Changing answers - I'd appreciate your quick thoughts

SDN Members don't see this ad. (About Ads)

On p. 20 of FA2012, it says, "studies show that if you change your answer, you are twice as likely to change it from an incorrect answer to a correct one than vice versa. So if you have a strong 'second hunch,' go for it!"

I have personally found that most every time I change an answer on marked questions, I end up getting the question wrong.

Each person obviously differs, but I just find it interesting that FA's advice has not worked for me on Rx QBank.

Does anyone have any thoughts? / What do you find best?

I also change it to the wrong answer more frequently. I did an experiment with this when I was studying for the MCAT and found I changed to the wrong answer about 85% of the time. I thought it was 50/50 at worst, but the reality is that the ones I change to right answers stuck in my mind more, and seeing an 85% fail rate shocked me. These answer changes are the last minute or last second (almost "reflexive") changes, and almost always wrong for me. So, at least for me and it appears also you, the answer change should be avoided. The hard part for me is actually sticking to the plan of not changing my answer, unless I know for sure what I put before was wrong.

I was actually thinking about this yesterday. On Rx it says I've changed from incorrect to correct more than 3x more than I've gone from correct to incorrect and 2x more than correct to incorrect + incorrect to incorrect. I think a lot of this has to do with the fact that I am much more lax about clicking an answer on the qbank questions than I am on a real test, so I wind up reading a question, picking one really quickly, and then really thinking about it in depth. Judging from my past real test experiences I suspect that I'm in the same boat as you two.

I think a lot of how it works for you individually is how you review your questions and your test as a whole. I think my hardest about every question as I come to it, I don't skip any, and I mark the questions I'm not 100% (or at least 95%) sure about. I then go back and look at the marked questions, but very rarely do I change an answer on any of them, it's more so I can see how many out of the total test I was borderline on or didn't know at all. I think my propensity to not skip any questions and do all of my in-brain tug-of-war before I pick an answer and move on to the next one makes me more likely to change to a wrong from a right answer. If someone is more inclined to just pick an answer after a certain amount of time, mark the question and move on to come back later though, it might be a different story.

Originally Posted by Phloston

On p. 20 of FA2012, it says, "studies show that if you change your answer, you are twice as likely to change it from an incorrect answer to a correct one than vice versa. So if you have a strong 'second hunch,' go for it!"

I have personally found that most every time I change an answer on marked questions, I end up getting the question wrong.

Each person obviously differs, but I just find it interesting that FA's advice has not worked for me on Rx QBank.

Does anyone have any thoughts? / What do you find best?

Actually if you use UWorld/FA/USMLERx, for every test you take, they will let you know how many times you change from wrong to right or vice versa. I was stunned to realize on all these testing platforms that I usually change my answer from wrong to right ... and by a large margin. Of course, I had always thought that I change my answer from right to wrong, or maybe it just seems that way because we're more likely to remember when we do that. So I am not surprised by the FA research at all. And I have changed the way I test because of it; I no longer have that "guilt" about changing my answers, because I know I will most likely change it from wrong to right. I actually thought as I finished the QBanks that it would even out, but it definitely hasn't. I've finished UW and I'm almost done with Kaplan, and on both I change my answers from wrong to right by a huge margin. I also change my answers from wrong to wrong a lot too lol. But changing from right to wrong for me is definitely not normal.

It's actually personality-dependent. I don't change my answers after the initial debate.

My friend changes it from right to wrong most of the time. Lol.

Originally Posted by Phloston

On p. 20 of FA2012, it says, "studies show that if you change your answer, you are twice as likely to change it from an incorrect answer to a correct one than vice versa. So if you have a strong 'second hunch,' go for it!"

I have personally found that most every time I change an answer on marked questions, I end up getting the question wrong.

Each person obviously differs, but I just find it interesting that FA's advice has not worked for me on Rx QBank.

Does anyone have any thoughts? / What do you find best?

It's about time there's a study to confirm my quirk! lol. Everyone used to say the opposite. I almost always change my answers from incorrect to correct. It's frustrating that my first instinct is wrong so often, but I find that when I look at the question a second time, I either misread it or I think it through more and find a different answer. Just to give you an idea, on my last school exam, I changed 11 of my answers when I went back over my exam. Of the 11, I changed 7 of them from incorrect to correct, three from incorrect to incorrect (lol), and only one from correct to incorrect.

I think it depends on the question. After doing enough questions over the course of my life, I think I've learned to tell when it's better to "go with your gut" and when the "second hunch" makes more sense.

studies show that if you change your answer, you are twice as likely to change it from an incorrect answer to a correct one than vice versa. So if you have a strong 'second hunch,' go for it!

There's even been a medical student participant study -

Originally Posted by http://carmine.se.edu/cvonbergen/Answertng%20Multiple-choice%20Questions%20in%20High-stakes%20Medical%20Examinations.pdf

Therefore, students need to be counselled to follow
their doubts and rethink their answers. The chance
that a substitute answer will be correct and will
increase an overall examination score is statistically
worth taking, providing the examinee does not
change horses more than once.

Phloston - are you just assuming that you change to incorrect that frequently or are you keeping meticulous data for all your answer changes? There's definitely a recall bias for those you change from right -> wrong to stand out in your mind

I think I may not have been specific enough the first time.

I too have many more "incorrect to correct" changes than "correct to incorrect" ones under the statistics in Rx, but these apply to any changes made while answering questions, including those when passing a question a first-time.

I was meaning to say that once I finish an entire block, when I revisit only the marked ones, it is then that the changes are predominantly "correct to incorrect."

When I think of the practice of changing answers, I think of it in terms of when I go back to review the questions with the remaining time left at the end of the block, not while answering questions on the first-pass.

I'm trying to understand whether it's generally good practice to change already uncertain answers when revisiting them with the x# of the minutes left at the end.