A study was conducted to determine if a person’s agreeability (whether they are generally an agreeable person or not) affects their income level. It also looked at the person’s gender to see if it was a factor as well. The variables are shown below:
Income: y = annual income in dollars
Agree: x1 = agreeability level with higher scores indicating a person is more aggregable and lower scores indicating less agreeable
Gender: x2 = 1 if male, 0 if female
Use the following information to answer the multiple regression questions.
Printout A: Pairwise Correlations:
Income Agree
Agree -0.2785 1.0000
Gender 0.8327 -0.0478
Printout B: Best Subset Regression Models for Income
Forced Independent Variables: (A)Agree (B)Gender (C)x1x2
Unforced Independent Variables: (D)x1sq (E)x1sqx2
Adjusted AICc -
P Cp R Square Min AICc Resid SS F P(F) Variables
4 4.6 0.7552 1798.32 5.805E+09 A B C
5 4.0 0.7593 1797.84 5.648E+09 2.64 0.1075 A B C D
5 5.3 0.7560 1799.22 5.726E+09 1.30 0.2568 A B C E
6 6.0 0.7568 1800.15 5.648E+09 3.31 0.0356 A B C D E
Cases Included 100 Missing Cases 0
Which of the following statements about multicollinearity is true from Printout A?
The data exhibit very little multicollinearity since there is a low correlation between agreement level (x1) and gender (x2). |
No multicollinearity exits since the model is statistically useful for predicting income (y). |
There is a high degree of multicollinearity in the data since the R2 for the model is high. |
The data exhibit some multicollinearity since there is a moderate correlation between agreement level (x1) and income (y). |
here
X1 (agreeability level) and X2 (gender) are the independent or regressor or predictor variables
Y ( income ) is the dependent variable or response variable
we consider multiple regression equation of Y on X1, X2 .
multicollinearity exists in a multiple regression model when two or more independent variables are moderately or highly correlated
From Printout A, we can see the correlation coefficient between X1 ( agreeability level) and X2 ( gender) is -0.0478 which is very low (only 4 % negatively correlated).
so the variables X1 and X2 are not correlated significantly.
hence,
the data exhibit very little multicollinearity since there is a low correlation between agreement level (X1) and gender (X2)
Get Answers For Free
Most questions answered within 1 hours.