Beat The GMAT - the MBA Social Network

I’ve been in full-on writing mode lately as we update our strategy guides (look for the 5^th edition in 2012, in time to start prepping for the Next Generation GMAT!). A couple of our teachers have been doing extensive research on all of the available official Critical Reasoning (CR) problems, and now we’re synthesizing everything. Although we’re still in development mode, I want to share some of our take-aways with you so that you can start to benefit from them right away!

One 4-step process for all CR questions

First, there’s one overall process we’re going to use for any CR question:

Step 1: Identify the question.

Step 2: Deconstruct the argument.

Step 3: State the Goal.

Step 4: Work from wrong to right.

Those steps might sound obvious to some people and very vague to others. I’ll explain each in more detail below, but I want to say first: each step is there for a very important reason, and each step has been split off from the others for a very important reason.

Identify the Question

Most arguments are followed by a question (there is actually one exception, but we’re not going to discuss that here). There are several different types of CR questions, and the wording of the question stem (the part below the argument itself) allows us to identify which type of CR question we’re about to have to answer. It’s critically important to identify that question type right away because we need to employ different kinds of reasoning depending upon the type of question we have.

We want to know, right from the beginning, how best to work through the current problem, and we can accomplish that by identifying the question Family and the specific question type (see below for more).

Deconstruct the Argument

Arguments are made up of building blocks, such as premises, counterpremises, assumptions, and conclusions. When reading the argument, we want to take very brief notes that allow us to deconstruct the argument into its building blocks. What kinds of information do we have and which pieces lead to which other pieces?

Knowing all of this allows us to do what we call “reading with a purpose.” We know what we need to find, so we can actively look for these things, and we know what kind of reasoning we need to do in order to answer the question, so we can begin to think things through a bit while we’re reading. (This assumes that we know what we’re doing for each question type; more below.)

State the Goal

This is a short but often overlooked step: what exactly am I trying to do when I answer this question? What’s my goal? I know what kind of question I have, I understand the argument and how it fits together now, I know my conclusion (if there is one)… now what?

At this stage, we need to remind ourselves what it is we’re actually trying to do when we start attacking the answers, and this goal depends upon the type of question that we have. Let’s say that we have an Inference question. Our goal is to find an answer that must be true given at least some (but not necessarily all) of the information in the argument. As we continue to study, we’ll also learn that trap answers on Inference questions often do too much; that is, they offer something that could be true or maybe is even somewhat likely to be true given the premises in the argument… but the answer doesn’t have to be true.

By the time we get to test day, we want our goals for each type to be so ingrained in our brains that we can just say to ourselves, “Okay, Inference. Go!”

Work from Wrong to Right

On verbal in general, we’re asked to find the “best” answer. We’re going to use a two-step process in order to accomplish this. First, we look through all five answers and eliminate as many “definitely wrong” answers as we can. On this first “pass” through the answers, we’re not actually trying to decide which is the right one, only which ones are definitely wrong.

If we only have one answer left, great; we’re done. If we have two or more answers left, then we do a second “pass” to compare those remaining answers. Because we’ve also reminded ourselves of our goal before looking at the answers, we’re better able to distinguish between the tempting but wrong answers and the right ones.

The 3 Major Question Families

We can group the vast majority of question types into 3 major Families that share common structures, reasoning requirements, and trap answers. There are some minor types that don’t fit neatly into the three major families; our new strategy guide will address those types but we’re not going to discuss them today.

The first major family is the Structure Family. We introduce this one first not because it’s the most common (it’s not) but because these questions are all based upon a solid understanding of the structure of the argument, and structure is the first major thing we learn about CR. Indeed, we can’t learn to deconstruct arguments until we understand how they’re structured in the first place. The two common question types in this Family are Describe the Role (more commonly called “boldface arguments”) and Describe the Argument. Both require us to identify the components of the argument and both employ “abstract” answers. For example, a Role answer choice might read:

The first is a judgment that counters the primary assertion expressed in the argument; the second is a circumstance on which that judgment is based.

Yuck. See what I mean about “abstract” language? It takes some work to learn how to handle these efficiently and effectively.

The second major family is the Assumption Family; this is the most frequently tested of the three families. Assumption Family questions rely on at least one assumption made by the author of the argument, and all contain conclusions. Assumptions, by definition, are not stated in an argument; they represent something the author believes must be true, but has not stated, in order to draw his or her conclusion. There are five different ways in which we might be asked an Assumption-based question, but all five have one thing in common: we have to identify an assumption in order to answer the question correctly.

We may be asked, for example, simply to identify an assumption (called the Assumption question type) or a Flaw (the author is assuming XYZ but we don’t know whether that’s true). We may be asked to identify something that Strengthens a conclusion. This would involve making an assumption explicit – actually stating (in the answer) that some assumption is in fact true. Alternatively, we could be asked to Weaken a conclusion, in which case an answer will tear down or call into question an assumption made by the author. Finally, we might be asked to Evaluate a conclusion: what hasn’t the author established that would be useful to know in order to decide whether the conclusion is a good one? Our goal here is to find a statement that would test an assumption made by the author in order to determine whether that assumption is valid.

The third major family is the Evidence Family. These questions all lack conclusions – they consist entirely of premises! On these, we’re essentially asked to find a new premise that must be true according to the information given (Inference questions), or a new premise that resolves a problem or contradiction (Explain the Paradox questions). (Note: if you are using our current materials, the 4^th edition guide, the Draw a Conclusion questions are the same as the newly-named Inference category.)

Take-aways for all Critical Reasoning questions:

(1) Read the question stem first. Identify the question type. Know what you’re supposed to be doing and what the traps tend to be for that type.

(2) Deconstruct the argument according to what your goals are. For example, if it’s a Structure Family argument, it’s important to label the different pieces of the argument carefully. If it’s an Assumption Family argument, you need to find the conclusion. If it’s an Evidence Family question, don’t waste your time looking for a conclusion!

(3) Remind yourself of your goal. At first, you might take 20 or more seconds to do this; as you get better, you’ll only need a few seconds to remind yourself.

(4) Cross off wrong answers first, then worry about finding the right answer. Don’t waste time trying to decide whether B is actually correct when you haven’t looked at C, D, or E yet. Eliminate, then compare any remaining answers and think about which is correct. Watch out for traps! During the preceding steps, you’ll have already reminded yourself of what traps to expect for this type.