<p>i’ve often thought it would be really interesting to do an experiment with the sat. take a group a folks and administer the test (different editions) to them four times, with one week in between each test. on the odd weeks, test at 8am, on the even weeks, test at 8pm. instruct participants to not engage in any additional studying during after the first test until the end of the experiment.</p>
<p>i think the data could answer some interesting questions, for example:</p>
<ol>
<li><p>how much improvement occurs with frequent retesting?</p></li>
<li><p>controlling for improvement over the course of the four tests, do certain groups show marked improvement during the mornings vs during the evenings? how much of an improvement do they show during their preferred vs their anti-preferred time?</p></li>
</ol>