"Admissions officers worried that their colleges were replacing standardized tests with metrics that were even more biased toward wealthier and white students, such as letters of recommendation and expensive extracurricular activities. One college purchased a data service that ranked high schools and factored those high school rankings into each application. Students from underserved high schools received a lower ranking, an admissions officer explained. It wasn’t a fair process.
Many admissions officers said that they were struggling with how to select candidates fairly and didn’t know how to weigh an application with test scores against one without. “I think the students that do have the strong test scores still do have that advantage, especially when you have a student that has strong test scores versus a student who doesn’t have test scores and everything else on the academics is more or less the same,” an admissions officer told Slay."
Curious what you all think…Do test optional policies make applications better or worse representations of who a student really is? In other words, do they make admissions more or less fair?
We have had the discussions about test utility on other threads, but want to point out the data from this report is based on 16 schools, and the final analyses aren’t yet published.
So, imo there’s not enough data to make any observations about TO decision making at the entire set of colleges.
While that’s true, this problem was identified by MIT when it reimposed its testing requirement. I do not know whether the 16 schools in this study were representative of others, but I think many selective schools face a similar quandry. I never understood the equity argument for abandoning test scores, so we could rely instead on factors even more skewed to SES status.
If you don’t want to take the test or if you test under average, you still have a chance.
If you smoke the averages and it’s a test optional school, you’re potentially in a better position.
If it’s test blind - then there’s no advantage to have.
Now - if you are saying does a high GPA but low test foreshadow academic issues - many schools have said that - hence the importance of the test. But no one wants to be told up front- that they’re not going to be a good college student, etc.
I’ve had the benefit of reading many posts by counselors and AOs about this article this week. Many schools invested quite a bit of time/money in training staff to read apps without test scores.
It’s still reasonable that some AOs are struggling. But, getting more eyes to read an app (to take one example in the article) is probably a good thing.
I’m not sure there’s agreement that grade inflation is more skewed to SES (even tho some data point that way there are plenty of examples that don’t), so will leave that discussion to the people who are experts in admissions decisions!
makes no sense to me to abandon using test scores but taking into account the demographic and geographic contexts. it’s just an excuse for colleges to accept whoever they want and not be scrutinized for it.
I am no expert on this question, but I am having a very tough time reconciling grade inflation and test-optional/no testing.
I know that there are studies that say standardized tests are not good predictors of performance in college. But if grades are inflated, how are grades a good predictor?
Perhaps I am misguided, but the standardized tests, IMO, are a consideration, but just one of many considerations. An admissions committee (at least in private schools) can determine for itself what weight to give to these test scores. For example, if a student has a high GPA and stellar LORs indicating that s/he is a fabulous student, but has a low test score, isn’t that enough for a college to say that they can “make the grade”, both literally and figuratively, at the college? Perhaps that student doesn’t test well, but there are other indicators that the student will do well. But why not have that test score as a consideration, just not the only consideration?
As I have said before, colleges should either do away with the test requirement altogether (e.g. U of Cal system) or require it (e.g., MIT). This TO “middle ground” seems to cause confusion for students and their parents, while allowing some colleges to game the system by encouraging submission of high scores to bump up their median scores and, ostensibly, their “elite factor.”
TO is better for some and worse for others, now that we are in a post-covid-crisis era when tests are easily accessible again(completely different story the year-ish that testing was not readily available in many places).
For kids with scores lower than the range their dream school accepts, it gives them a better chance than they had when tests were required, if they have the other factors(rigor, gpa etc).
For kids with scores within average range of the typically accepted students, they have less of a chance now if they submit, as only about 60-75% of “slots” tend to be for kids with scores, depending on schools. And, they have the added doubt of when to submit and when to not, due to confusing and vague wording depending on the school.
For kids with sky-high scores, at or above every school’s 75th%ile on Verbal and Math, it probably makes little to no difference in their chances compared to pre-TO, since this super high scoring range tends to be admitted at higher rates anyway, even at tippy top schools, presuming all the other factors are consistent with being a top kid.
Personally, for top-25ish/top-LAC type schools, since that is primarily what our family has considered, I agree with MIT’s assessment and decision and wish they all went back to requiring them, with the obvious caveat that they are just one factor in holistic admissions and have to be considered in context.
Those studies claimed that standardized test scores alone weren’t as good a predictor as high school grades alone. There’re multiple issues:
When the studies were carried out, high school grades weren’t as inflated.
No one has ever denied that standardized tests enhanced the predictability of college success.
Since grades in college are also inflated and uneven (among courses with different levels of rigor), college GPAs or graduation rates alone aren’t sufficient to measure college success.
I agree with that. If a college thinks test scores aren’t useful, it should ignore them completely. Or it should require them if it thinks they’re helpful. If tests are optional, colleges would have to treat their applicants who submit differently from those who don’t. It makes admission decisions more arbitrary.
If you look at how well different metrics of application evaluation correlate with SES, nearly all are less correlated with SES than test scores. For example, I’ve previous linked to the Ithaca study at https://www.ithaca.edu/wpm_preview/site/ir2/docs/testoptionalpaper.pdf . Using ALANA and first gen as a proxy for SES, in lists the following correlations, ranked from strongest to weakest. Scores were more correlated with income than both GPA and measures of course rigor.
SAT W: -0.15 fist gen, -0.13 ALANA
SAT (M+V): -0.13 ALANA, -0.08 fist gen
Number of AP Credits: -0.12 ALANA, -0.03 first gen
HS GPA: -0.08 ALANA, +0.01 first gen
Rigor of HS Courses: -0.04 ALANA, +0.05 first gen
Using, LDC hook as proxy for high income, the Harvard lawsuit analysis suggests scores are also more correlated with income than EC rating, LOR ratings, etc.
Studies of test optional colleges consistently show non-submitter admits at test optional colleges who are admitted based on the only the additional non-score criteria average lower income than test submitter admits at the same college.
Perhaps, @Data10. But that is hard to square with the Hechinger report, and the other report of LACs which were test optional for 10 years, concluding that it made little difference
The article doesn’t evaluate the correlations with income listed in my post. The report instead links to the study at https://journals.sagepub.com/stoken/default+domain/RMRIR6XS4QICVZDDIAZM/full , which states the following in the abstract. The linked study found Pell Grant totals increased 3.2% at more selective colleges and 3.6% at less selective colleges, hence the “3 %to 4%” range stated in the abstract. The HechingerReport article calls this 3% to 4% range only a 1% increase in low income students, . They also call the 10-12% increase in URMs listed in the abstract a 1% increase. This should give some idea about the accuracy of the news article, which seems to be published on a small, non-mainstream media type website.
I find that test-optional policies were associated with a 3% to 4% increase in Pell Grant recipients, a 10% to 12% increase in first-time students from underrepresented racial/ethnic backgrounds, and a 6% to 8% increase in first-time enrollment of women.
Well, then I am glad it works for them. Might be more efficient to utilize low-income specific programs like Questbridge instead to achieve that goal, but whatever.
At selective colleges, the college can adjust the portion low income as needed, regardless of test optional/required policies. For example, if a selective college wants to increase portion of students that are Pell grant they can give a greater boost for Pell grant (if it is part of evaluation) or other criteria that are well correlated with Pell grant, even if they stay test required.
Similarly if a selective college wants to hold Pell grant % and full pay % at a roughly constant value upon going test optional, this can be accomplished fairly easily, particularly at need aware colleges. Going test optional doesn’t guarantee a change in income distribution of the student body. However, one can expect that the kids who are admitted test optional will average lower income than the kids who are admitted test required at that college, even if income distribution of full student body doesn’t change.
Questbridge probably picks up disproportionately what some call the “privileged poor” who are in environments where college prep starts early enough for them to accumulate all of the needed application items to make the very early Questbridge application (as opposed to those with little access to counselor, teacher, parent, and peer knowledge about applying to college other than the nearby not-very-selective state schools that accept late rolling applications, high school records, and test score reports if required).
Of course, as @Data10 writes above, colleges have plenty of other ways to adjust their admission class for more or fewer lower SES admits by how much they consider correlates of SES (e.g. legacy correlates to higher SES, while first-generation-to-college correlates to lower SES).
I think it’s very hard for an evaluator to “unsee” the scores that the student has when comparing between students.
So take two students and one with a 34 Act and one test optional. Ecs are the same and recs the same. Unless the higher score students recs or Ecs or essays are really bad or not up to par, I would think the one with the sent in scores would fair better.
I think the discrepancy between the 3-4% vs. 1% reporting and the 10-12% vs. 1% reporting is not due to Hechinger being inaccurate, but rather due to Hechinger reporting the absolute % change instead of the relative % change.
This is what the report you linked to said:
“In addition to the relative changes for individual student groups, it is also worth considering the implications for the student body composition as a whole. As shown in Table 1, the proportion of Pell Grant recipients and URM students at matched institutions during the pretreatment period is comparatively low. As a result, the estimated increases in Pell Grant recipients and URM students amount to relatively modest gains as an absolute share of the student body. Specifically, these estimates correspond to an increase of roughly 1 percentage point in terms of both the share of students receiving a Pell Grant and the share of students from URM backgrounds. With women accounting for the majority of students at private institutions, their enrollment shift following test-optional policies represents a larger absolute increase—amounting to approximately 4 percentage points as a proportion of all FTFT students.”
There were many schools that were test-optional before COVID and they were test-optional for a reason. They were schools that felt holistic measures gave them the student body that did well at their school. Now many other schools were able to (forced to ) give it a try. I’m sure some will stick with it and some won’t. As someone that always favored test-optional schools, I will be interested to see who stays test optional. As for better/worse, the good thing is that students have a choice to apply to test-optional schools if that’s what they so choose. They just have to be a bit more flexible when making their lists.
I’d love to see schools take an even deeper dive into whether they stay test-optional or not. For example, while large state schools may find that test scores are very important to predict the successful completion of an engineering degree, they may also find that it has no correlation or an inverse correlation to other degrees such as a BFA, business or education degree.
While I know this is an old thread, I wanted to share an observation.
TO just kicks the can down the road.
We were lucky than when we went thru the college application process, our kids (3) all had to take standardized tests, and combined with their good grades in rigorous programs, they were prepared for college and thrived there. TO took hold just after all our kids were already in college.
When they graduated (one is still in undergraduate), they pursued STEM careers and employers were beginning to conduct their interviews via a series of tests, where the candidate had to film themselves (to illustrate they took the test without assistance). Only after passing these tests and proving competence/mastery of the subject matter, did a job offer follow.
I understand this level of testing has become commonplace.
Tech companies are not TO and I believe so because the need to know what they are getting.
The tests used in those hiring processes are very subject matter specific, not multiple choice tests used for all applicants to all types of jobs.
Closest resemblance would be exam questions in relevant college courses. Closest high school standardized test resemblance would be free response AP test questions for the most relevant AP course.