LSAT Argument Strategy: Some Trade-offs and Perspectives (Part I)

Part of the value of the process of blogging my way through the LSAT's is to make myself more aware of which explanation strategies are working for me and which ones aren't.  I am midway through my explanations for LSAT 25, the second test I am attempting, and I am finding myself running into two mutually exclusive ideas about how to deal with the argument questions that I keep getting wrong.  So, to resolve the conflict between these two ideas, I thought I'd post a little bit of an aside column about the LSAT's "Argument" section design and some differing ways of approaching that design.  One approach is covered in this column, and the other will be covered in the next column.

To start off, I think the best way to understand something is to first define it.  In this case, if I were asked the question, "What is the LSAT Arguments section about?" I would be tempted to make some sort of pithy but ultimately worthless statement about "reasoning" or "logic" or "deduction."  However, any definition along this front would only invite questions about defining any of those three words, which would take lots of time and would probably rely on more vague words that would need to be similarly defined.  So to escape this morass of empty definitions, I would instead say:  The LSAT Argument section is concerned with arbitrary categories and the connections and distinctions between them.  Of course, it is the reasoning process that allows us to make those connections and distinctions, but now the focus is on the categories themselves, which allows us to anchor our discussion of reasoning.

In a typical question, the speaker introduces some arbitrary set of categories.  These categories either group themselves into some sort of complete argument, a partial argument, or just a semi-related collection of statements.  This argument can either be correct or incorrect as presented in the question stem (the block of text before the actual question is asked.)

All the information in the previous paragraph might tell you about the question, but it certainly doesn't really tell you anything about how to digest what you have read.  This is where the two different opinions on how to approach the test crop up.  The first approach is something I call the deductive approach.  It is the approach that seems to be advocated by many of the major test prep books I have read (The PowerScore Arguments Bible, as well as the Kaplan and Princeton Review books), and it generally consists of categorizing each argument into one of several categories, which is usually done by question type.  If you've ever thought about dividing the arguments section into categories like "strengthen" questions, "weaken" questions, and "parallel-the-flaw" questions, then you are implicitly thinking along these lines.

The deductive approach assumes that every question type has a certain analytical framework for figuring out the answer.  That is, that every single "parallel-the-flaw" question gets solved a similar way.  The tutorial approach using the deductive method is to train someone to recognize the categories, then practice a lot of problems within those categories.  When a test taker sees a question, he or she has to "deduce" which category the question is in, and then apply whatever framework is suggested by that category.  By the time the test taker has completed a bunch of those types of problems, the reasoning goes, he or she should be ready to encounter it on the actual test.

This approach has a lot to recommend it.  It is simple, straightforward, and everyone appears to be recommending some sort of variation on it.

Nevertheless, I still can't get it out of my head that this approach might be wrong.

When I first began tutoring the SAT's 8 years ago, I would have classified myself as a deductive approach guy.  Show a student some circle problems, some triangle problems, some systems of equations problems, and soon enough, boom, you've got the SAT math covered.  A similar process existed for the verbal section, and when it was added, the writing section.

The more time I spent tutoring, the more I realized two disturbing trends:  one, there were all of these problems that didn't fit cleanly into one category, and two, that simply providing a category of a question's content doesn't provide any sort of way to separate an easy problem from a hard problem.  If the whole point of tutoring is to help someone figure out tough problems, then a category system that doesn't distinguish easy from hard is pretty useless.

These two realizations started me down a long path of thinking about standardized tests outside of their content.  It also urged me to start constructing my own approach to standardized testing, which I'll outline in the next column.

Thoughts? Please share.


For the past eight years, Charles Williamson hasn't met a standardized test he didn't like, helping hundreds of students in everything from the SAT and ACT to the SSAT to the GMAT to AP Calculus. A longtime student of the thought process that defines performance on standardized tests, Charles got off to a good start, receiving a perfect score on his SATs in high school. He then graduated from Brown University, earning bachelor's degrees in computer science and history.  He has long been fascinated by the intersection of education and technology, and when not writing about standardized tests and educational policy, he will be happy to speculate rampantly about how the Internet will affect the ways that we learn.  Charles blogs for the Ivey Files about test preparation.