STATMed Q&A: Our Theory on How Working Memory Can Cause Smart People to Fail the Boards

← Back to the Blog

We see plenty of super-smart medical students and doctors fail the boards. They’re prepared and knowledgeable — so what gives?  

In this video interview, STATMed founder Ryan Orwig and instructor David LaSalle discuss a unique theory Ryan has developed over more than a decade spent working with medical students and professionals who’ve failed the boards or other tests. 

They share how your working memory may actually be to blame and identify three patterns “bad test-takers” can fall into.

Looking for more advice for passing medical boards exams? Don’t miss our series on the 5 common test-taking problems that medical students and doctors face.


David LaSalle – Hi everybody. I’m David and I’m the lead instructor at STATMed Learning. And I’m sitting down today to talk with Ryan, the founder of STATMed Learning about our philosophy on how working memory impacts people’s test-taking. So Ryan I know this is something you’ve given a lot of thought to. Can you start off, maybe just tell us a little bit about what working memory is?

Ryan Orwig – Working memory is very, very, very short term memory. This is where we are solving problems. Everything that goes into our memory has to pass through and come out of working memory. Think of it as maybe seven plus or minus two slots, seven containers to hold onto pieces of information. You might be on the low end with five. You might be on the high end with nine. I think most really good test takers are on the very high end of working memory. And the traditional thought about working memory is the higher the IQ, the higher the working memory. That’s not, not all working memories are created equally. You can be a really smart, successful person. You can be a smart, successful physician med student, but you might not have that really robust working memory. You might not be able to hold on to nine, 10, 11 pieces of information at once. My pet theory on this is that as these medical board exams have evolved, there was just sort of this blind unintentional expectation of robust working memory. That’s why when you look at these passages, I mean just went from step one to step two, the size of these things. I mean, if you’ve ever seen internal medicine passages and these things are epic in their length and volume and density and the amount of information in these things. So I think just as they’ve evolved, there was just this sort of an unintended consequence expectation that people can hold on to all this information. A group of these bad test takers, one of the central issues is they have less robust working memories. So whereas somebody sitting beside you might be able to hold on to nine, 10, 11 pieces of information. You can only hold on to seven, six, you know, that’s gonna make working a problem, these massive cumbersome clinical vignettes so much harder. So if working memory is the problem, what are you gonna do? Are you gonna fix the working memory? I mean, what have you seen on as far as just this idea of fixing working memory?

David – I mean, I think the answer to that is maybe, you know, there’s sort of soft science on it. There’s a lot of soft science out there about augmenting your working memory and you can buy apps, you know, on your phone and play brain games. And you can hope that six months or a year down the road, maybe instead of holding seven items in your working memory, you can hold seven and a half or eight items working memory. But that doesn’t seem very pragmatic for people who have board exams coming up.

Ryan – It’s just so far away, right? It’s just, you’re, practicing this thing way over here. And what we wanna do is be good over here. Our take is that the people know their stuff, they just can’t translate it on these board exams. So yeah, the idea of chasing some working memory augmentation stuff just doesn’t make sense. I’m just not seeing anything that makes me believe it. And again, we just wanna address the specific issue. Like, let me get better at these tests and by, you know, learning how to manage all the flow of that information so that it’s not overloading, your working memory seems to be at the heart of helping people get over this. So the question, I guess that, brings us back around to like, well, if someone’s listening to this, how do you know if it’s me, right? How do I know if I’m a, I think I’m a bad test taker, but is it this working memory thing? So that brings us around to talking about three of the main patterns we see. So if you experienced any of these three paths we’re gonna talk about then possibly have some working memory issues. So what would be the first one?

David – Well, you had mentioned earlier the prompt, the last sentence in the vignette, the question, the actual specific question being asked. And one of the things that I see over and over again, I see clients missing questions because they end up answering the wrong question and it’s not because they didn’t read and process the prompt initially they read and process the prompt. But then as they work their way through all those clues, as they work their way through all those answers, they have lost track of what that question was. For example, there’s a question that I use with clients and the prompt is, what part of this patient’s history would you be most concerned about? And somebody can read and process that and say, okay, well, which part of part history are you most concerned about? It’s pretty specific. And then as they go start working their way through processes, start dealing with the actual clues, you lose track of that prompt and it ends up in their heads turning into well, what part of patient history would be good for me to know about? That’s a much broader question.

Ryan – Or what’s most likely wrong with her.

David – Sure.

Ryan – Your brain is not going to allow for there to be no question being answered. Is that just going to, it’s not like, Oh, there’s a blank that I lost. The brain will overwrite and force a question. But once that working memory boots it, and you lose it, who knows what you’re gonna bring back into play. And it’s not gonna be like some flashing alert about like, hey, watch out. Like, and since there are no key, you know, 25 cent medical terms in the prompt, in the last sentence, you’re less likely to go back if you’re a bad test=taker.

David – They call that sort of oftentimes the rounding down of the prompt, right? Because one of the strategies that your brain will use, let’s say that you have a completely average working memory. You’ve got seven in working memory, but you’re trying to cram eight, nine things in there. Your brain is gonna try its best to accommodate that. One of the ways that your brain is gonna try and help you out is by sanding off some of the rough edges. Losing some of the specificity, taking maybe the prompt and making it a little more generic, a little smoother and easier to fit. It can also do that same thing with the clues from the passage. So another way that this manifests a second type of error that we see pertaining to working memory is that instead of losing track of the prompt, maybe we start to see some sort of a generifacation, some rounding down of the specific clues from the passage. So somebody might’ve read a passage and let’s say, for example, that it has in it the detail that this patient has pinprick purple spots on their lower backside, they read that they process that they take that on, but as they continue to read more clues they’re trying to cram everything in. And by the time they get down to answers, they’re looking at answers like toxic shock syndrome. And I have a lot of clients who will say, Oh yeah, toxic shock syndrome, I kept that around. ‘Cause you know, kids got a rash. Nowhere in the passage, does it say the kid had a rash? What it says in the passages, pinprick purple spots on the lower backside. But what your brain has done is it has sanded off the rough edges of that clue to try and cram it and to help you out. To try and fit it into working memory. So that makes an answer that we know is wrong. Because if I ask these people, does toxic shock at pinprick purple spots? They’ll say, no, no, it doesn’t have that. That’s not the right rash. But it takes an answer that we know is wrong and makes it look potentially right.

Ryan – Yeah and it was just generification. It’s the rounding down of something very specific and pointing to something generic, smooth, and round. And this lets them flex on their knowledge. It lets them validate something. ‘Cause then maybe the other clues really fit toxic shock. And then this is the one clue that would knock it out. And then what they do instead of ruling out, they’re ruling it in. And these are all patterns that we teach. these patterns are part of the overlay. They’re part of the blueprint. And then they’re, keeping that thing in play because it’s like, or even if it is like a rash, even if it says rash, it’s like this kind of rash or however the description could be interpreted as a rash. You’re then taking this very specific description and generectifying it into rash. And then you’re like, Oh, this has a rash over here too. And these things are all happening within seconds and it gets really ugly. And that’s how we end up as one of the sub patterns where someone misses that question, they go back, they read the explanation. They’re like, Oh wait, what did they do? They’re like, I guess I need to study toxics. I need to learn more about toxic shock. No, you knew enough. They say, well, that was a dumb mistake. Well, that’s not discreet enough. If you don’t understand the cause behind the behavior, you’re gonna keep on repeating that same behavior over and over and over on an infinite mix. Or you’re not gonna trust yourself. Once you start not trusting yourself what is it all worth? And that’s all the bad test taker, I guess compensates one of the main compensory mechanisms is they just know stuff so much more than the person sitting next to them ’cause you don’t need to know everything about every question to get a question right. And we talk about all this stuff elsewhere. Well, another example that I think of, but this one is like we something, whereas one of the initial clues in the passage is this person comes in. One of their chief complaints is muscle weakness in their left leg, something like that. And then by the time they get down through the end of the labs and all this stuff, they’ve compressed it because of working memory. They turn left leg weakness into leg weakness into leg weakness. And then when it unpacks again, it becomes default to like a bilateral thing. So they go from thinking unilateral or it’s very specifically unilateral to now in their head they’re thinking about it as bilateral. They get down and see like Guillain-Barre Syndrome. And the other clues fit well enough. Okay, keep that in play. And as they keep it in play, then they go to a tie break and they pick it. They miss it. And you’re like, and then you can ask them like, you know, lateral bilateral, like we’ll usually get bilateral. Well why’d you pick it? They don’t know what you’re talking about. Then they go back and get to you highlight the thing like, Oh, we’ll shoot.

David – Yup, yup for sure. So we talked about losing track of the prompt. We talked about having clues in the passage round down, even after you’ve read and process the passage, we still have to deal with our initial thoughts about what we think is going on when we add everything together, that status check that that you mentioned, and then we have to deal with the actual answers. We haven’t even talked about the answers yet. So I know that working memory impacts both of those things. I know that. I mean, I, and I say, I know this, I know this because I see people do it. I see people take a presentation and you know, they’ll go ahead and they’ll read the prompt and they’ll read the vignette. And if I had just stopped them at that point, before they even looked at answers and said, “Hey, What do you think about what’s going on here?” They’d be able to say, okay, well, you know what? This looks like a maybe an infection and it’s pretty acute and it’s pretty severe. But then as they go down to work answers, they’re looking at answers on the table that are not as acute or severe. We have a question that I use in the board’s workshop that deals with Henoch Schonlei purpura. Is one of the answer options that people really gravitate to in this particular presentation. And then I’ll ask them like, what would this kid be this sick, this fast for Henoch Schonlei purpura? And I’ll say, Oh no, no, they wouldn’t be. But you knew that you knew that all the way up here, they lost track of it. What happened is, they did that mental math of sort of adding all the clues together landing on maybe a hypothesis maybe not even something as specific as hypothesis, maybe just some general facts about what was going on. But then that falls off the table as we go to work the answer options again, because we’re overloading that system. So that’s an example of that initial thoughts losing track of that status check thinking as we work answers. But it can happen down in the answer set as well, right?

Ryan – Oh, absolutely. It can happen anywhere along the chain of the blueprint. We need to see where things fall off as they’re working through. I mean, you know, if that happens up in the passage on top of the clues, then you’ve got the wrong coordinates. Anything else downstream is not gonna matter at that point. Now you could get to the very end and then you sort of botch it, you lose a clue. You’d round something down. Anywhere along that chain, this stuff can happen. And again, we need to be aware of these things and it can be more subtle. It could be more in your face, but any time that this working memory gets overloaded and crashes, it’s going to put you in a bad spot. So solution wise, I mean, I really don’t have a lot of tips for it other than, ’cause I don’t believe in tips. I don’t like tips. If I were in trouble and I needed help I wouldn’t want tips. I’d want an end to end system that’s what the board’s workshop offers. But I would at least make sure that the system I’m using is not heavily burdening my working memory. That’s advice that helps somebody that’s great, right? But otherwise it isn’t a nuanced thing to bring this on board. This is one of my big, big philosophies. This is something that I sort of came up with. I’ve not read it anywhere else and we’ve built a huge part of STATMed Learning around this theory. And it does make a lot of sense and we get validation from that every day I would say with working with our students and clients from all over the country and beyond, right?

David – For sure, yeah.

Ryan- Good, good.

David – So if this sounds like a, maybe this could be you, then you probably wanna reach out and get in touch with us. If you go to our website all of our contact information is there. You can give us a call, talk to one of us directly and ask these questions. We can help you figure out if these are the kinds of problems that you’re encountering and what you can do to fix it. And you can also find additional test taking and study strategies on our blog, on our social pages, we’re on Facebook, we’re on Twitter. We’re on Instagram. We’re on LinkedIn. The best sort of clearing house for all of our information is So drop on by and give us a shout.

– Thanks for watching. We appreciate it.

– Bye bye.

Have a topic you want to see covered in our Q&A Series? Let us know