CLEP Sociology grrr

[Hometeacher628]

So DD #1 took the CLEP Intro to Sociology yesterday. No less than 7 questions on WEB du Boise. REALLY??. She was getting high 67's on the REA practice tests, yet barely squeaked by on actual test. OK, it's not her field of study, so eeeking is still a pass. But 7 questions? Has anyone else had a testing experience where something was overemphasized on a test, way more than your study prep guides? This one came out of left field.

[Jonathan Whatley]

Remember, some of the questions on CLEPs are pretest questions that won't be scored. The College Board is testing the reliability of the questions themselves. It's very possible that they're "pretesting" one or more pretest W.E.B. Du Bois questions. If students who do well on the rest of the sociology test consistently do well on the pretest W.E.B. Du Bois questions, this suggests those new questions are reliable and good to add to the bank of "real" test questions in the future. If students who do well on the rest of the sociology test are, comparatively, all over the map on the new W.E.B. Du Bois questions, the new questions are probably not reliable.

[Hometeacher628]

Remember, some of the questions on CLEPs are pretest questions that won't be scored. The College Board is testing the reliability of the questions themselves. It's very possible that they're "pretesting" one or more pretest W.E.B. Du Bois questions. If students who do well on the rest of the sociology test consistently do well on the pretest W.E.B. Du Bois questions, this suggests those new questions are reliable and good to add to the bank of "real" test questions in the future. If students who do well on the rest of the sociology test are, comparatively, all over the map on the new W.E.B. Du Bois questions, the new questions are probably not reliable.

Oh, interesting. But these questions were well into the test, and she felt missing these were probably what dinged her score so much lower than her practice scores.

[Leebo]

Non-scored questions are mixed throughout the exam. You never know which ones count and which ones don't.

[Jonathan Whatley]

Oh, interesting. But these questions were well into the test

As Leebo says, that shouldn't have anything to do with it, pretest questions are not distributed in any obvious way. A test-taker will have no idea whether a question is a "real" question that contributes to their score, or a pretest question that has no effect. That's intentional – and helpful to the goal of constructing a good test. If test-takers could recognize pre-test questions, many test-takers wouldn't bother giving the pre-test questions an "honest try." An important part of pretesting is that students give the pretest questions "honest tries."

and she felt missing these were probably what dinged her score so much lower than her practice scores.

Given a 100-question test, leaving aside 7 questions still leaves 93% of the test…

[BGSU_Alum_86]

I cannot tell you how many of my CLEP testing candidates look amazed when they hear me remind them that not all of the questions they will be answering will be scored. This tells me people are not thoroughly educating themselves on the CLEP system! I am so happy to hear that some of you are informed about this, but I think it is still not common knowledge.

[hypertweeky]

So DD #1 took the CLEP Intro to Sociology yesterday. No less than 7 questions on WEB du Boise. REALLY??. She was getting high 67's on the REA practice tests, yet barely squeaked by on actual test. OK, it's not her field of study, so eeeking is still a pass. But 7 questions? Has anyone else had a testing experience where something was overemphasized on a test, way more than your study prep guides? This one came out of left field.

Don't stress about it. I took the sociology CLEP and scored a 67.. I had only studied for 5 days. It's one of the easiest.
I do not know anyone who has failed it. You will be ok!

[Hometeacher628]

[quote=Jonathan Whatley]As Leebo says, that shouldn't have anything to do with it, pretest questions are not distributed in any obvious way. A test-taker will have no idea whether a question is a "real" question that contributes to their score, or a pretest question that has no effect. That's intentional – and helpful to the goal of constructing a good test. If test-takers could recognize pre-test questions, many test-takers wouldn't bother giving the pre-test questions an "honest try." An important part of pretesting is that students give the pretest questions "honest tries."


Good point, but it threw her off when she missed so many. It seems she got rattled. She'll take the pass (albeit not a spectacular score) and move on. I really appreciate all the help this forum has given me. I never in my wildest dreams thought dd#1 would be where she is now, and I quite simply couldn't have done it without this forum's wise advice. She should have her A.A. locked up by December, then graduates high school the following May. This Sociology test was a pick yourself up and dust yourself off, and start again lesson! Don't allow yourself to become too complacent with supposedly easy tests!

[Nerheim55]

Has anyone that has taken this exam feel that Peterson practice tests are closely related to the real test? I'm planning on studying instantcert's flashcards and also the practice tests from Peterson. I took the first Peterson practice test tonight, and scored a 46% sadly.

[LaterBloomer]

While I don't know how the questions are chosen, I wouldn't be surprised to learn that questions are randomly selected. If that's the case, it wouldn't be out of the realm of possibility to have seven questions on WEB du Boise.