Sunday, April 16, 2006

Are LSAT Answers Accurate?

Took the fourth part of my first sample LSAT today, went 20 for 25, with the two answers I changed on further consideration going from correct to incorrect. I wrote something last week about how I was leaning towards interpreting "best answer" as "only answer" but today I was proven wrong. Adding up the four sections yielded a rough score of 77 and an LSAT adjusted score of 162, which I suspect isn't too good. Blowing half of the third section on faulty graphing and time management certainly didn't help, so it will be interesting to see if I improve on the next run through.

We've all taken plenty of standardized tests in our life, and taken the scores they assign us on faith. Faith that the grading equipment isn't broke, faith that there's no intentional shenanigans, and faith that the people who design the test actually know the right answers to the questions they are asking. That's the facet of the LSAT I'm going to address today. One possibility that I'm not considering here is that the answer guide has a typo, and that the answer I'm choosing as being correct was the one used when the test was officially offered. The following example in #24 from the 4th section of the first test in "The Next 10 Actual, Official LSAT PrepTests(TM)"

24. Medical researcher: As expected, records covering the last four years of ten major hospitals indicate that babies born prematurely were more likely to have low birth weights and to suffer from health problems than babies not born prematurely. These records also indicate that mothers who had received adequate prenatal care were less likely to have low birth weights than were mothers who had received inadequate prenatal care. Adequate prenatal care, therefore, significantly decreases the risk of low birth weight babies.

Which of the following, if true, most weakens the medical researcher's argument?


(Here I'll give only their answer and the answer I chose)

(B) Mothers giving birth prematurely are routinely classified by hospitals as having received inadequate prenatal care when the record of care is not available.

(C) The hospital records indicate that low birth weight babies were routinely classified as having been born prematurely.


I initially chose (B), the correct answer as listed in their answer guide, but I changed to (C) based on the following two reasons, keeping in mind that we are looking for the answer that "most weakens" the researcher's argument.

First, for (B) to have an impact, there would have to be some number of mothers without available records of care. Nothing is said about this number, whether it was five, fifty or five hundred or more.

Second, (C) affects the reliability of the whole study. If low birth weight babies are routinely termed premature, which would occur far more frequently than instances of mothers of low birth weight babies without records, the whole argument about adequate prenatal care having an effect on birth weight is built on a castle of sand.

Therefore, while (B) would be a reasonable answer in the absence of (C), the quantitative effect of (C) is going to be greater, and would do more to undermine the researchers argument.

Do the LSAT people accept appeals about answers? I don't know if they even release the test takers answer sheet on demand, so I'd be pretty surprised if there's any reasonable way to go about an appeal. So, it looks like we have to take our LSAT scores on faith, even if it's bad faith.