<p>Wednesday, December 16, 2009 7:32 PM
By Ashley Merryman
Recently, University of Michigan professor Stephen Garcia and Haifa University professor Avishalom Tor published a little noticed but remarkable study in the preeminent journal, Psychological Science.</p>
<p>Garcia and Tor took the College Board’s 2005 SAT data for all 50 states, and then they compared each state’s mean SAT score to what they called “test-taker density.” Essentially, test-taker density was the number of kids taking the SAT in each state, divided by the number of locations in that state where kids could take the test. (The researchers considered it a fair approximation of how many kids would be taking the test at a site at a given time.) The researchers discovered that the higher the state’s test-taking density, the lower the SAT scores.
In other words, the more kids take the SAT in the same place at the same time, the lower their scores will be.</p>
<p>Now, there are state differences in who takes the SAT; it could be that only the best and the brightest take the SAT in those rural states. But Garcia and Tor included ACT/SAT prevalence (and ACT scores) into their analysis. Even controlling for the ACT, the kids in the low-density states still had higher scores.</p>
<p>Garcia and Tor also addressed other between-state differences. They made statistical adjustments for parental education and the percentage of kids who were a minority. They also included more systemic controls, including the states’ rate of SAT score improvement over the past decade and the amount of state and federal funding going into schools.
With all these things taken into account, the kids who were taking the SAT in smaller, less crowded venues still had higher scores.</p>
<p>Garcia and Tor wondered if the lower SAT scores were not a mere reflection of the larger pool of test-takers.</p>
<p>What if taking the SAT in a more crowded room actually caused kids’ test scores to drop?</p>
<p>So the researchers pulled out data on short cognitive tests given to University of Michigan students. On these tests, they could analyze the data on a per-classroom basisthe cognitive score of the students compared with how many students were in the room when they took the test.
In those scores, Garcia and Tor found the same pattern as in the SAT data. The more students there were in the exam room at the same time, the poorer the students’ scores.
Garcia and Tor now call this phenomenon “the N-effect.” The larger the "N"the number of participants involved in a taskthe worse the outcome for the individuals who are participating.</p>
<p>This is actually pretty interesting. I actually took the SAT in a cafeteria that was pretty full and scored lower than I expected as did a couple of my friends.</p>
<p>There are different explanations for this phenomenon though. Here is a possible one:</p>
<p>States in which kids have to take standardized tests in “denser” conditions could just be states who have schools with less resources. Hence these schools/states are forced to use the few large rooms they have to administer the testing. The differential in scores could simply be a reflection in the comparatively poor school systems of the state. Have they done anything to isolate the millions of variables like this? I know it says they made some “adjustments,” but there are literally millions of variables like this. And, personally, I have a hard time understanding this. Short of a disability of some sort, I would not buy this argument from anyone. Its one test, you, and a proctor. If the other kids were perhaps coughing/sneezing/tapping their foot/etc… a lot (yet another compounding variable) sure. But I don’t think the difference comes just from the fact that they tested with a lot of other kids.</p>
<p>And while I know one person’s experience isn’t the same as a statistical study, I’m like the anti-thesis to this lol. My score went way up after I tested in a HUGE room (probably close to 300 kids). They went up not because of room size, but rather due to luck and preparation imo.</p>
<p>^It doesn’t have to be a huge score decrease, just a statistically significant one. Even a average decrease of 15 points on the SAT would be quite significant statistically. That means answering less than even one question wrong compared to less dense rooms.</p>
<p>When your in a room with 100+ people, the coughing and season add up. Especially in the winter time. These distractions lead to silly mistakes on the Math section, and break focus, which is critical on the Reading portions. All and all this study isn’t ironic.</p>
<p>Ha, I was always that kid who was sniffling loudly during SATs (and APs too, for that matter). I have this theory that I’m allergic to standardized tests…</p>
<p>Hahaha, this article makes me laugh. But, it’s true, I got a 120 point increase with no studying in between just because the second time had 10 people instead of 40 people in the room. :)</p>