a (3) In the Eastern European country of Romania, admission to university is based in large part on scores on a national

[answerhappygod]

A 3 In The Eastern European Country Of Romania Admission To University Is Based In Large Part On Scores On A National 1 (49.38 KiB) Viewed 25 times

a (3) In the Eastern European country of Romania, admission to university is based in large part on scores on a national exam called the Bac. Scoring at least a 6.0 on the exam makes it much easier to attend university than scoring below this, so much so that the differences in the likelihood of attending university are large between students who score just below and those who score just above this value. To simplify this example, we'll assume that all students who score 6.0 or above go to university and all those who score below 6.0 do not go to university (it's actually the case that over 2 85% of those scoring narrowly above 6 attend university about 20% of those scoring narrowly below 6.0 attend university, but, again, for the purposes of this question we'll pretend that everyone above 6 attends and nobody below 6 attends).

The plot below shows the results of a study in which people are surveyed at the age of 30 and asked questions about their tolerance for corruption. Assume that researchers also were able to verify each survey respondent's Bac exam score from when they were in high school (and again assume that everyone who got at least a 6.0 on the Bac went to college and everyone who got below a 6.0 did not go to college). The figure shows on the y-axis the average value of the dependent variable that is tolerance for corruption (i.e. how much a person things various corrupt acts such as bribery are acceptable, with higher values meaning more tolerance for corruption). These points show the average value of tolerance for corruption among those who have each unique score value on the exam (exam score is on the x-axis, note that only students with scores relatively close to the passing threshold of 6.0 are included in the data here). Average Tolerance for Corruption 0.0 0.5 1.0 1.5 -1.0 5.8 5.9 6.0 6.1 6.2 Bac Score

(a) If you were to estimate the relationship between Bac score and average tolerance for corruption using only observations below 6.0 on Bac score (i.e. those who do not attend university), what would you estimate in general terms? (It might help to draw a line on this plot but you don't have to turn in this picture just describe what that line would look like - you don't have to give exact numbers just explain generally) (b) If you were to estimate the relationship between Bac score and average tolerance for corruption using only observations above 6.0 on Bac score (i.e. those who attend university), what would you estimate? (c) Now think about the predictions of those lines you drew and how what each of their values would be at the passage cutoff of 6.0. Explain how this might tell you about the effect of university 3 attendance on tolerance for corruption, including what the (rough - estimated from looking at the figure) size of this effect would be estimated as. (d) Why might one trust the estimated effect of university attendance from this regression discon- tinuity analysis more than one from a simple survey (ignoring Bac scores and just surveying the general public) where people are asked whether they attended university and what their views on corruption were? In other words, if we just said, ignoring this exam, "in general people who attend university on average have a lower tolerance for corruption than those who don't attend university" give some examples of reasons why we might see a relationship between these variables even if university attendance may not really affect this dependent variable. (i.e. how might people who attend university and people who don't attend university be different on average in ways other than their university attendance)