In Ability, Gender, and Performance Standards: Evidence from Academic Probation, Jason M. Lindo, Nicholas J. Sanders, and Philip Oreopoulos find evidence using regression discontinuity methods that being placed on academic probation increases the probability that a student voluntarily leaves the university. Figure 1 below shows the estimated discontinuity in probation status at the end of the first year, and Figure 2 shows the impact of being placed on academic probation on the decision of male students to permanently leave the university at the end of their first year. Figure 1 Figure 2 1) Why doesn’t the mean difference between the rates of voluntary leaving for students placed on probation and those not placed on probation provide an unbiased estimate of the effect of being placed on probation? 2) Based on Figure 1, which type of RD do you think this study is using, sharp or fuzzy RD? Explain the difference between the two. 3) Briefly explain why the regression discontinuity estimates provide an unbiased estimate of the effect of being placed on probation? 4) In this study, GPA is used as a running variable. Describe the role of the running variable in a sharp regression discontinuity design.