Ellen's Choice: How I Feel About the SAT Redesign

A few weeks ago I posted an article about the impending SAT redesign and the changes that have been announced. I mentioned at the end of that article that I'd be posting another one soon with my thoughts on the redesign, once I'd had time to think more about them. Well, this is that article.

Overall, I think the motivation for the redesign is good – that the College Board's heart is in the right place and they're acknowledging some of the very real problems that the current SAT has. I'm very happy with their partnership with Khan Academy as well. I'm happy to hear that they acknowledge that students really do need some kind of prep help for the SAT, and that if they're going to force every student who wants to apply to college to take it, they should be offering free prep help for everyone who wants it. Not everyone can afford a private tutor, and money should not be a limiting factor in every student's ability to thoroughly prepare for the test. (That said, I am always willing to adjust my rate for a student with a special circumstance, so if you're in need of help, don't hesitate to contact me!)

As for the specific changes they're making to the test itself, I have more worries there. Again, I think their intentions are good, but I suspect some of these changes will have unbalancing ramifications that the College Board is unaware of. I'll detail a few of my biggest worries below.

Passages About School Subjects

In the current SAT, passages for the reading comprehension are deliberately selected to be obscure and deal with topics the students wouldn't have heard about before. This is done to ensure that students are forced to draw their information exclusively from the passage, and to prevent students from having an unfair advantage due to prior knowledge of the topic. In the 2016 redesign, passages will now deal with topics from school subjects such as science, history and social studies. The College Board says that this will make the test more relevant to the information that students are learning in school – but I was under the impression that the whole point of the obscure passages was that they weren't relevant to the information from school – you couldn't get lucky and get a passage you knew a lot about already.

Here's an example: suppose two students, Student A and Student B, are given the same booklet on the same test date. This booklet, by luck of the draw, contains a majority of history-based passages, including one about the Haitian Revolution. Student A loves history, and actually wrote a term paper about the Haitian Revolution the previous year. Student B, on the other hand, hates history and has an awful time reading about it – the names and dates just jumble together into a big mess in her head. Obviously, Student A has an unfair advantage overall on this test – her love for history means she will likely be better able to find the information needed in the passage. Student B will be at a disadvantage, and will probably struggle to make sense out of the information that Student A found so enthralling. On the other hand, of course, Student A's extensive knowledge of the Haitian Revolution from her term paper might lead to her relying on information not present in the passage, which could cause her to miss some questions. Or she could still get the right answers, but not even need to use the critical reasoning skills that the test is supposed to be evaluating.

Overall, I'm reserving judgement on this topic until we hear more from the College Board about the process of creating the new test. I do see ways in which they could balance the test – make sure there are exactly the same number of passages and questions about each subject, for example, or make the topics discussed come from higher-level topics than the students would have covered. However, I would want to know that the test structure has been heavily beta-tested before it is unleashed on the general student population.

No Penalty For Guessing

The concept of “being penalized for guessing” versus “not being penalized for guessing” is actually a gross oversimplification. It's not that the current SAT can somehow tell whether you knew that answer was right when you picked it – it's all about probability. There are 5 answer choices for each question of the SAT, and a wrong answer gets you a ¼ - point penalty. This means that if you can narrow it down to 3 choices or less, you're actually better off guessing, statistically-speaking. So the SAT doesn't penalize guessing, it penalizes blind guessing. It actually rewards educated guessing. By removing the penalty for wrong answers, they are no longer offering an incentive to avoid random guessing. The College Board says this change is to encourage students to offer “the best answer they have to every question” - but that's what I already teach my students to do with the current system! The only real difference, if taught appropriately, is that in the last minute or so of each section students should now go back through and randomly pick answers for any questions they were completely clueless on, since leaving a question blank is now, statistically, a bad idea.

So this change will have very little effect for students who are already learning smart test-taking strategies – so why am I worried? Well, it could artificially inflate the scores of students who are less prepared. Unprepared students who are guessing blindly could now conceivably get lucky and happen to pick the right answers enough of the time to raise their scores, which would give an inaccurate assessment of their abilities and knowledge. Without the discouragement of wrong-answer penalties, you just encourage blind guessing. Most of the changes in the SAT over the years have been in response to either substantially decreasing or substantially rising test scores, and I worry that eliminating the penalty for guessing will cause scores to artificially rise, possibly resulting in an increasing number of perfect scores such as the ones that prompted the 2005 redesign.

The “Great Global Conversation”

My final worry is a somewhat different one. The introduction of the new “Great Global Conversation” sounds as though it is intended to encourage students to learn about and participate in a nation-wide discussion of the principles our nation was founded on and how they have developed over time. While this is an admirable sentiment in general, I can't help feeling it is really not the business of a standardized test to encourage it. If we want our students to participate more in the conversation about our nation and where we should be headed, that's fine – but why is the College Board worrying about it? Shouldn't that ball be in the court of the country's social studies teachers? I'm not convinced that putting the Gettysburg Address or Martin Luther King's “I Have A Dream” speech on the SAT and asking questions about it would really affect the “conversation” that much. Especially since they'd be treated like any other passage on the SAT – that is, the test-taker would not be supposed to bring outside information into their process, but simply look for evidence in the text itself. It's not as if the questions about the Gettysburg Address would be asking for the test-taker's opinion on the topic; even a passage written by another notable figure about the Gettysburg Address would be followed with questions asking about that author's opinion, not the student's. It is certainly an admirable goal to have for students' futures, but I don't think the College Board should be trying to achieve it through a standardized test.


Hi Ellen,
I am wondering whether you've ever worked for a test development company like the College Board or the Educational Testing Service.  I worked as a consultant for test development companies for about 10 years, and although you make very reasonable and pertinent points about the SAT and how it is being redesigned, the primary reason for redesigning it is to make more money for the test development companies.  One might argue that they are "not-for-profit" organizations, so that attaining maximum profits would not be part of their agendas.  Regardless, everyone has to eat and have suitable shelter.  And many other things.  And I've just written a "fragment," which would undoubtedly affect my score on the writing portion of the SAT. 
Book-length studies could be written on the politics of "high stakes testing," and, in fact, there have been.
Keep up the good work!
Hi Alan,
No, I haven't ever worked for a test development company.  I definitely see your point about the profit aspect of the redesign, and I'd be interested to hear some more detailed thoughts on the redesign and the issue of profit versus stated mission from someone with experience in the field.  These are just my thoughts as a tutor.
if (isMyPost) { }