2. You are working with a healthcare provider who provides patients with a free annual health screening. The provider would like to better understand the factors that drive participation in the screening program. You use logistic regression to develop a model that predicts participation based upon an individual’s marital status and ethnicity.

The results of the model are shown here:

Call:

glm(formula = participated ˜ age + maritalStatus + ethnicity, family = binomial, data = patients_train)

Deviance Residuals:

Min 1Q Median 3Q Max

-1.739 -1.256 1.018 1.027 1.590 Coefficients:

Estimate Std. Error z value Pr(>|z|)

(Intercept) 1.424848 0.567979 2.509 0.0121 *

age 0.000498 0.002121 0.235 0.8144 maritalStatusMarried -0.195182 0.159257 -1.226 0.2204 maritalStatusNot Known -1.150035 0.175621 -6.548 5.82e-11 ***

maritalStatusSingle -0.770244 0.168187 -4.580 4.66e-06 ***

maritalStatusWidowed -0.441739 0.290676 -1.520 0.1286 ethnicityAsian -1.019093 0.543590 -1.875 0.0608 .

ethnicityBlack or African American -1.187287 0.544551 -2.180 0.0292 *

ethnicityHispanic -0.984501 0.545999 -1.803 0.0714 .

ethnicityNative Hawaiian or Other Pacific Islander -12.230119 196.968421 -0.062 0.9505 ethnicityTwo or More -1.060614 0.561182 -1.890 0.0588 .

ethnicityUnknown -1.217726 0.554415 -2.196 0.0281 *

ethnicityWhite -0.880737 0.536667 -1.641 0.1008

---

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

(Dispersion parameter for binomial family taken to be 1)

Null deviance: 8464.2 on 6111 degrees of freedom Residual deviance: 8223.3 on 6099 degrees of freedom AIC: 8249.3 Number of Fisher Scoring iterations: 10

a. In this model, which variable has the greatest effect on the outcome?

b. For that variable, rank-order the levels from the group least likely to participate in the assessments to the group most likely to participate in the assessments.