There was a flurry of re-shares last week for this article about Yale shutting down a site that aggregated student course evaluations, which is fine as far as it goes, but repeats a stat that really bugs me:
About 43 percent of college letter grades in 2011 were A’s, up from 31 percent in 1988 and 15 percent in 1960, a 2011 study found. Over roughly the same span, the average amount of studying by people enrolled in college declined almost 50 percent, a 2011 study found, from 25 hours per week to 13 hours.
This is less bad than the usual, thanks to including the 1988 point, but it's still kind of deceptive, suggesting a steady monotonic increase in A's from 1960 to the present. In fact, though, the historical trend is lumpier, as you can see in the "featured image" at the top of this post, which I've blogged about before. In fact, nearly half of that increase occurred very rapidly, during the Vietnam era-- The fraction of A's increased from 15% to over 30% in just 14 years-- in fact, the 1988 fraction is probably slightly less than the early-70's peak. But saying "increased from 33% in 1974 to 43% in 2011" doesn't sound as dramatic as "increased from 15% in 1960 to 43% in 2011."
Now, as I said, the recent piece is better than a lot of others in quoting one intermediate point, and you can argue that if you drew a straight line from 1960 through 1988 you'd hit the 2011 point, or close to it. But then, if you look at the trend in the early 1960's, before things really go nuts around 1966, and extrapolate that you would also come pretty close to the modern number. Which kind of undercuts the claim that any modern phenomenon is to blame-- back in the idyllic days when the professors who taught my professors were walking uphill through the snow to adhere to rigorous classroom standards, grades were creeping upward at more or less the same rate they are now.
This is not to say that the core point of the Post blog is wrong-- there are lots of reasons to think that student opinion surveys are a lousy measure of the quality of a course, and lots of reasons to think that Yale was kind of stupid about what they did. But the grade inflation stat, as usually presented, is not the strong argument for this that people seem to think.
As noted above, I've made this point before, but it seemed to surprise a bunch of folks on social media when I mentioned it last week, so it's worth repeating. If you want to see what dramatic grade inflation looks like, you need to look at the college days of the Baby Boomers, not the kids these days with their baggy pants and hip-hop music who won't stay off the damn lawn.
- Log in to post comments
What is an A anyway?
Percentage I understand, but the meaning of an A grade is different in different countries, and can be different within the education systems of individual countires. This is to say nothing of non-educational systems which use such grades (AAA+ bonds anyone? BB-?)
So what is the expected percentage attainment in exams which should lead to an A? 90%? 85%? 80%? Is this the percentage for which an A grade is in practice awarded after "fitting to the curve"? I suspect not in practice.
Grade inflation and the main secular(as in time) forces behind it are a problem. But part of the problem is the nebulous nature of grades to begin with. Until you nail down what an A -- and more importantly what a B, C, and D -- should be, you can't begin to define what the problem truely is.
While these schools have expanded over that time, it seems doubtful they have expanded as much as the population meaning they would be taking more cream from the top, so how much of this even represents grade inflation? One might assume some improvement in the quality of education over time as well, or do people expect grades should correspond to percentiles of population regardless of quality? At the same time, knowledge has expanded greatly meaning they probably know less as a whole than ever before even if more than those before them so should they be downgraded on this account?
If you plot the number of men who can run a mile in four minutes it will show a steady increase from 1954. Does this mean that the mile is getting shorter? No, it means that training methods are getting better. Is it possible that students today are simply better trained for writing exams, or perhaps more motivated to do better than their elders?
Perhaps things were easier in the past. My children certainly have more homework at a younger age than I ever had.
If you plot the number of men who can run a mile in four minutes it will show a steady increase from 1954. Does this mean that the mile is getting shorter? No, it means that training methods are getting better. Is it possible that students today are simply better trained for writing exams, or perhaps more motivated to do better than their elders?
Perhaps things were easier in the past. My children certainly have more homework at a younger age than I ever had.
While these schools have expanded over that time, it seems doubtful they have expanded as much as the population meaning they would be taking more cream from the top, so how much of this even represents grade inflation?
Since the 1990's at least, the fraction of the college-age population (18-24-year-olds) has been steadily increasing-- see, for example, this fact sheet. That means that the college population has been increasing faster than the general population, so if anything we would need to be dipping lower into the milk, not just skimming more cream.
The Vietnam war compromised honest grading. An honest evaluation could kill young males. Add to this DIVERSITY whose singular qualification is objectively proven inability,
http://dailycaller.com/2013/11/26/prof-corrects-minority-students-capit…
"all knowledge is subjective and based on one's position in society"
http://www.amren.com/news/2014/01/brooklyn-school-cutting-gifted-progra…
Either "fairly"distribute earned grade points, or only fail the proven able. Socially-mandated stupidity plays a pat hand.
But there aren't any more Yales than there were, so one would expect Yale to be drawing from a larger population. There are more new schools including places like Phoenix U and community colleges that would not have had much of a prior history and these would not necessarily use the same class weighting. We don't have B schools or C schools, each wants to rank its students according to its entrants. Growth at the lower rankings of colleges will increase the number As without representing inflation. Yale does give out more As, but does it give out more As as a proportion of population, not of its entrants, but as of all college entrants?
It is not that surprising after all. The boomers were about to enter college which needed to staff up to handle them so they had to give out more As to provide staff for this expansion, which then remained the standard until the late 90s.
The boomers were about to enter college which needed to staff up to handle them so they had to give out more As to provide staff for this expansion
Except that the expansion in question happened mostly before the Boomers were of college age. The golden age of university expansion in this country was a roughly 20 year period after World War II, fueled largely by returning soldiers attending college on the GI Bill. That would be the parents of the Baby Boom generation. The era of expansion was pretty much over by the end of the 1960s (part of why the physics job market went so abruptly from "faculty positions for anybody who wants one" to "no faculty positions available, except maybe if you have a Nobel Prize, and even then no guarantees"). The oldest Baby Boomers only entered college in 1964. The beneficiaries of this expansion were mainly from the Silent Generation, who went through college before the big grade inflation of the late 1960s.
Your first paragraph is more on point. Some of the elite private universities have expanded capacity, but not nearly by enough to keep up with the expanding college-age population, so these schools can be ever more selective about who gets in the door. For many of these schools, fewer than 10% of applicants are offered admission. (For my own undergraduate alma mater, the admit rate for this year's freshman class was about 1 in 12, compared to about 1 in 3 for my freshman class.) It's true that high school seniors are applying to more colleges, thanks to internet applications: 12 is typical these days, versus 6 for my cohort (who had to send checks along with applications sent by snail mail, as well as provide postage for the envelopes their teachers would send recommendations in). But that increase is also not enough to account for the drop in admissions rates, so assuming that the skill level of the two cohorts on entering college was similar, todays elite university students really are better, on average, than students enrolling in those same universities in my day.
Given the Flynn Effect, US college students, like other people their age, are smarter, on average, in terms of school than their grandparents. Unless college courses have become tougher, we should expect their grades to be higher.
Expansion of colleges or expansion of college population? The boomers would have greatly expanded demand even if no more likely to attend college. The demand for professors would have peaked before this since they would have to graduate before the boomers to teach them and upcoming demand encountering limited supply would induce lowered standards. The increased supply of boomers would tend to tighten standards since they could not expect as great a demand following them, so one should expect the inflation to precede the increase in population.
What is an A anyway?
If you plot the number of men who can run a mile in four minutes it will show a steady increase from 1954. Does this mean that the mile is getting shorter? No, it means that training methods are getting better. Is it possible that students today are simply better trained for writing exams, or perhaps more motivated to do better than their elders?
Perhaps things were easier in the past. Percentage I understand, but the meaning of an A grade is different in different countries, and can be different within the education systems of individual countires. This is to say nothing of non-educational systems which use such grades (AAA+ bonds anyone? BB-?)
So what is the expected percentage attainment in exams which should lead to an A? 90%? 85%? 80%? Is this the percentage for which an A grade is in practice awarded after “fitting to the curve”? I suspect not in practice.
Grade inflation and the main secular(as in time) forces behind it are a problem. But part of the problem is the nebulous nature of grades to begin with. Until you nail down what an A — and more importantly what a B, C, and D — should be, you can’t begin to define what the problem truely is.