<p>OK, so now we have established that the CB will no longer deduct points on AP exams for
wrong answers. Here are four questions to discuss and debate:</p>
<ol>
<li>Why hasn't the CB made an official announcement?</li>
<li>How will this change affect the percentage of 5s and 4s on AP exams?</li>
<li>What motivated the CB to make this change?</li>
<li>Will the CB extend this policy to SAT I and SAT II tests?</li>
</ol>
<p>The curve will be made much higher so this isn’t particularly a good thing.</p>
<ol>
<li>I think they will later in the year.</li>
<li>I believe the curves will shift right so the percentages stay the same.</li>
<li>I believe it’s an effort to compete with the ACT for market cap.</li>
<li>If my #3 is correct, then they eventually will.</li>
</ol>
<p>wow AP… way to **** us over… I would’ve done so much better with the elimination of point reduction…</p>
<p>Does the point deduction apply to this year’s AP tests?</p>
<p>^ Yes, it does.</p>
<p>Blimey, I wish I was born a year later.</p>
<p>
</p>
<p>You just stole my line from this thread <a href=“http://talk.collegeconfidential.com/ap-tests-preparation/940263-2011-exams-no-more-deduction-wrong-answers.html[/url]”>http://talk.collegeconfidential.com/ap-tests-preparation/940263-2011-exams-no-more-deduction-wrong-answers.html</a></p>
<p>
</p>
<p>The point deduction is still in place (unfortunately) for the May 2010 exams.</p>
<p>
</p>
<p>Not exactly. I looked into the new curves for released exams and this is what I found: <a href=“http://talk.collegeconfidential.com/1065022031-post47.html[/url]”>http://talk.collegeconfidential.com/1065022031-post47.html</a></p>
<p>Just in case the link in the link doesn’t work (there’s a link in the above linked post that links to a revised 2009 AP Euro curve), do this:</p>
<p>Try this link: <a href=“Store App”>http://store.collegeboard.com/sto/productdetail.do?Itemkey=090083123&category=259&categoryName=AP�&secondCategory=281&secondCatName=History&thirdLevelCategory=&thirdLevelCatName=</a></p>
<p>Look for the “NOTE” near the bottom (above “People who viewed this item also viewed”) and click the link in the note.</p>
<p>
</p>
<p>An0maly is wise beyond their years. The change in policy won’t make a difference.</p>
<p>The change from a 5 on the old curve to a 5 on the new curve is only about 5 points on average. It gets bigger for a 4,3,2,1…</p>
<p>I feel like I am going to get a 3 now… greaat…</p>
<p>Basing how the curves will change on their revised sheets is inaccurate. These curves were only changed in order to take into account the change of not subtracting the 1/4 penalty. When students know that they will not be penalized for guessing, I believe that the increased amount of guessing will result in higher curves. Only time will tell.</p>
<ol>
<li><p>I don’t know why. They should have announced this earlier, but I guess the best time would be after the AP Score Reports are released. </p></li>
<li><p>There would be a slight decrease in the percentages of 5’s, but it should not be significant though. </p></li>
<li><p>I believe that College Board is trying to make the AP exams more comparable to the tests at colleges and universities, where there is no penalty for a wrong answer. However, this is a horrible reason in my perspective.</p></li>
<li><p>I don’t think so. The new policy, I believe, was decided by the AP Committee.</p></li>
</ol>
<p>
</p>
<p>I don’t see why.</p>
<p>I believe the percentages of 5’s and 4’s will stay the same, the curve will just be harsher. With the guessing penalty, a lot of people choose to leave items blank. However, obvious statistics dictate that for those questions they leave blank, they can rule out some answers, even if its something like 0.25 answers (Which means that for every 4 questions they guess on, they can rule out 1 answer).</p>
<p>When they leave the item blank, they get an expected value of 0 points. But if they guess with the “ruling out” aspect factored in, they get an expected value of 0.25*x/(5-x), which is greater than zero when x > 0, where x is the number of answers they can rule out per question (It can be something like 0.25, like I just said above).</p>
<p>Here’s the math behind it.</p>
<p>Let’s say you can eliminate x choices for every question (Obviously, x < 5)
Thus, the chance of getting the question right is (Total correct answers)/(Total answers). Total correct answers is just 1 (Since there’s only correct 1 answer behind every question). Total answers is just 5 - x because you eliminated x of the choices. Thus, we have a probability of 1/(5-x) of getting it right. The probability of getting the question wrong is just 1 - [1/(5-x)].</p>
<p>If we do [Probability of getting it right] - 0.25 [Probability of getting it wrong], we get the expected points per question (Basically, we’re just adding the correct points up, and subtracting 1/4 points for every incorrect answer). Simplifying, we end up with 0.25*x/(5-x). Thus, if we have x=0, or essentially leaving the question blank, we get 0 as our expected value. However, if we get 0<x<5, which is basically saying that you can eliminate some answers, we get the expected value to be greater than 0. Thus, it is always better to guess.</p>
<p>Now, if we are now free to guess without penalty, then people will, on average, get more questions correct. This makes the average score higher. However, I think the percentages of 4’s and 5’s will stay the same (That is just speculation), so in order to compensate for the higher average score, there will be a harsher curve, making it tougher to get 4’s and 5’s.</p>
<p>Of course, I think this only makes it marginally tougher, possibly bumping down those SUPER borderline 5’s instead of bumping them up.</p>
<p>^Great post!</p>
<p>For an identically difficult test, the curve <em>must</em> change (and become tougher); otherwise, the standardization of the test is lost (i.e., a 4 in 2011 needs to mean the same thing as a 4 in 2010, namely, a “B” grade level of achievement). The percentage of people scoring 5’s (or any other score) will change (more than that associated with random fluctuations) only if the quality of the students taking the test significantly changes.</p>
<p>The two curves published for the same test (eg Euro history) are instructive because they suggest (roughly, at least), the extra number of points that would be expected due to the lack of a guessing penalty.</p>
<p>
How can we discuss and debate this unless one of us are CB representatives, which I highly doubt is the case? </p>
<p>As somebody mentioned in the [original</a> thread](<a href=“http://talk.collegeconfidential.com/ap-tests-preparation/940263-2011-exams-no-more-deduction-wrong-answers.html"]original”>http://talk.collegeconfidential.com/ap-tests-preparation/940263-2011-exams-no-more-deduction-wrong-answers.html) discussing this topic (I don’t understand why we need more than one thread to discuss the same thing): " The AP Annual Conference is July 14-18. Perhaps they’re making the big announcement there. "
It probably won’t. The adjustment of the curve will help maintain comparable distribution.
Probably the fact that the average performance on the exams is decreasing.</p>