<p>what is this i'm seeing about how some sat exams have a curve depending on which month you take them? is this true??</p>
<p>I don’t think so, all SAT scores are based off some chart. The month does not matter.</p>
<p>While the month itself doesn’t matter, how others scored in that same month is compared with the average to see how “difficult” the test was. The raw score is then set on a curve to give you the scaled score depending on the overall “difficulty” of the test as determined by the average score, standard deviation, etc.</p>
<p>^^That is completely incorrect, physicsnut. EVERY question is pre-tested numerous times – that’s what the experimental section is for – and CB knows with absolute certainty the ‘hardness’ of every question on every test. The scaled score is predetermined before you ever walk into the room.</p>
<p>I must not have made myself clear. When I say “others” I am referring to the fact that CB tests out these tests on students to determine the test’s difficulty. I’m really not sure how that is “completely” incorrect when the only thing not clear is a single word, others. My main idea stands: the difficulty of the test (which, to actually answer the OP’s question, is different every month) determines the curve and the scoring.</p>
<p>Wait a second, just read your post again. What do you mean “that’s what the experimental section is for”? Are you suggesting that they put the actual questions from the experimental section on future SATs? Haha, I think you mean to say “every question type” is pre-tested numerous times on us students in the experimental section.</p>
<p>Yes, they use the experimental section to test questions.to be used in future real tests. Think about it: they put in a question about trogoldytes in the experiemental section and they find that 99% of the gal get it right but 99% of the guys get it wrong. Ooops, bad question; deleted. But questions that end up not being gender-race-ethnic-geographic biased, can be used later bcos they are considered “fair”. But, CB also then knows the hardness of the question because they know how many of the hundreds of thousands testers got it wrong/right.</p>
<p>Bluebayou is correct. They pre-test all the questions, not just the question type.</p>
<p>Not disagreeing, but would you mind citing that?</p>
<p>Yeah you guys seem so sure of your bold conclusions, yet none of use have any substantial evidence.</p>
<p>Bluebayou is partly correct. All questions do appear on an experimental section before they are used in a graded section. This is done to determine the difficulty of the question as well as its value (it can’t be too hard or too easy and it must be able to distinguish low scorers and high scorers and it must be fair across ethnic groups etc). See for example, [this</a> College Board paper](<a href=“College Board - SAT, AP, College Search and Admission Tools”>College Board - SAT, AP, College Search and Admission Tools) for more info.</p>
<p>However, the difficulty of the test as a whole (math, reading, and writing) is unknown until the test is administered; the scaling curve cannot be generated until after the test is given. See my comments and links in [thread=940259]this thread[/thread] for more info.</p>
<p>The curves do vary from month to month because the whole-test difficulty varies a little from one month to another.</p>