I recently encountered someone who, after many years of hearing tutors advise students to “pick the shortest” answer on ACT English and SAT Writing, decided to see how often that option actually was correct. After going through a bunch of ACTs, she discovered that the shortest answer was in fact correct only a relatively small percentage of the time. She was quite incensed about this fact, and took it as evidence that students should not be encouraged to select their answers based on length.

Now, for a tutor who advises a blunt, just-pick-the-shortest-answer-if-you’re-not-sure approach, this is a reasonable criticism.

Otherwise, however, I think it misses the point.

Fundamentally, “shorter is better” is a general guideline; it is not intended to be an ironclad rule for choosing answers. If the shortest answer were indeed always correct, even just on rhetoric questions, then SAT and ACT grammar would be far too easy to game, and many more students would receive high scores than is actually the case.

Working from the shortest answer is about adopting a particular, systematic way of thinking, NOT about mechanically picking answers. It is not intended to mean that the shortest answer is always right. Rather, it implies that shorter answers are more likely to be correct than longer and more complex ones, and that a test-taker who does not spot the correct answer right away, or who isn’t sure what a question is testing, can use that as a rule of thumb to work through answers in order from shortest to longest, with the understanding that the shortest option that fits grammatically and logically will—barring extreme awkwardness—be correct.

That might be the shortest answer; it might be the second- or third-shortest; and in rare instances, it might even be the longest.

Basically, you make an assumption based on a common pattern, you test it out, and if it doesn’t work out, you reassess. Then, rinse and repeat, knowing that if you keep working that way, you’ll eventually hit on something that fits.

The ability to approach questions this way is even more important on a test like the GMAT, where certain Sentence Corrections can be quite long and, at first glance, fairly impenetrable. In the absence of reliable clues as to what is being tested, checking answers in order of length offers a way in, a starting point. It also prevents test-takers from becoming overwhelmed and going into panic mode.

There is also, of course, a particular subset of questions on which the shortest answer is virtually always correct, namely those testing wordiness: if all the choices are grammatically acceptable and convey the exact same information, then one can identify the most probable answer with, oh, I don’t know, perhaps 98% certainty on visual grounds alone, usually in about a second (5-10 on the GMAT). Overcomplicating the process is not helpful in such cases.

In a few cases, when I started work with an ACT student who was overly inclined to dawdle over answers, I would temporarily make them answer certain rhetoric questions based on length alone (“just look, don’t read”), just to jumpstart them into moving quickly and override their habit of overthinking things.

When I was training tutors and someone saw me push a student through parts of an English section at warp speed, they were initially kind of taken aback (when I go fast, I go fast); but once they saw how quickly kids changed their approach to the test and started getting a lot more questions right, they understood why I was doing it. As a short-term measure, it was extraordinarily effective at helping kids to see that the test was simpler than they realized. Almost everyone I worked with this way ultimately scored 30+ in English.

But otherwise, with the exception of extreme time crunches, I would never advocate that a student choose a final answer without actually plugging it back into the sentence to ensure it works. Exceptions do exist, and it is always a good idea to be prepared for that possibility.

Multiple-choice tests may have their shortcomings, but there is undoubtedly some value in practicing this type of disciplined, methodical process, and in learning when it is reasonable to make a choice based on one’s initial assumptions vs. recognizing when a situation calls for further investigation. Presumably, these are the sorts of  vaunted “critical thinking” skills everyone spends so much time oohing and ahhing about.

That said, I would caution against viewing the skills as something that can be taught in the abstract. A student whose understanding of grammar and mechanics is too lacking will not be able to transfer this type of process to a test effectively. In a few cases early in my tutoring career, when I had a very limited number of sessions with a low-scoring student and tried to use this approach, it inevitably backfired because they didn’t have enough hard knowledge to hang the strategy on. They also couldn’t distinguish between questions where it was relevant and ones where it was not.

As a tutor, you have to be able to determine what approach is most appropriate for a given student relative to their current level, their goals, the number of sessions you have with them, their motivation level, and their ability to assimilate and apply new information. That’s a lot of balls to juggle, and sometimes you have to simplify. If a student whom you only have a few sessions with spends a lot of time considering long and complicated answers that stand almost no chance of being right, then you may need to go to the other extreme and have them ignore those answers entirely. They might not get everything right, but assuming they have the basics down, they’ll probably see a moderate amount of improvement and get less fatigued to boot. Given the choice between a 23 and a 27 on ACT English, which one do you think they’re going to take?