The Inequality Machine: Difference between revisions
(Created page with "<blockquote>Among economists, though, this question has long been a matter of debate. Proponents of a more laid-back approach to college admissions base their argument largely...") Â |
No edit summary |
||
Line 1: | Line 1: | ||
CHAPTER II: Getting In | |||
<blockquote>Among economists, though, this question has long been a matter of debate. Proponents of a more laid-back approach to college admissions base their argument largely on a 2002 paper, updated and extended in 2011, by Stacy Dale, a researcher, and the late economist Alan Krueger. Dale and Krueger analyzed records from twenty-seven selective colleges and universities, as well as a national database, and concluded that for well-off white and Asian students, there was no benefit at all, in terms of future earnings, to attending a more selective college. Yes, a graduate of Princeton will earn more than a graduate of Penn State, on average--but, the researchers said, that has everything to do with the talents and ambitions of the students whom Princeton admits.</blockquote> | <blockquote>Among economists, though, this question has long been a matter of debate. Proponents of a more laid-back approach to college admissions base their argument largely on a 2002 paper, updated and extended in 2011, by Stacy Dale, a researcher, and the late economist Alan Krueger. Dale and Krueger analyzed records from twenty-seven selective colleges and universities, as well as a national database, and concluded that for well-off white and Asian students, there was no benefit at all, in terms of future earnings, to attending a more selective college. Yes, a graduate of Princeton will earn more than a graduate of Penn State, on average--but, the researchers said, that has everything to do with the talents and ambitions of the students whom Princeton admits.</blockquote> | ||
*** | |||
<blockquote>Hoxby has focused her research for more than a decade on the college-admissions process. But for many years, she avoided the debate over the relative value of attending a selective college. She was always somewhat skeptical of the methods that Dale and Krueger used, but she considered the available data too muddy either to confirm or rebut their conclusion. More recently, though, that muddy picture clarified for Hoxby, when she found a way to leverage many years of tax records--using the same IRS database that Chetty draws on for his research--to calculate "value-added" estimates for thousands of American colleges and universities. Â | <blockquote>Hoxby has focused her research for more than a decade on the college-admissions process. But for many years, she avoided the debate over the relative value of attending a selective college. She was always somewhat skeptical of the methods that Dale and Krueger used, but she considered the available data too muddy either to confirm or rebut their conclusion. More recently, though, that muddy picture clarified for Hoxby, when she found a way to leverage many years of tax records--using the same IRS database that Chetty draws on for his research--to calculate "value-added" estimates for thousands of American colleges and universities. Â | ||
Line 6: | Line 10: | ||
Her conclusion? Dale and Krueger were simply wrong. High-prestige colleges do pay off for the students who attend them, and in fact they pay off in a big way.</blockquote> | Her conclusion? Dale and Krueger were simply wrong. High-prestige colleges do pay off for the students who attend them, and in fact they pay off in a big way.</blockquote> | ||
*** | |||
<blockquote>CollegePoint counted an advising relationship as a success only if the student ended up enrolling in an Aspen 270 college. But the advisers I spoke to found the Aspen list to be rigid and arbitrary; in the real world of high school students, they believed, sometimes a school that wasn't on the list was a better choice than a school that was. For instance, the only historically black college on the Aspen 270 list was Spelman, a women's college.</blockquote> | |||
CHAPTER III: Fixing the Test | |||
<blockquote>It was true that studying for twenty hours on Official SAT Practice was associated with an average score gain of 115 points. But what the College Board didn't tell the public was that very few students had studied that many hours. The College Board sent invitations to 1,075,000 students in the high school class of 2017, encouraging them to link their College Board account to a free account on Khan Academy. Fewer than a quarter of those students -- 250,000 -- actually linked their accounts. Most of those 250,000 students studied for just a few hours, or not at all. Only 8,000 students across the country did at least twenty hours of practice and achieved that average gain of 115 points... | |||
This is where dealing with the College Board grew especially frustrating--and my faith in the organization's good intentions began to wear thin. If only a tiny percentage of students studied for twenty hours, why lead your public presentation with that group's results? When I spoke later than summer with Aaron Lemon-Strauss, the College Board executive who led the collaboration with Khan Academy, he explained that the decision had been driven by a desire for positive media coverage. There was an original draft of the press release, Lemon-Strauss explained, that featured a chart comparing the effect of zero hours of Khan study (an average gain of sixty points over students' PSAT scores), with the effect of five hours (an average gain of eighty-four points) and ten hours (an average gain of ninety-seven points). That chart would have indicated much smaller gains than the numbers the College Board eventually released, but they would have been more representative of what typical students actually achieved using Official SAT Practice. | |||
Lemon-Strauss told me he liked that version, but he explained frankly that it was rejected by the College Board's communications team.</blockquote> | |||
*** | |||
<blockquote>And because the effect of an hour of study on a student's score gain was consistent among all these groups, the extra hours of study that the more privileged students took advantage of meant that their scores increased more, as well. On the whole, according to the data from the hackathon, Official SAT Practice on Khan Academy had increased the existing gaps between students in groups that tended to score higher on the SAT and students in groups that tended to score lower.</blockquote> | |||
CHAPTER IV: Fitting In |
Revision as of 14:22, 30 April 2021
CHAPTER II: Getting In
Among economists, though, this question has long been a matter of debate. Proponents of a more laid-back approach to college admissions base their argument largely on a 2002 paper, updated and extended in 2011, by Stacy Dale, a researcher, and the late economist Alan Krueger. Dale and Krueger analyzed records from twenty-seven selective colleges and universities, as well as a national database, and concluded that for well-off white and Asian students, there was no benefit at all, in terms of future earnings, to attending a more selective college. Yes, a graduate of Princeton will earn more than a graduate of Penn State, on average--but, the researchers said, that has everything to do with the talents and ambitions of the students whom Princeton admits.
Hoxby has focused her research for more than a decade on the college-admissions process. But for many years, she avoided the debate over the relative value of attending a selective college. She was always somewhat skeptical of the methods that Dale and Krueger used, but she considered the available data too muddy either to confirm or rebut their conclusion. More recently, though, that muddy picture clarified for Hoxby, when she found a way to leverage many years of tax records--using the same IRS database that Chetty draws on for his research--to calculate "value-added" estimates for thousands of American colleges and universities.
Her method, briefly, involves combining the IRS data showing which college each student attends and their earnings a decade after graduating from high school with information from the College Board on where students apply, plus data from a national clearninghouse that reveals which college each student actually winds up attending. By merging these data points, and comparing them for millions of students who graduated from high school between 1999 and 2003, Hoxby was able to estimate, quite precisely, the causal effect of the college a student attends on his or her average lifetime earnings.
Her conclusion? Dale and Krueger were simply wrong. High-prestige colleges do pay off for the students who attend them, and in fact they pay off in a big way.
CollegePoint counted an advising relationship as a success only if the student ended up enrolling in an Aspen 270 college. But the advisers I spoke to found the Aspen list to be rigid and arbitrary; in the real world of high school students, they believed, sometimes a school that wasn't on the list was a better choice than a school that was. For instance, the only historically black college on the Aspen 270 list was Spelman, a women's college.
CHAPTER III: Fixing the Test
It was true that studying for twenty hours on Official SAT Practice was associated with an average score gain of 115 points. But what the College Board didn't tell the public was that very few students had studied that many hours. The College Board sent invitations to 1,075,000 students in the high school class of 2017, encouraging them to link their College Board account to a free account on Khan Academy. Fewer than a quarter of those students -- 250,000 -- actually linked their accounts. Most of those 250,000 students studied for just a few hours, or not at all. Only 8,000 students across the country did at least twenty hours of practice and achieved that average gain of 115 points...
This is where dealing with the College Board grew especially frustrating--and my faith in the organization's good intentions began to wear thin. If only a tiny percentage of students studied for twenty hours, why lead your public presentation with that group's results? When I spoke later than summer with Aaron Lemon-Strauss, the College Board executive who led the collaboration with Khan Academy, he explained that the decision had been driven by a desire for positive media coverage. There was an original draft of the press release, Lemon-Strauss explained, that featured a chart comparing the effect of zero hours of Khan study (an average gain of sixty points over students' PSAT scores), with the effect of five hours (an average gain of eighty-four points) and ten hours (an average gain of ninety-seven points). That chart would have indicated much smaller gains than the numbers the College Board eventually released, but they would have been more representative of what typical students actually achieved using Official SAT Practice.
Lemon-Strauss told me he liked that version, but he explained frankly that it was rejected by the College Board's communications team.
And because the effect of an hour of study on a student's score gain was consistent among all these groups, the extra hours of study that the more privileged students took advantage of meant that their scores increased more, as well. On the whole, according to the data from the hackathon, Official SAT Practice on Khan Academy had increased the existing gaps between students in groups that tended to score higher on the SAT and students in groups that tended to score lower.
CHAPTER IV: Fitting In