Grade Inflation, Covid, Smartphones, and AI

There has been a bit of a media explosion in America of late with articles about the unprepared state of the university student.  Here we try to unpack things a bit.

Grade inflation has been a growing pestilence that has its roots in university professors increasing the grades of young men in the 1960s to keep them out of South East Asia.  This accelerated over Covid as schools chose not to admit the fact that kids didn't learn much of anything over that period and have profound gaps in all academic domains.

Now, grades don't mean much of anything.  As this important report from UC San Diego reveals, enrolment in the most remedial of math classes has skyrocketed even as students enter university with straight As. So there are really a number of things coming together:

1. A long-term trend of grade inflation and with many grades based on "effort" rather than actual mastery.  Score a B?  You can re-test or write a supplemental essay to make that into an A...  Teachers who give lots of Bs and Cs don't get good reviews, and might be out a job.

2. Covid accelerated this trend and has resulted in high school graduates with straight As who can't read properly or solve 3x + 5 = 17.  AP test scoring has been "recalibrated" to ensure that kids continue to take them as they are the source of over half of College Board's US$1 billion revenue.  (Kudos to the IBO that raised standards back to pre-Covid levels.)

3. Smartphone usage has proliferated, and laptop use is ubiquitous, which have reduced attention span in additional to any number of other negative externalities.

Now... Enter the AI.  This is the first year of university students who have had ready access to a form of AI since their first year, and the early reports are dire.

Despite the noble efforts to detect and deter flagrant AI cheating (this is our favorite) a majority of university students now use AI in one form or another daily.  While AI will be part of our lives, using it to complete the bulk of university assignments brings in to question the very point of attending university.  Doing so is like going to the gym, and paying a robot to lift weights for you.  Students are increasingly more interested in networking and getting the right summer internship than in actually reading anything, perhaps because they are keenly aware that AI will be taking away many entry-level consulting and finance jobs...

Universities are slowly addressing the reality and blue book exam booklet sales are booming as there is a movement back to hand-written assessments.  We don't have a solution for any of this beyond a call to get technology entirely out of the classroom and a return to requiring the reading David Hume and Karl Marx - not likely. Sweden had taken the lead in embracing all manner of technology but now is very quickly running back to spiral notebooks.  Hopefully more will follow.  The alternative is a bleak "post literate society".

(Rant over.)  (No AI was used in writing this.)

Please refer to the our blog for more information as well as our LinkedIn feed for more updates from the surprisingly active world of testing and admissions.

The Downside of ED

Early Decision (ED) is under a bit of legal peril but it will be around for a while anyway.  ED allows students to apply Early (it's in the name) and get a Decision (!) well before the Regular Decision (RD) deadlines.  The expectation is that students accepted under ED will commit to attending said university well before Christmas.  This aids universities in filling dorms in a very predicable manner and hence bumping up their yield percentages.  Yield is the percent of accepted students that actually attend the school and is a core component in most rankings.  The policy also lets families complete the entire application process before the end of the calendar year and not apply to 18 other universities. 

Sounds great for all concerned, doesn't it? Nope. 

Students seldom pay the full sticker price for university and the price of the product is a core consideration in choosing the university for students in the RD cycle.  A student might get into 5 schools and the family would then have some degree of bargaining power to extract the best "package" from the university.  This would help inform the monumental decision of which school to pack off to for four years.  Students committed to ED do not have such privilege as they are expected to attend the ED school if accepted.  This makes the policy clearly biased towards the very wealthy for whom cost isn't really a concern (and who can afford expensive test prep).  University rankings probably take a bit of a hit through marginally inflated admit rates, but the financial imperative of getting bums in seats early and in a predictable way more than outweighs that minor concern.  Yield percentages have become a bit unpredictable (see below) so getting students "double-confirmed" is a great idea for schools.

An ED arms race of sorts has emerged with many competitive universities leaning very heavily on the policy to fill over half of their entering class.  The admit rate for ED applicants is often over triple the RD rate so if a student really has university A as a first choice, it would be folly not to apply ED.  Unless, the family doesn't have the financial means to potentially pay full fare.  

We think the policy should be banned as it is clearly discriminatory in a larger system that is rife with similar barriers to entry for the non-plutocratic majority.  Not all schools have ED and many have a dizzying, shifting array of non-restrictive Early Action (EA), EA II, RD, REA, etc. that you need to read about carefully on the school websites.   As always, check your in-school university advisor to understand everything as best you can.  They are the ones that have the unenviable task of keeping up with all this.

All we can do here is share some data we have compiled.

Covid's Impact on American University Admissions

We have done some fairly involved analysis of data available from university Common Data Sets, yearly statistical snapshots released by most universities covering enrolment and admissions.  A graphical narrative of our story can be found here.  (Note this tale is most relevant for the 40-50 American schools that are, for some reason, most popular among domestic and (especially) international students.  The overwhelming majority of American universities are excellent, and not competitive to get into.)

The basic narrative is as follows:

- A small number of American universities were becoming harder and harder to get into as there was a slow and building race to perceived "quality" among applicants.

- Covid happened and all schools shifted to a test optional as they couldn't require a test that kids couldn't take.

- This led to an appreciable increase in the number of applications to said schools as many students who would previously have been reluctant to apply due to low test scores went ahead and had a try.

- Universities largely did not increase enrolment and hence the admit rate among these schools dropped even faster than pre-Covid trends.

- Admit rate is a key metric of the many university ranking systems that have far more influence than is merited.  

- The majority of students enrolling in these schools did submit test scores, though the actual admit rates with and without scores is not public.

- Universities know that if they start requiring tests then there will be a drop in application numbers, a rise in admit rates, and fall in their rankings - perhaps leading to a cycle of even lower application numbers, rinse, repeat.

- A few universities have reverted back to test required based on several fairly weighty studies that indicated test scores are a better indicator of potential collegiate success compared to GPAs that have been further inflated by Covid.  These have generally been schools that don't need to worry about application volume as their high rankings are assured (Harvard, Yale, MIT, CalTech, etc,).

- Universities in the test optional camp often list test scores as "important"" in admissions decisions yet continue to have test "optional" as their stated policy.  Are test scores important, just as good grades, academic rigor, and CCAs are also important?

- High schools do not know which students submit test scores and which students do not, and universities do not reveal this.  This makes the admissions process more opaque and confusing for all concerned as is makes the question "do I submit scores" into one that is sometimes impossible to answer.

To help families better understand the admissions process, we ask that universities reveal the following information:

- International, in-state, and out of state admissions rates (many are now revealing this!)

- Admit rate for students that submit scores and for those that do not.

Universities will not start doing this until based on the words of a test prep provider.  We call on all students, parents, and university advisors to ask every single university representative they meet to reveal this information.  This will help families make better sense of the admissions process and make more informed decisions.

Domestic vs. International Admit Rates

Some universities have started including international admit rates in their Common Data Sets! This is exciting to data nerds like us and confirms what we already know: admit rates for foreign students are much lower than domestic, sometime much lower. It would also be very interesting to know how many of the 229 accepted international students are Princeton were helped by "institutional priorities", such as athletics.

We haven't included all the data but some schools get thousands of applications from international students but as the data show, only admit less than 4% of those students. Drumming up applications from international students who realistically have a small chance of admission is a great way to lower headline admit rates, as well as a decent source of revenue at US$100 per application or so.

Sadly, many schools do not report this data as the common data set initiative is technically optional. Please ask every university representative that you speak to from schools that don't report, "why are you being disingenuous (SAT word!) and not revealing data like your peers do?" Hopefully more will report in future, but they certainly will need to be pushed.

More on Testing Trends

 The number of applications to the most rejective American universities has been growing for the last 20 years.  Interestingly, the number of Americans going to university is reaching a demographic peak, but the number of applications has been growing as students apply to more universities, and more students from abroad also apply - many of whom might not necessarily enrol.

Universities will never tell a student not to apply.  The more applications received, the lower the admit rate, and the lower the admit rate, the higher ranked a school will be.  These silly, subjective, flawed rankings have led to an arms race of sorts where universities seek applications from around the globe to "keep up" with the school down the road that is doing the same. 

Covid resulted in universities largely adopting a "Test Optional" policy as they couldn't realistically ask for a test that students couldn't take.  This led to many students applying to MORE schools than before, including those that they probably have zero chance of getting into.  Test scores used to act as a rough filter of sorts to reduce frivolous applications, but if the test is "optional" then why not send in an application from a straight A student? (A majority of students at many schools are "straight A students" now...)

This accelerated the growth in application numbers further with those to the most popular schools surging upwards of 20% from 2020 to 2023.  The number of places has been fixed so the result is even lower admit numbers as the number of applications to schools like NYU have more than doubled in 10 years.  All the while, as previously reported, the majority of those that get into said schools are submitting SAT scores.  Test "optional" is actually test "preferred".

Switching away from test optional would result in a sudden dip in the number of applications received, a rise in the admit rate, and a fall in the rankings.  Only a select few schools have been brave enough to adjust their policy to reflect the reality on the ground - Cornell being the latest on April 22nd.  We expect more to follow, but it is a collective action problem with universities effectively side-eyeing their peers.  It is too late now for any other changes for this coming 2024-2025 admissions cycle, but more schools are sure to follow.  Perversely, some schools might follow as doing so would group them with the very rejective schools that have already done so and perhaps help with application numbers!

An excerpt from our related research drawn from university Common Data Sets with charts and graphs can be found below.

We will be mining this data for more thoughts in the future.  For now, please look at schools outside of some arbitrary ranking table or those that are "popular" at your school - only about 75 schools in America are actually hard to get into!