GMAT Tip: Remember there are experimental questions

While I agree with your advice that test takers shouldn't let one unusual question throw them off (the real test always has many unusual questions), I'd be worried test takers might be misled by how you describe experimental questions. I've seen many people describe experimental questions in a similar way, using words like "wonky" and "weird" and "strange" and I think many people draw the wrong conclusion from the word "experimental"; that word makes it sound as though GMAC is experimenting with new question types. That's not at all what they're doing.

Before a question can be used as a real test question on the GMAT, they need a lot of response data from test takers. Without that data, they don't know the difficulty level of the question, among other things, and the algorithm needs to know the difficulty level of a question to know what a right or wrong answer means. That's why there are experimental questions on the test. They're newly-written questions, nothing more. Experimental questions aren't any weirder than other questions, and if you see a strange question on the test, it's just as likely to count as all of your other questions.

I mention this because if a test taker believed experimental questions were the unusual questions, that test taker might very reasonably arrive at a strategy: the test taker might think it's fine to guess at all the weird questions, because experimental questions don't count. And if you could identify experimental questions, that would be a great strategy. But you can't identify experimental questions, because they don't look any different from other questions.

Sometimes people use the term "diagnostic question" instead of "experimental question", and at least "diagnostic question" doesn't incorrectly suggest a question is in some kind of avant garde format. But I don't care much for that term either, because it suggests the question is used to diagnose the test taker, and that's what all the test questions besides the diagnostic questions are used for. So both of the terms are misleading in one way or another.

edit: just after I posted this reply, I looked back at the first post in the thread, and it had been mostly erased, and replaced with a blog link, so now it might look like I went to the blog to quote from it. I can't imagine I'd ever do that; what I quote above is part of what was posted initially in this gmatclub thread.