A very, very long time ago — so long ago that many of the people who stumble across this post were probably in, gasp, middle school — I wrote a post about the infamous marshmallow experiment. For those of you unfamiliar with the experiment, it involved giving a group of preschool students a marshmallow and then telling them they could either eat it right then or, if they wanted to wait, could have a second marshmallow. A follow-up study revealed that the children who had elected to wait had higher SAT scores than those who ate the marshmallow immediately, thus suggesting a correlation between the ability to delay gratification and long-term academic achievement.

That correlation is something I observe pretty regularly. A student who jumps to choose the first answer she thinks sounds plausible without really considering what it’s saying is is obviously going to have difficult doing well. (By the way, I’m not just trying to be politically correct by using the female pronoun here — interestingly, I’ve actually seen this problem occur more frequently among girls than boys.) But the one place on the entire SAT that I consistently see this problem most clearly is in sentence completions.

Oddly enough, I wasn’t fully conscious of that weakness presented itself until I started writing dozens (and dozens and dozens) of sentence completions for my Sentence Completion Workbook (yes, that’s a shameless plug). The more time I spent analyzing how answer choices were constructed, though, the more I realized how those questions are set up to exploit students’ tendency to jump to conclusions before fully thinking things through.

Let’s try an experiment. Look at the following question:

There has been little ——- written about de la Mare; indeed, that which has been written is at the two extremes,
either appallingly ——- or bitterly antagonistic.

(A)  hostile . . ambiguous
(B)  recent . . illogical
(C)  fervent . . complimentary
(D)  objective . . sycophantic
(E)  temperate . . censorious

This isn’t the easiest question, but it’s pretty doable if someone has a solid vocabulary and, much more importantly, can stay calm long enough to figure out what the sentence is actually saying.

The second blank is a little more straightforward than the first, so it makes sense to start with it. It’s the opposite of “bitterly antagonistic,” which has to be something good. Even if you don’t know what “antagonistic” means, you can make an educated guess because good things aren’t normally described as “bitterly.”

Now, when a lot of solid, 500-600 students look at the right-hand blank, something like this happens:

(A) no, ambiguous means “unclear”

(B) no, “illogical” just doesn’t make sense

(C) “complimentary” is good, so it fits! It’s the answer. Ok, done.

When students do bother to look at (D) and (E), they can often get rid of (E) because they know that “censor” is bad. Then they look at (D), and I hear something like, “Well, I don’t know what “sycophantic” means, but “syco” sounds like something bad (like a psycho), so it must be (C).

Which of course it isn’t; otherwise, I never would have chosen this question to discuss.

(C) vs. (D) is actually a classic case of easy synonym vs. hard synonym. It is, shall we say, an ETS favorite, primarily because it plays on the oh-so-common tendency to grab at the first thing that looks like it could work.

In reality, “sycophantic” means “excessively complimentary” — as in, so over the top that it’s borderline creepy. In reality, the second side of either (C) or (D) could work; the answer hinges on the first blank, which is opposed to “two extremes.” The word must therefore mean something like “not extreme,” and between “fervent” and “objective,” only the latter fits (“fervent” means “passionate”).

There is, however, an interesting phenomenon that can be observed when one looks only at the right-hand answers.

(A) . . ambiguous
(B) . . illogical
(C) . . complimentary
(D) . . sycophantic
(E) . . censorious

The words in (A), (B), and (E) have nothing to do with one another. They’re somewhat random, even if they are all negative. (C) and (D), however, have similar meanings — (D) is simply much stronger than (C). In addition, it’s much more obscure, and that’s the part that counts. Given the choice between word that clearly fits and a word that could mean anything, most people will choose the word that clearly fits.

Furthermore, it’s not a coincidence that “complimentary” is presented before “sycophantic.” Plenty of test-takers stop as soon as they hit that word; it doesn’t occur to them that there could be another possibility later on.

But here’s the rule: Different answers to two-blank sentence completions typically contain “easy” and “hard” synonyms that could work equally well for one of the blanks. When this occurs, the more difficult synonym is usually correct. This is particularly true as you get closer to the end of the section (unless, of course, a second meaning is involved) — the answer to number two might be something very straightforward, but the answer to number seven…? Probably not.

So the bottom line:

One, don’t choose an answer until you’ve looked through ALL of your options.

Two, don’t choose an answer just because you know what it means, especially if the word for the other blank doesn’t quite fit.

And three, if you’re close to the end of a section and happen to spot an easy/hard synonym pair in different answer choices, it’s usually a safe bet to start out by assuming that the answer that contains the harder word is right. You can always reevaluate if necessary.