Wednesday, September 22, 2010

LSAT Reading Speed Meta-Analysis

The more that I look into the problem of LSAT reading speed, the more I realize that my normal channels of inquiry are coming up dry, at least in terms of verifiable data from trustworthy sources. The sole LSAC-sponsored research on the matter is currently on my to-do list, and will be part of a future post about the few key findings I'm confident in.
This study examines the question whether test-taking speed is a variable that affects performance on both the Law School Admission Test (LSAT) and actual law school exams.
However, for now, I'll discuss the dearth of relevant sources found during my normal online search patterns.

There is an article series on Slate that covers the basics of why speed reading is a sham--it's really just speed skimming.  Four hundred words per minute seems to be the upper bound of human potential before you start sacrificing comprehension. Skimming might be enough for casual reading, but it will certainly screw you up on the LSAT.  Even in the best of circumstances, 4 of 5 answer choices will be stuff that you skimmed and you'll have to refer back to the text to find the answer.

Problem is, that article series and the accompanying foray into speed reading software is so old that the author--non-satirically--refers to getting his "CD-ROM".  I eventually confirmed the smoldering crater when the most recent version of the software is compatible with Windows XP.

I turned to scholarly sources, digging up a meta-analysis of the comparisons between reading on screens and on paper. Again a problem with immediacy: no sources were cited beyond the 1990s. However, the following detail is important regarding the link I posted last week about the "best" free online reading speed test I could find.
Figures vary according to means of calculation and experimental design but the evidence suggests a performance deficit of between 20% and 30% when reading from screen.
Clearly, I'm going to have to utilize Harvard's library systems to find something relevant.

After taking the online reading speed test myself, I ended up with a words per minute rate between 300 and 350 on both tests with a comprehension rate of 100%.  According to the test, that clocks me in at "above average", and combined with the meta-analysis about paper versus screen, hovering slightly above 400 words per minute on the LSAT.  That rate seems to be the upper bound for college students, and the graph below shows how long it takes humans to reach that speed.

This speed test result at the upper bound for words per minute appears to correlate well with my LSAT reading comprehension speed and peformance, since I usually finish with at least 5 minutes to spare and rarely miss any points in the section.  Perhaps the LSAC designed the LSAT to max out at this point based on empirical research or just balanced Reading Comprehension sections there after years of experience setting curves.

That begs the next question: does reading speed have anything to do with that performance, or is it more about something else: skills, comprehension, notetaking, memory? Or, more likely and more depressing for a quantitative approach, is reading such a personal and complex task that everyone approaches it differently?

And if speed does have a positive relationship with LSAT performance, what can we do to improve reading speed?  The best answer seems to be timed practice (a la speed training for runners) and subject matter familiarity so that processing takes up less time.  What this means is that knowledge about physcial and social sciences, arts and humanities, and law may indeed be important to your reading comprehension score by controlling reading speed.

A quick "NYTimes reading speed" Google search brings up the--unfortunately named--Speed Read blogish series and our own suggestions for how to use the NY Times to increase your LSAT reading speed. Irony is searching for persuasive information and finding yourself.

My findings relevant to the LSAT:
  • ~400 words per minute is the approximate upper bound for humans at 100% comprehension, and that is about the speed the LSAT expects
    • The LSAT is thus built so that most humans--and many testtakers--will not be able to finish
  • Online reading is slightly slower than paper reading, by about 20-30%
    • You can find your online reading speed--and thus calculate your paper reading speed by dividing by .75--at
  • "Speed reading" claims above 500 words per minute are hoaxes for LSAT purposes because you dip dangerously in comprehension and retention
  • The more familiar you are with the subject matter, the more quickly you read