Least Squares Linear Regression of JobSat

Predictor

Variables Coefficient Std Error T P VIF

Constant 41.6019 18.5465 2.24 0.0749 0.0

SupFdbk 0.60863 0.51319 1.19 0.2889 174.5

Salary 5.973E-04 5.053E-04 1.18 0.2903 201.2

SalSup 1.694E-05 1.710E-05 0.99 0.3672 6445.7

SupSq -9.116E-03 9.260E-03 -0.98 0.3701 2024.1

SalSq -1.020E-08 7.590E-09 -1.34 0.2368 1386.3

R 0.9880 Mean Square Error (MSE) 6.74382

Adjusted R 0.9761 Standard Deviation 2.59689

AICc 63.655

PRESS 134.66

Source DF SS MS F P

Regression 5 2784.46 556.893 82.58 0.0001

Residual 5 33.72 6.74382

Total 10 2818.18

Cases Included 11 Missing Cases 0

What is the interpretation of the coefficient for salary in Printout?

For every one-minute increase in supervisor feedback, we expect salary to go up $59.73

For every $59.73 increase in salary, we expect job satisfaction to go up one point, holding supervisor feedback constant

We should not interpret this beta because the interaction term is not useful

We should not interpret this beta

If we believe there will be a positive relationship between salary and employee satisfaction, what would we do with the p-value of the t-test of this relationship?

divide it by 2

multiply it by 2

interpret it as a positive correlation

nothing