The data in the following table come from an article by Bethel et al. (1985). All subjects are asthmatics. For the model with FEV1 as the response and HEIGHT, WEIGHT, and AGE as the predictors,

a. Examine a plot of the studentized or jackknife residuals versus the predicted values. Are any regression assumption violations apparent? If so, suggest possible remedies.
b. Examine numerical descriptive statistics, histograms, box-and-whisker plots, and normal probability plots of jackknife residuals. Is the normality assumption violated? If so, suggest possible remedies.
c. Examine outlier diagnostics, including Cook's distance, leverage statistics, and jackknife residuals, and identify any potential outliers. What course of action, if any, should be taken when outliers are identified?
d. Examine variance inflation factors, condition indices (unadjusted and adjusted for the intercept), and variance proportions. Are there any important collinearity problems? If so, suggest possible remedies.

Transcribed Image Text:

Height Weight FEVİ Height Weight FEVİ Subject Sex k() AGE 5 AGE Subject (yr) Sex (cm) (kg (L) 24 M 75 78.0 4.7 4.3 28 F 171 98.0 3.5 70.5 75.0 68.0 65,0 26 M 180 29 M 63 33F 31 M 180 З0 M 180 70.4 22 M 27 M 168 91.2 46 M 178 36 M 173 620 3.5 12 13 14 15 1G 17 18 19 M172 67.6 0 2.6 2.0 180 25 M 166 65.5 26 F 166 65.0 .2 22 M 176 655 .7 185 85.5 43 76.3 4.7 36 M 79. 5.2 4.2 168 63.0 3.9 3.0 4.5 2.4 27 M 27 M 171 67.0 10 24 M 18 88.2 Forced expiratory volume in I second.