Admissions directors of graduate business programs constantly review the criteria they use to make admission decisions. Suppose that the director of a particular top-20 MBA program is trying to understand how she and her staff discriminate between those who are admitted to their program and those who are rejected. She collects data on each of the following variables for 100 randomly selected individuals who applied in the past academic year: whether the applicant graduated in the top 10% of his or her undergraduate class, whether the admissions office interviewed the applicant in person or over the telephone, the applicant's overall GMAT score (where the highest possible score is 800), and the applicant's undergraduate grade-point average (standardized on a four-point scale). These data are provided in the fileP17_03.xlsx. Use logistic regression to discriminate between those who are admitted to this MBA program and those who are not on the basis of these variables.

Using all the variables, what are the logistic regression model coefficients? Round your answers to one decimal place, if necessary. If your answer is a negative number, enter a "minus" sign.

Regression CoefficientsCoefficientConstantTop TenInterviewGMAT ScoreUndergraduate GPAThe logistic regression model classifies the applicants very well. However, all the p-values are very high, due to collinearity. Refine the model by removing one at the time all the collinear variables until all variables left are significant at any reasonable level of significance. What is the final model? Enter a coefficient of zero for any variable removed from the model. Round your answers to one decimal place, if necessary. If your answer is a negative number, enter a "minus" sign.

Regression CoefficientsCoefficientConstantTop TenInterviewGMAT ScoreUndergraduate GPA

https://docs.google.com/spreadsheets/d/1Ohfty9ekXtUc5t7nth5nhwNE6JaBb-kA07H53b20H0k/edit?usp=sharing