Flawed Reasoning in LSAT Logical Reasoning Questions: Exceptional Cases and Overgeneralizations

    LSAT Logical Reasoning, LSAT bibles, PowerScore LSAT prep

    [Read Part 1 and Part 2 of this series here.]

    Before we continue with Part Three of our examination of common flaws found in LR questions on the LSAT, let’s once again take a brief moment to review why it’s so important to understand these argumentative errors. Here’s how I began the first post in this series, where we looked at Source Arguments:

    Considering the vast majority of LSAT Logical Reasoning questions will have an argument in their stimulus, and the vast majority of those arguments will contain some sort of flawed reasoning, I thought I would take a moment to address a variety of the flaws that tend to appear with some frequency. In a series of posts I’ll examine a number of common mistakes that authors on the test make, which should prove useful for both Flaw in the Reasoning questions (a type that accounts for about 15% of all LR questions), as well as other question types that require you to respond to argumentation.

    With that in mind, let’s turn our attention to another type of mistake that LSAT authors—and a LOT of real-world people—make with some frequency: Exceptional Case/Overgeneralization arguments. A stimulus with this flaw typically presents a small number of instances (maybe a single case, but never more than a few) and treats those instances as though they substantiate a broad, often sweeping, conclusion.

    The error in that type of argument stems from a fundamental truth of valid conclusions: they can never be more strongly worded or absolute than the evidence given in support of them. Put more simply, a conclusion is limited by its premises. So observing a single instance of some event is not enough evidence to conclude that said event “always occurs,” or even “is likely to occur” similarly in the future, any more than the individual absence of an event would be enough to justify a belief that said event “will never occur,” or “is rare.”

    Sometimes I envision this idea as almost a ladder, or staircase, of increasing certainty, where the lowest rungs/steps are reserved for ideas of mere possibility (“could occur,” “may be the case,” etc), and as we climb we find ourselves reaching heights of greater restriction, and thus greater certainty in our beliefs: above possibility is likelihood (“it will probably be the case,” “this will usually occur,” “rarely will we find,” etc), and only at the very top rungs/steps do we reach our absolutes, where our evidence is so persuasive, so strong, that we feel confident in making claims with extreme language (“it is the case that,” “this will occur only when,” “we can expect a complete absence of,” etc). We were able to climb to that height of certainty only because we had sufficient support to hold us up. Occupying a height below the last rung of evidence is easy—if I can prove something is nearly guaranteed to occur, I can certainly say with confidence that it’s possible; I simply cannot continue to safely climb above it.

    I want to pause briefly here to make something very clear: this idea of recognizing the strength of language, and its limiting effects on an author’s conclusion, is one of the most crucial considerations you must make for EVERY stimulus! They must all be viewed through this particular lens, where you contemplate the scenario described by the author and the nature of the words used to describe it—literally how the language an author chooses dictates exactly what is, and what is not, reasonably supported. So that aspect of this discussion transcends our look at common flaws, and should be a primary concern regardless of the specific stimulus or question type. As an example of the broad applicability, note that one of the most common answer choice traps in Must be True questions is an answer that is simply more extreme in its language than the premises in the stimulus allow for, which is the exact same mistake as is being described here.

    Of course, this is meant to be a look at a specific argumentative error, so let’s get back to the task at hand. What exactly might an Exceptional Case or Overgeneralization flaw look like on the LSAT? Consider an example:

    “My friend swears the mechanic at that shop overcharged her last week and after looking at her invoice it seems she’s right. So don’t take your business there, as you’ll definitely get ripped off.”

    Or one more:

    “You know, for years people have been telling me that cigarettes are bad for my health. But look at my uncle Frank: he smoked two packs every day for 83 years and outlived all my other relatives by nearly two decades!”

    In either case, do we have enough evidence to draw those conclusions—“take your car there and they’ll overcharge you” and “smoking isn’t bad for my/your health”—with language that absolute? No, of course not. Granted, the first conclusion seems a little more reasonable than the second, but we’ve got only a single instance in both examples…not nearly enough support for an argument that strong.

    Granted, those aren’t the most elaborate examples to work through, but they do illustrate the underlying idea and that’s what is fundamental to your ability to succeed when things get trickier. And the good news is that, just as the error is hopefully easy to recognize above, so too will it be easy to see on test day.  

    If you do spot an argument where the conclusion generalizes beyond the evidence provided, you’ll want to be able to quickly scan the answers for one that describes this flaw. Here are a few examples of how that description would appear on the LSAT:

    “bases a general claim on a few exceptional instances”

    “supports a universal claim on the basis of a single example”

    “Too general a conclusion is made about a medical procedure on the basis of a single operation”

    A final point about arguments of this type: this is a very specific type of mistake and follows from a very consistent construction, so unless you see a generalized (broad) conclusion following a limited number of instances, do NOT pick an answer like those above. They are very commonly presented as traps following stimuli with other reasoning errors; in fact, answer choices describing this error are more often incorrect traps than they are correct, so be wary when you inevitably encounter them.

    Keep an eye out for additional posts addressing a wide variety of other flaws you are likely to encounter on test day. Commit them all to memory and you’ll find yourself well prepared to respond to nearly any argument you come across.