Is the SAT fair and most people don’t know the content, or is the SAT unfair because its racial bias

I’ve heard peoples’ opinions about the SAT, from it being purely inaccurate of one’s college readiness to it being an accurate, fair college admission test no matter one’s background. Though this, I believe that it’s a fair test, but I just wanted to ask if anyone else had similar or different opinions and why?

I feel it’s a fair and very necessary test. I say this as a minority.

Whether or not you think the SAT is fair or has racial bias itself, its results will include the effect of existing unfairness or racial bias that had other sources.

1 Like

The critiques of the SAT/ACT revolves around its use as a college acceptance tool. However, it would seem that there has to be some mechanism to understand where students’ knowledge stand relative to one another. This is the classic, “don’t throw the baby out with the bathwater”, scenario.

1 Like

Increasingly, the negativity about it is that it measures how well one prepares for it and access to preparation is uneven. Just read these boards and you’ll see how much goes into studying, tutoring, retaking, etc. And yes, the scores go up.

When some of the original concerns about the test were raised decades ago, test prep was non-existent. But the word analogies were classist. Efforts were made to change the test. Yacht: stern came off.

Personally, I feel like it could be a useful data point given how different high schools are. It indeed measures something - the question is how valuable that is to predicting college readiness and success. I would guess that students applying to engineering programs will still need some validation of their math abilities.

Remember that as schools become TO, it doesn’t prevent students who do well on these tests to submit scores.

1 Like

It’s not a simple choice of either SAT is purely inaccurate or purely accurate prediction of college readiness. Instead SAT has a non-zero degree of correlation with college readiness and related college outcomes. The magnitude of that non-zero correlation depends on a variety of factors.

For example, the study at https://journals.sagepub.com/doi/full/10.1177/2332858416670601 compares the predictive ability of SAT, state achievement tests, and HSGPA in predicting measures of college success among ~20k kids in the NY and Kentucky public school systems. Results are summarized below.

10k Kids in CUNY System FY GPA
SAT alone explains 14% of variance in first year GPA
HS NYS Regents test explains 16% of variance in first year GPA
HS GPA explains 25% of variance in first year GPA
HS GPA + SAT explains 28% of variance in first year GPA

10k Kids in Kentucky Public Colleges FY GPA
SAT explains 16% of variance in first year GPA
HS KCCT test explains 17% of variance in first year GPA
HS GPA explains 32% of variance in first year GPA
HS GPA + SAT explains 34% of variance in first year GPA

In this study, SAT alone explains ~15% of variance in FY GPA,. The ~15% variance explained is statistically significant, but it is still only explaining a small portion of first year college success. In this study, SAT alone was also less predictive than both the state achievement tests and HS GPA in isolation. The combination of HSGPA + SAT was better than either alone – explaining 2-3% more variance in FY GPA than HS GPA alone.

This 2-3% increase in variance explained beyond HS GPA alone is statistically significant, but it comes at a cost since SAT in isolation is far more correlated with wealth than HS GPA in isolation. The correlation with wealth partially relates to wealthy kids truly being better prepared for college on average, but it also partially relates to things like wealthy kids on average having better prep, taking the test more times, ,more likely to have special accommodations, more likely to attend a HS that teaches to the test, etc. I think the wealth concerns are more valid that concerns about racially biased test questions and such.

Perhaps more importantly, the colleges that are the focus of this forum almost never admit based on HS GPA in isolation + SAT, as was evaluated in the study above and most similar studies. Instead selective colleges usually consider the full transcript, rather than just HSGPA in isolation. Admissions may consider how a student’s GPA compares to GPA range for their HS, rigor of curriculum, number of AP classes, which classes had the lower/higher grads, upward/downward trend, etc. One of the key benefits of the SAT is helping to standardize GPA against different HS grading systems and different levels or rigor; so if the admission officer considers the HS’s grading system and the student’s level of rigor, the incremental benefit from the SAT score decreases. Selective colleges also often consider things like LORs, essays, ECs/awards, activities outsider of the classroom, etc. The more factors you consider, the incremental benefit test scores have beyond the combination of those measures.

It’s also worthwhile to note that colleges aren’t particularly focused on maximizing average first year GPA of students prior to effects of a curve, as was evaluated in the study above and typical SAT college preparedness studies. Instead they are more concerned with things like students persisting until graduation. The correlations become much smaller for chance of a particular student graduating than for first year GPA, often less than 5% of variance explained by SAT in isolation… and near negligible incremental benefits.

1 Like

Having taken both the old and new SAT, I think the new SAT in some way shows “college readiness,” but overall is not the best tool for admissions. I treated the new SAT as a game, which I believe it is. It is a game to find out patterns in question types and answer options, and to effectively answer before the last grain falls down the hourglass. But this is merely measuring problem solving skills, not knowledge, writing ability (hardly anyone requires the essay which is contrived anyhow), personal passion, or life experiences. While I personally enjoyed the nature of taking the test, it is very stressful and difficult for many students with learning or physical disabilities. Although CB tries to be accommodating, the fact that it is a timed multiple choice test means the structure is fairly rigid and hard to morph to be fair to all students.

It will be interesting to compare grades, retention, and graduation rates between previous college classes and the upcoming one whose admissions cycle will probably discard SAT as an entrance requirement.

1 Like

https://commons.princeton.edu/livinglaboratories/2016/10/24/the-origins-of-the-sat/

It is relevant to remember the SAT was originally created by a racist who engineered the initial tests to support his belief that minorities were inherently less intelligent than white people. There are many articles about this on the internet. The link above is to a student article on the Princeton University site, but there are many other sources.

Every test after those initial tests were based on a test intentionally biased against minorities. In the 1980s-90s, insubstantial efforts were made to de-prejudice the test. Those efforts continued until we have reached a point where the most blatant questions were removed. Today, it is understood that while the test may be less prejudiced against minorities specifically, it remains prejudiced against students from lower SES families.

Here is another student article about the SAT, this time from Boston College: http://www.bcgavel.com/2019/11/05/the-racist-history-of-the-sat/
I’ve chosen student articles to link to because I think it’s important to know what students think of the tests.

Recently, more and more universities have been moving to a Test Optional policy. With Covid-19 resulting in the cancellation of most SAT/ACT tests over the past few months, many more universities are moving to a Test Optional policy for this coming application season.

For now, view it as a necessary evil for admission to some universities. If the universities you choose to apply to require the test, you must take it to get admitted. As such, you should prepare for it as best you can.

Lastly, I recommend an article Teen Vogue ran in Oct 2018 about this topic. It also approaches the topic in a manner to aid students going through the process: The History of the SAT Is Mired in Racism and Elitism

1 Like

Like any other standardized test, the most important thing that it measures are A, how well prepared people are for the test, and B, how talented people are at taking this particular type of test.

Mastery of the material from which the test is taken helps with A, and general academic talent helps with B. However, these are only a part of what goes into high SAT or ACT scores.

Unfortunately, SAT and ACT scores are taken by many people, including in college admissions, as unambiguous and objective measures of mastery of all core academic material and of academic talent, and do not accept that there are extensive other factors which have as much, and more, weight.

The strongest indicators that SAT scores are not these are,
A. the fact that experience with the test increases the score,
B. the fact that test prep works,
C. the low amount of variation in first year college GPA that is explained by SAT/ACT scores.

At the end of the day, a test for which you can prepare is not a good indicator of innate abilities, and a three hour standardized multiple choice test is not a good test of mastery of 11 years of education, nor is it a good way to test whether a person has the ability to go through four years of tough academics.

I’d rather get opinions of professionals who know what they’re talking about.

1 Like

It’s socio-economic bias. Either you have access to better schools, better teachers, and better prep…or you don’t. If you don’t have access then the odds are stacked against you, regardless of race.

That said, life isn’t always fair. It’s only one tool. Success depends on so many more variables.

Personally, I find grade-inflation an annoyance.

I remember compiling data for the MBA admissions office when I was a grad student. Undergrad GPA and GMAT scores were about equally correlated with MBA GPA if I remember correctly. Somewhere around a .6 something if I remember correctly. Combined it was higher of course.

The data that you have presented @Data10 is compelling, but I always wish that similar data sets had 1 other variable (College major) added to the mix. I have heard many arguments and debates around those with higher standardized test scores choosing more academically rigorous majors, which may lower the standardized test correlation in comparison to 1st year college GPA, but that logic has never made sense to me (at least not without data to support it).

As far as the discussion around racial and SES status bias, their has been a lot of data published on both “phenomenon”. The data shows higher SAT scores as income is increased without fail. It also show a gap in test scores based on race. Some of that could be attributed to the lower SES status of African-Americans as a group, but the statistic that Black households with incomes over 200K have lower SAT scores on average than White households with incomes under 20K means that other factors are at play besides socio-economic status alone.

My own African-American family has adapted to the racial bias by exposing our children to the cultural differences early on. The way that I communicated as a youth, what I spent my time reading, how I comprehended what I read, and what I learned in my spare time away from school was completely different from my kids. I can remember reading a SAT reading passage once on skiing while trying to prep for the test (which I took once) and not knowing what skiing was (I find this slightly funny and sad at the same time but at least I am from the South). Once I looked up what skiing was, the context of the passage made more sense, but I ran into that problem quite often.

“Like any other standardized test, the most important thing that it measures are A, how well prepared people are for the test, and B, how talented people are at taking this particular type of test.”

That’s not really the definition of a standardized test, it’s basically to ask the same questions to the test takers and score consistently so you can compare students, that’s the most important thing.

“I have heard many arguments and debates around those with higher standardized test scores choosing more academically rigorous majors”

It would depend on your definition of most rigorous majors, and there probably isn’t data aggregated over all colleges, but individual colleges show the test scores by school and engineering, science usually have higher scores than the other schools.

It’s not the definition of “a standardized test”, but it is the invariable result of producing a standardized test. If you want to create a set of multiple choice questions which do not differ much in difficulty between years, it limits the topics and the types of information that can be tested. It also limits the depth of the material.

Multiple choice questions also limit the amount of synthesis that can be required, and there is a lot more very straightforward questions on knowledge. There is no place for innovation or originality, either.

Standardizing the test has limited it to a specific set of knowledge that lends itself to these types of questions. Moreover, the questions have been standardized in format and language, which further limits the range of ways in which a problem is presented, both in language and in terminology.

All types of testing are helped by experience with that particular type of test and talent at test taking. Standardized tests simply make those two factors much more important.

“Multiple choice questions also limit the amount of synthesis that can be required, and there is a lot more very straightforward questions on knowledge.”

Agree on the multiple choice part, the SAT added an essay and made some of the math problems write-in to address that.

“Though this, I believe that it’s a fair test”

It may not be the fairest of tests, but colleges don’t apply it uniformly, but more in context - high SES, educated parents, Asian which is common here in the bay area would need to score much higher scores to get in to the selective colleges.

It has nothing whatever to do with race. It’s a standardized test to determine how prepared you are for college. No more, no less. If you score low, then you haven’t studied much in high school, and you probably wouldn’t do very well in college if you started the next fall. If you eliminate the SAT, then you would get discrimination. Colleges would have to rely on GPA, which varies widely from school to school. This means they would have to do school rankings. The ones ranked the lowest would be poor minority communities, effectively shutting them out of selective schools…all in the name of solving racism.

This seems to be a common theory among forum posters, but I’ve never heard of it occurring in practice. Many colleges are test optional and have been long before COVID-19. These existing test optional colleges have a wide variation of different degrees of selectivity and a wide variation of different admission systems. When test optional colleges publish stats comparing test submitter class members to non-submitter class members, the non-submitters always have a significantly lower average income than test-submitters, as well as larger portion of Pell grant kids or other indicators of low income. This pattern has occurred at every college I’ve ever heard of that published income stats. Lower income kids as a whole don’t seem to be shut out any more than they were with test scores being considered.

As an example, Ithaca does a review that includes both the applicant population and admitted population at https://www.ithaca.edu/ir/docs/testoptionalpaper2.pdf . A summary is below:

Applicants
Test Submitter – Mean family contribution = $37k, 10% Pell, 26% URM*
Test Optional s – Mean family contribution = $31k, 17% Pell, 40% URM*

Admits
Test Submitter Admits-- Mean family contribution = $37k, 15% Pell, 22% URM*
Test Optional Admits – Mean family contribution = $30k, 29% Pell, 35% URM*

Enrolled
Test Submitter Enrolls-- Mean family contribution = $34k, 18% Pell, 19% URM*
Test Optional Enrolls – Mean family contribution = $29k, 30% Pell, 31% URM*

*Ithaca includes Asian students as part of their URM category. Only ~4% of Ithaca kids are Asian.

If it were a test of how much a student studied in high school, people who are prepping for the test would not be able to get scores which are any higher than students who did not prep.

If it were only a test of how much a student learned in high school, there would not be a clear pattern of increase in scores between tests, no matter how little extra material was learned in the meantime.

It isn’t that there is no relationship between SAT scores and what a students learns in high school. After all, the material is based on what a student learns in school.

However, the SAT score is also affected by the talent to perform well on standardized tests (yes, this is a thing), levels of anxiety, the conditions in the room, the number of tests that a student has taken, prep, extra time (that wealthier students are more likely to receive), and whether the test questions cover specific areas in which a student is particularly strong, or particularly weak. The effects of these factors are create additional variance which substantially weakens the relationship between the SAT scores and the knowledge that a student has gained in school.

That means that, while answering 90% of all SAT questions correctly will give a student a score of about 1500, the range of the actual scores of students who have mastery over 90% of the material is much wider.

All things being equal, and student with better mastery over more material will get a high SAT score. However, all things are not equal.

Then there is an issue of all the things for which the SAT does not test. SATs do NOT cover at all are the ability to persevere in class, hand in all the assignments on time, to participate in class, to go beyond the material that is provided, to do extra work, etc. They don’t cover the ability to synthesize, demonstrate deeper understanding, demonstrate scholarship, etc. All of these are as critical for success in college as they are for high school, and not one of these are tested in the SAT, but they are included in the HSGPA.

Does a kid who gets a 1550 have better study habits than a kid who gets 1410? Are they more likely to participate in class? Are they better at writing the essays which form a substantial part of grades in colleges? We don’t know, since the SAT scores don’t show that. However, it is pretty likely that a kid with a GPA of 3.95 is better at all those things than a kid with a GPA of 3.75.

I think the SAT naturally has some advantages towards the wealthier crowd, but unfortunately that’s an unavoidable aspect of the college process; I compare it to how kids from expensive prep schools will receive much better recommendation letters from counselors; it’s just a part of the system, unfair as it is.

Yet, with that said, by far my biggest criticism towards the SAT policy at elite schools is the ability to “Superscore” tests. I get it, you might not be able to flip the switch on reading critically and solving Algebra problems, but there are kids who can, and just because you can shill out the time and money for two extra stabs at the test doesn’t mean you should have a higher chance at a certain university. Wondering what anybody else thinks about this?

“However, it is pretty likely that a kid with a GPA of 3.95 is better at all those things than a kid with a GPA of 3.75.”

@MWolf

There’s no way to make a statement like that unless you know the rigor of the two kids and how harsh the grading the system is in the high school. And that’s only a couple more Bs than the 3.95 student.

The other flaw is you think that GPAs are not gameable like the SATs, when there’s more gaming that goes on with GPAs, here’s a few:

  • cheating, the more competitive the hs, the more cheating, there are a myriad of ways students can cheat today with technology than I back in the day. there's cheating on the sats/acts too, but h/s cheating is more prevalent
  • teachers afraid to give Bs or Cs because of the complaints they get from parents
  • related to the above, the classes being setup so that you need a B or a B+ on the test to get an A in the course because of h/w, labs etc..
  • avoiding hard teachers
  • test prep for AP and honors classes, if you think people do too much test prep for the SAT/ACT, there's more that goes in to the harder classes.
  • taking a prep class or the actual class itself in the summer

"Are they better at writing the essays which form a substantial part of grades in colleges? "

Doesn’t apply to stem - they may take a writing class or two (and many times you can take those pass/fail), but 99% of their time is spent solving problems.

In my grand dream of colleges admission reform, I’d limit the sat/acts to only be taken once, like basically every other country.