Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

Let p represent the probability that a student engages in binge drinking. After

ID: 3322954 • Letter: L

Question

Let p represent the probability that a student engages in binge drinking. After running a logistic regression, we have the model: Logit(p) = 3.726 + 0.283(White)-0.434(Black)-0.2 34(Asian)- 0.209(age)+0.449(Fraternity) - .028(Male) Assuming all coefficients are significant, which factors increase the risk of binge drinking? A. There is not enough information to answer this question B. Belonging to a fraternity, being male ° C. Being Black or Asian, aging: being male D. Being white; belonging to a Fraternity Reset Selection uestion 7 of 34 The 95% confidence interval for a log-odds ratio is (-343, .573). The odds ratio is O A. Statistically significant at the 5% level O B. Statistically not significant at the 5% level Reset Selection

Explanation / Answer

Answer to the following with detailed explanations as below:

6.

Factors that have positive coefficient increase the risk of binge drinking. Only the following have +ive coefficient :
Fraternity and White

D. Being White, Belonging to Fraternity is correct


7. If the CI has 0 in it, which in this case it has because LCL and UCL are on 2 sides of 0: 1 +ive and the other -ive then we conclude that the coefficient isn't statistically different from 0. If it is not statistically different from 0 then it is not significant at that alpha level.

So, B. Statistically not signifiacnt at the 5% level is correct

8. In(odds) >0 then probability is greater than .5
True

Please see the probability S-curve for logit regression.