B)Which of the predictors, if any, are statistically significant (using the alpha of .05 like we have the whole semester)?

C) Based on the results above, if someone were to have scored a 0 on test 1, 2, 3 and 4...what would be their predicted jobscore (the dependent variable)? Hint: think about the definition of the intercept/constant.

D)Which of the predictors (tests 1-4) appear to have a negative relationship with the dependent variable (jobscore)?