Part 1: Admissions to medical school are typically made on a hybrid model involving clinical and statistical components. MCAT scores and grades are examined to determine which candidates will be interviewed, then candidates who make that cut visit the medical school for interviews with staff and faculty. An admissions committee then combines the information clinically and decides whom to admit.
• If you were asked to make the process "more statistical;' what recommendations would you make? Do you think the current model is an effective one? Why or why not?
Part 2: As an external I-0 consultant, you are asked by a manager in the Human Resources department of a large organization to calculate whether there is adverse impact in a selection system that has been used over the last year. You are provided with the following data:
# of majority group members that applied: 300
# job offers to majority group members: 100
# minority group members that applied: 75
# job offers to minority group members: 15
• What is the adverse impact ratio? Given what you found, is there evidence of adverse impact? If there is adverse impact, what would you recommend in terms of next steps that the organization takes?