Question

Based upon the scatterplot matrix for the seven regressors, comment on the possibility of multicollinearity between some of the regressors. Note, scatterplots can also be useful in spotting outliers and other data anomalies.

| Multivariate Correlations X1 X1 1.0000 X2 -0.1919 X3 0.4887 X4 0.6729 X5 -0.7360 X6 -0.8167 X7 -0.1780 X2 -0.1919 1.0000 0.6302 0.1765 0.1658 0.1271 0.2551 X3 0.4887 0.6302 1.0000 0.8137 -0.4634 -0.5586 0.0733 X4 0.6729 0.1765 0.8137 1.0000 -0.5775 -0.6646 -0.1141 X5 -0.7360 0.1658 -0.4634 -0.5775 1.0000 0.8852 0.1442 X6 X7 -0.8167 -0.1780 0.1271 0.2551 -0.5586 0.0733 -0.6646 -0.1141 0.8852 0.1442 1.00000.1414 0.1414 1.0000

Answer 1

Answer:

Based upon the scatterplot matrix for the seven regressors, comment on the possibility of multicollinearity between some of the regressors. Note, scatterplots can also be useful in spotting outliers and other data anomalies.

Pearson pairwise correlation coefficients were varied from a weak correlation (0<=|r|< 0.3) to a moderate correlation (i.e., 0.3≤|r|<0.7) and a strong correlation (i.e., |r|>=0.7).

Therefore buy considering 0.7 as cut off value,

x1 is highly correlated with x5 and x6.

X3 highly correlated x4

X5 highly correlated x6

Observing the strength of the correlations, x6 is highly correlated with x5 and the x1. Therefore it is to see the variable x6 removed and then run the regression to see any problem of multicollinearity.