Tuesday, March 6, 2012

Logical Reasoning: Most Similar in Flawed Reasoning

Today we'll introduce one of the most infamous tasks on the LSAT, one that frequently closes out a difficult LR section with a 400 word monster: most similar in flawed reasoning.

This Zen task is easy to recognize but difficult to complete, if only because the entire text for a question can take up one column on a logical reasoning section's page. In addition to their length, the questions turn on your ability to take abstract logic, identify the logic's flaw, and then compare it against multiple subtle distortions of that logical structure: many people's idea of mental torture.

Identifying the task is exceedingly easy, however. On a modern LSAT, the following stem would be common:
Which one of the following is most similar in its flawed
reasoning to the flawed reasoning in the argument above?
In order to accurately complete this task, you should focus on positively identifying the reasoning error exhibited in the argument; thankfully, these are usually much easier than the stand alone fix by removing questions. As you read, you should also take note of the actors, their explicit relationships, and the degree of certainty for those relationships. The argument's conclusion is also helpful to identify, and it almost always simple to do so for even the most difficult most similar in flawed reasoning questions.

Once these elements are identified, you can turn the stimulus into a pre-phrase similar to a principle or into a formal logic statement with symbols similar to a logic game diagram, making special note of the reasoning error. These tools can then be compared against the answer choices until a perfect match is found, starting with the explicit conclusion (but be careful with contrapositives!), following up with the argument's structure, and fact-checking the error.

For example, suppose the following argument:

Boston is America's largest higher education market.
Police in Boston arrest far more people during the academic school year than they do during the colleges' winter and summer breaks. Thus, Boston police arrest more students than non-students.

The flaw: associating a raw number change with a relative comparison between actors, without any evidence for that comparison.

Type of conclusion: comparative (more students arrested than non-students)
The actors: police, academic and non-academic time, students and non-students

The relationship: arrests increase during non-academic and decrease during academic

Degree of certainty: raw number changes
Combining all of these elements together, you can start eliminating answer choices. Some example distractors might be eliminated because their conclusions aren't comparitive:
  • e.g., a definitional conclusion, as if the students were criminals, not that more of them are arrested
  • e.g., a relative conclusion, as if the students were more likely to be arrested than non-students
Other types of distractors might be based on miscontruals of the stimulus' flaw, such as comparing the two different actors without any evidence linking them or utilizing cause and effect.

Finally, the argument structure distractors might introduce extra actors or omit the bridge.

Once you apply these subsequent layers of elimination, there should only be one correct answer choice left; if there are multiple, you should not try to repeat the same process. Once you only have a couple of answer choices left, you should focus on the relevant factors in a side-by-side comparison of the answer choices, rather than comparing them back to the stimulus. Thinking about the answer choices in a different way will almost certainly help you more than thinking harder about the same process.

As we mentioned earlier, many test prep companies recommend skipping this question type for anyone with significant timing issues on the logical reasoning sections.  While this may be solid advice for those not finishing 5 or more questions per section, we don't generally advocate skipping tasks for those hoping to score above 165.