8. For the following questions circle ALL of the correct answers (questions may

Question

8. For the following questions circle ALL of the correct answers (questions may have moret may have more tharn one correct answer or no correct answers). Each sub-item is worth 1 point. a) to determine model adequacy for a multiple linear regression model with 5 predictors, one could i. look at residual plots. ii. look at a qq-plot of the residuals. iii, use the AIC value. (b) To choose which predictor variables to include in a model, one could i. look at Cook's D values. ii. look at R2 ii. use the likelihood value. iv. look at residual plots. (c) A valid way to deal with multicollinearity in predictor variables is to i. perform a transformation. ii. leave out one or more predictor variables. iii. fit with generalized least squares.

Explanation / Answer

a) All three (i, ii, and iii) are corrects. For constant variance we need to used residual plot, for normality assumption weneed to used qq plot of residuals, and for multicollinearity assumption we need to used AIC values.

b) The first option (i) is the only correct option.

R2 is used for how many variation in dependent variable explain by all independent variable. likelihood values are the estimates of regression coefficients and residual plot used for constant variance that is used for checking the assumptions of homoschedastic.

c) The first two option are correct (i, and ii). Non linear transformation may help to eliminate linear relationship between the predictors. Also we can remove one of the highly correlated predictor.

Related Questions

Navigate

• Browse (All)
• Browse 8
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.