Logistic Regression You are an institutional researcher for the Go County School District. The superintendent is interested in predicting whether graduating seniors will be admitted into an Ivy League...

Logistic Regression

Logistic Regression You are an institutional researcher for the Go County School District. The superintendent is interested in predicting whether graduating seniors will be admitted into an Ivy League college (e.g., Harvard, Yale). You collect data on 400 seniors which contains information on whether the student the student was admitted to an Ivy League college (admit), the students’ GPA (gpa), standardized test score (test), and their class rank in four categories (rank). Treat rank as a categorical variable and use the last group “lowest rank” as the reference group. Use the admit.sav dataset on Canvas and regress the acceptance rate on students’ GPA, test score, and class rank. Use an alpha of .05 to answer the questions below. 1. What proportion of seniors were admitted to an Ivy League school? (2 points) 2. What is the regression equation predicting admission to Ivy League schools? (2 points) 3. Is the full regression model an improvement over the null model? What information did you use to come to this conclusion? (2 points) 4. What proportion of cases are correctly classified using the full logistic regression model? (2 points) 5. Report and interpret the results of the Hosmer and Lemeshow Test. If the test is statistically significant use the contingency table to diagnose the source of misfit. (2 points) 6. Calculate the squared correlation between the model’s predicted value and actual values (i.e, Efron’s R square value) and compare it with the values reported for Cox & Snell and Nagelkerke R Square values. (2 points) 7. Interpret the regression coefficient and odds ratio for GPA. (2 points) 8. What is the regression coefficient for test score telling you? (2 points) 9. Interpret the effect for class rank. In other words, is the main effect for rank statistically significant? What information did you use to come to this conclusion? Also, substantively interpret each dummy/indicator variable for rank. (4 points)