GMAT Tip of the Week: Thinking Like the Testmaker

GMAT Tip of the WeekEarlier this week, in creating a blog post for our friends at Poets & Quants, we wanted to punctuate the Data Sufficiency lesson in the post with a fairly-basic sample problem that would have these four characteristics:

– More than half of users would get it wrong

– Of those users, the vast majority would pick the trap answer that corresponded with one particular mistake, the subject of the post

– After reading the rest of the post, they’d easily understand the mistake they made

– Naturally, the question had to be perfectly valid and the trap answer couldn’t be a “dirty trick” but rather had to be a valuable lesson

Now, you’d think it might be hard to get more than half of users – those who are taking additional time to study for the GMAT, so they’re clearly taking things seriously and putting in the work…this isn’t a late night Leno or Kimmel sketch in which we find the least-educated, least-worldly tourists on Hollywood Blvd. to answer a current events question! – to get a fairly basic question wrong. But in all actuality? It took less than 5 minutes to come up with a question that would end up hitting all those metrics. Upon creating the question we posted it in the Veritas Prep Question Bank and 60% of users to date have gotten it wrong, with just over 50% of users picking the prescribed trap answer. The lesson?

Writing GMAT questions that smart people get wrong is easy, because GMAT test-takers are extremely predictable.

That’s the point of the Veritas Prep “Think Like the Testmaker” theme that reappears through our lessons. The GMAT can bait you into trap answers over and over again, because your mind is predictable. A standardized test lends itself quite well to a standardized set of mistakes. The questions themselves will very often seem unique to you, but the same blueprints come up over and over again for the testmaker. “How do I trap 60% of users? Well, let me dig into my playbook, which includes:

  • Hide a valuable piece of information in the question stem of a Data Sufficiency question. For added effect, make one of the statements just algebra-intensive enough that people have to spend 45 seconds on it and it steals their attention away from that juicy nugget in the question…they’ll have totally forgotten about it.
  • Use a strange – but correct – grammatical structure as the first 5-7 words of the correct Sentence Correction answer, and people will eliminate it quickly and end up choosing between two incorrect answers that both passed the “I prefer this phrasing” test up front.
  • Set up a calculation for a word problem that leads people to solve for an intermediate step first, then make that number – the number they get right before they should take one more step – an answer choice, and prey on those who aren’t keeping track of which question was actually asked.
  • Create a problem that seems like it’s just testing a concept that everyone memorizes via flashcards, but make that memorized concept only one of two correct options.
  • Use a modifier or qualifier to eat up the first 7-10 words of the correct Critical Reasoning answer, knowing that impatient test-takers will quickly make their decision that it’s “out of scope” before reading the whole thing.

These testmaker traps – and more – come up very frequently on the GMAT, but test taker who are paying attention can learn to sniff them out before they fall into them. That’s the payoff to Thinking Like the Testmaker – if you get a geometry question wrong, it’s not necessarily that you don’t know the geometry (although most people will jot down “study more triangles” as their takeaway), as it could very well be that the testmaker beat you by knowing how you think instead.

For example, the lesson in the aforementioned blog post was “make sure you note important information in the question stem of a Data Sufficiency question”. Why? We see it all the time – students are in such a hurry to start “doing math” that they quickly skim through the question stem, then hustle to the statements. And this blueprint will get around half of test-takers just about every time even if the math isn’t that hard:

1) Embed a piece of information in the question stem; make it something that isn’t obvious…that the student will have to think about just a little to make it “actionable”

2) Make one of the statements require multiple steps and “satisfy the intellect” of the test-taker once they’ve finished those steps

3) Make it so that the question stem piece of information is critical

So our question:

If the product xy is not equal to 0, what is the value of x?

(1) y(x^2) + 4xy + 4y = 0

(2) y = 6

How does this fit the blueprint?

1) That “xy is not equal to 0” line requires just a bit of unpacking. For xy to not be 0, neither x nor y can equal 0.

2) In order to unravel the quadratic in statement 1, you have to do a few steps of factoring:

Factor the common ys: y(x^2 + 4x + 4) = 0
Notice that the x terms form a common quadratic – it’s (x + 2)^2 —> y(x + 2)(x + 2) = 0

By this point, many have forgotten about the fact that y can’t be 0. They may still look at this and say “either x = -2 or y = 0”, and the stats show that more than half of users don’t think this is sufficient. The main reason? They haven’t unpacked that question stem – had they, they could eliminate y = 0 as a possibility (meaning x must be -2) or they’d have just divided both sides by y in the original (you can’t do that if y might equal 0, but since we know it can’t you can divide it out) and been done with y forever. Most people (52% or so) pick C, and several pick E (perhaps not seeing that there’s only one solution for x).

The bigger lesson? It pays to not just see your mistakes in terms of content – those who picked C on this problem do not have a major problem with quadratics! In order to pick C, you have to be able to factor out statement 1 and realize that there’s only one value of x. If you picked C – as most do – you didn’t get beaten on content, you got beaten because the GMAT knew it could sneak xy isn’t 0 past you in the question stem. So as you study, pay attention to your mistakes and Think Like the Testmaker. The more you recognize these blueprints for trap answers, the bigger you’ll smile when you see the GMAT setting you up for them.

Plan on taking the GMAT soon? Try our own new, 100% computer-adaptive GMAT practice test and see how you do. And, be sure to find us on Facebook and Google+, and follow us on Twitter!

By Brian Galvin