kitchen table math, the sequel: PTSD & SAT

Tuesday, November 22, 2011


Exactly what I would have predicted:
Given the widespread use and high-stakes nature of educational standardized assessments, understanding factors that affect test-taking ability in young adults is vital. Although scholarly attention has often focused on demographic factors (e.g., gender and race), sufficiently prevalent acquired characteristics may also help explain widespread individual differences on standardized tests. In particular, this article focuses on the role that posttraumatic stress symptoms (PSS) potentially play in standardized academic assessments. Using a military sample measured before and after exposure to war-zone stressors, the authors sought to explain test-taking ability differences with respect to symptoms of PTSD on two cognitive tasks that closely match standardized test constructs. The primary method for this analysis is based on an item response theory with covariates approach. Findings suggest that the effect for PSS is significant on both tasks, particularly for those who experience the highest levels of PSS following war-zone exposure. Findings provide potentially valuable information regarding the nature of the relationship between PSS and verbal and logical reasoning test performance.
Previous research on college-age groups suggests that educa- tional attainment is negatively impacted by anxiety disorders (Kessler, Foster, Saunders, & Stang, 1995); however, less is known about the specific effects of anxiety disorders on test-taking ability, particularly from a prospective approach. The current study sheds light on this issue and suggests that after controlling for predeployment PSS and a number of possibly confounding factors, PTSD symptoms adversely affect test-taking ability in study par- ticipants, and that there is a dosing effect in which more severe symptoms are associated with poorer test taking.
Posttraumatic Stress Disorder and Standardized Test-Taking Ability - Leslie Rutkowski et al - Journal of Educational Psychology - 2010, Vol. 102, No. 1, 223–233.


Roland said...

I'm not quite sure what you're trying to say here. The article's claim that PSS adversely affects test performance is, like you said, not a huge surprise, but saying that being stressed prior to or during the SAT because of factors related to the SAT (its role in college admission, the speed at which one is expected to work, anxieties caused by being in a test-taking situation, etc.) is a reason to discount the SAT isn't really fair. When 50-80% of final grades in many college courses are determined by performance on midterm and final exams, which are similarly grueling and high-pressure, being able to perform well on long tests, in spite of any test anxieties one might have, is an important ability to have. While many classes rely on projects or papers instead to determine grades, it's extremely unlikely that a student will go through all four years of college without having to sit a 3-hour final that will have a significant impact on their grade.

There are definitely a lot of problems with the SAT, but I don't think the way it tests one's ability to perform well on long exams, under pressure, is one of them. A number of the readers of this blog are parents who want to see their kids do well enough on the SAT so that they can attend good universities and colleges, but it's important to emphasize the role that test taking has at these (and all) universities. I recently graduated from a HYPS school, and I'd say that 60-70% of my classes involved midterms/finals. This includes classes like my philosophy, film, and psychology classes, which had midterms and finals as long and as difficult as my engineering classes (I was in an interdisciplinary major; I also took a bunch of random classes out of interest). If a student is really struggling with the SAT because of test anxieties, they're going to have to learn to turn those anxieties into productive behaviors, or college is going to be a really nasty surprise, regardless of how smart they are and how well they do on their homework assignments and papers; at the top colleges, everyone is smart and does well on their homework assignments and papers.

One could also argue that emphasizing test performance is just bad news and that students should be evaluated by different systems. This is an argument that I'm actually partial to (ooh am I glad that I won't be taking any more tests for a while), but it's not really relevant – regardless of whether college courses should evaluate on the basis of long tests, the fact is that they do, and being able to take tests, under pressure, is a necessary skill.

gasstationwithoutpumps said...

"at the top colleges, everyone is smart and does well on their homework assignments and papers"

Having been a grad student and a professor at 3 different top colleges, I can assure you that not everyone does well on homework assignments and papers. That said, I agree that the ability to do well on 3-hour (and longer) exams is critical to success at most colleges.

Allison said...

Roland, this was really well said.
" If a student is really struggling with the SAT because of test anxieties, they're going to have to learn to turn those anxieties into productive behaviors, or college is going to be a really nasty surprise, regardless of how smart they are..."

very very few people have "test anxiety" without having other anxiety issues. Anxiety issues can be crippling in education as in the rest of life. The solution is not to take away the sources of stress or fear. The solution is to raise competent children who can negotiate the imperfect, difficult, unfair world they live in.

To be competent, they must learn to succeed by having overcome failures. They do tht by learning success depends on effort and talent, not just talent. They need to learn not to make excuses for their performance, but to understand they must succeed in real work imperfect conditions.

They must learn to handle difficult people as much as difficult situations. We help them learn that by preparing them for the road, not constantly preparing the road for them. We need to teach them about their temperament so they can recognize their own needs, and pick the best environment for themselves.

If, say, like me, you had massive anxiety disorders as a child, then the appropriate solution was not to go 3000 miles away to MIT at 16, as I did, because it produced the expected disastrous results. It also isn't appropriate to blame the SAT. The solution would have been to be a waitress for a year or three and grow up more, handle the world, and then handle the artificial tests.

For all of the complaints about credentialism, there's a reason top institutions hire people who have top credentials: the credentials, coupled with someone who demonstrates cheer and health, indicate a person of high energy, decently high intelligence, decently high interpersonal skills who can handle difficult environments and difficult people. Not all of us can.

SteveH said...

ETS uses tricky questions and limited time to separate students near the top end. Part of that separation is related to the need for specific SAT test preparation and part of that is related to stress. Many students, however, take the SAT test with little stress (because they know that they are going to end up in the middle of the pack), they get accepted to some college, and they do quite well. The stress in college might come if you are close to failing a course. Stress on an individual test might give you a poor grade, but I've never seen it as a deciding factor. The stress on the SAT is a different animal. The stakes are much higher.

As I mentioned in "Bubble this, part 2", does the added stress needed to get a top SAT score correlate with the stress encountered at the colleges that SAT score allows you to get into? No. The stress on the SAT is much higher. The difference between a 780 and a 680 score is tiny and the payoff (or lack thereof) is much larger.

I think the key point is that stress has now become a very large variable on the SAT. It does not correlate with anything significant in college. After 7 1/2 years of full-time college, I never had one test with that level of stress and resulting consequences.

The stress on the SAT is like the stress in a championship basketball game with 3 seconds on the clock and you are two points behind. The stress in college is more like the stress in a marathon. If you do poorly on a test, that is just one grade. If you do poorly on many tests and risk flunking out, then stress is not the problem. Something else is the problem.

I don't think that ETS is trying to test one's ability to deal with stress. I think it's a natural result of the huge college demand, ETS trying to separate students, and the enormous differnce in payoff for very small numbers of correct or incorrect answers. The stress variable is way too large and seeing some sort of correlation does not explain what is going on.

SteveH said...

"...very very few people have "test anxiety" without having other anxiety issues. Anxiety issues can be crippling in education as in the rest of life."

Only people without (what level?) of test anxiety are worthy of being accepted into a particular college? Is college a performance degree? For a music performance degree, it's all about the audition. It's all about performance. Few degrees are like that. Few careers are like that. Life is not like an SAT test.

The SAT induces much more than a normal level of anxiety. It has evloved that way due to demand and the need for ETS to separate students. The anxiety variable has grown way out of proportion to its meaning. How many Harvard grads from 20 years ago could even make the wait-list cutoff now? If college demand dropped way off, would life then require less ability to deal with stress? What absolute scale are we talking about here?

SAT is a huge game now with enormous consequences for small changes in score. Just because you see some sort of correlation doesn't mean that it hasn't gotten way out of hand. I'm pissed off that my son has to prepare as much as he does just to show that he is equal to his peers. I'm not looking at the SAT game and seeing something well-planned and normal, as if ETS has it all figured out.

Just because the variables look reasonable and you see some correlation doesn't mean that the game is not pushing those variables to extremes.

Many credentials are crap. Many college degrees are crap. Some fall back on credentialism because they are lazy or they want to cover their asses. Don't look at this issue from the top-down. Look at it from the bottom-up. Look at individuals, not averages. Colleges do look at individuals, and that does mitigate the vagaries of the SAT, but the SAT score variable is still huge. You still have to play the game.