The table below refers to a survey by the Wright State University School of Medicine and...

  

Transcribed Image Text:

The table below refers to a survey by the Wright State University School of Medicine and the United Health Services in Dayton, Ohio. The survey asked 2276 students in their final year of high school in a nonurban area near Dayton, Ohio, whether they had ever used alcohol (A), cigarettes (C), or marijuana (M). Alcohol. Use Cigarette. Use Yes No. Yes No Yes No The information below is from the logit model treating M as the response (1 for Yes, 0 for No) and C and A as the predictors. Deviance Residuals: Call: glm (formula = cbind (yes, no) C + A, family 1 Marijuana. Use Yes No Coefficients: CYes AYes 2 3 0.03353 -0.09697 -0.34739 0.49273 4 911 538. 44 456 3 43 2 279 = binomial, data = Estimate Std. Error z value Pr(>|z|) (Intercept) -5.3090 0.4752 -11.172 < 2e-16 *** 2.8479 2.9860 0.1638 17.382 < 2e-16 *** 0.4647 6.426 1.31e-10 *** Signif. codes: 0 '*** 0.001 '**' 0.01 * 0.05. 0.1 1 Null deviance: 843.82664 on 3 Residual deviance: 0.37399 on 1 AIC: 25.101 Number of Fisher Scoring iterations: 4 (Dispersion parameter for binomial family taken to be 1) degrees of freedom degrees of freedom dfb) (a) Find the estimated odds ratio for alcohol use (Yes vs. No) with marijuana use (the AM odds ratio). Interpret the odds ratio by using it in a sentence. [4 pts] (b) Is the AM odds ratio you reported conditional or not? If so, what is it conditional on? [4 pts] (c) Does the estimated AM odds ratio depend on the value of C? Why or why not? What type of association between A and M is this called? [4 pts] (d) Perform a goodness-of-fit test for this model. Does it fit well? How is this test related to the test for interaction between A and C? [4 pts] (e) Calculate the marginal odds ratio for AM (note be careful about what values you are summing together). Does this differ from the AM odds ratio estimated by the model? Is there evidence of confounding by C on the AM odds ratio? [4 pts] The table below refers to a survey by the Wright State University School of Medicine and the United Health Services in Dayton, Ohio. The survey asked 2276 students in their final year of high school in a nonurban area near Dayton, Ohio, whether they had ever used alcohol (A), cigarettes (C), or marijuana (M). Alcohol. Use Cigarette. Use Yes No. Yes No Yes No The information below is from the logit model treating M as the response (1 for Yes, 0 for No) and C and A as the predictors. Deviance Residuals: Call: glm (formula = cbind (yes, no) C + A, family 1 Marijuana. Use Yes No Coefficients: CYes AYes 2 3 0.03353 -0.09697 -0.34739 0.49273 4 911 538. 44 456 3 43 2 279 = binomial, data = Estimate Std. Error z value Pr(>|z|) (Intercept) -5.3090 0.4752 -11.172 < 2e-16 *** 2.8479 2.9860 0.1638 17.382 < 2e-16 *** 0.4647 6.426 1.31e-10 *** Signif. codes: 0 '*** 0.001 '**' 0.01 * 0.05. 0.1 1 Null deviance: 843.82664 on 3 Residual deviance: 0.37399 on 1 AIC: 25.101 Number of Fisher Scoring iterations: 4 (Dispersion parameter for binomial family taken to be 1) degrees of freedom degrees of freedom dfb) (a) Find the estimated odds ratio for alcohol use (Yes vs. No) with marijuana use (the AM odds ratio). Interpret the odds ratio by using it in a sentence. [4 pts] (b) Is the AM odds ratio you reported conditional or not? If so, what is it conditional on? [4 pts] (c) Does the estimated AM odds ratio depend on the value of C? Why or why not? What type of association between A and M is this called? [4 pts] (d) Perform a goodness-of-fit test for this model. Does it fit well? How is this test related to the test for interaction between A and C? [4 pts] (e) Calculate the marginal odds ratio for AM (note be careful about what values you are summing together). Does this differ from the AM odds ratio estimated by the model? Is there evidence of confounding by C on the AM odds ratio? [4 pts]

Answer rating: 100% (QA)

Find the estimated odds ratio for alcohol use Yes vs No with marijuana use the AM odds ratio Interpret the odds ratio by using it in a sentence 4 pts 4 049273 The estimated odds ratio for alcohol use ... View the full answer