Memory - Hacking your memory

[Auto-generated transcript. Edits may have been applied for clarity.]

Loving everyone. Happy Friday again.

People seem lively, which is nice. Um, looking forward to the weekend.

So this is the last lecture in the series of four that I've been doing on memory.

And today, what I want to do is bring together some of the principles that you've heard about in the

first three and talk to you a bit about how you can apply them at university and beyond.

And obviously, I've talked a bit about that already because I'm quite keen on this topic.

Um, but the example studies that I'm going to go through today are more applied.

One so testing the principles in more applied settings and showing how they work.

So what we're going to cover is, um, paying attention and practice and how to take notes.

We're going to talk a bit more about some applications of deep encoding and elaboration,

and how you can take advantage of what you already know by how you prepare for lectures and how you process things during and how you revise.

We're also going to consider applying practice testing and when to do it.

Um, and what a lot of this ties in with is basically the principle that you will need to generalise your memory eventually.

And we have to talk about this even though we don't have a lot of time to think about the mechanisms by which this works.

But what I want to do is to point you to other techniques that I'm suggesting a good to use for learning,

or ones that also help you generalise your memory, generalise your memory, and I'll explain a bit more about what that means.

So is anyone here? Maybe people who come in from the States, but has anyone heard about laptops being banned in class?

I think this was popular in the United States at one point, and some lecturers telling people, you mustn't have your laptops.

Um, partly because they're obviously worried that you might be doing something other than paying attention to what they're saying.

Um, but there might be another reason behind it, which we're just going to talk about now.

Um, unless it's just, you know, people, old people being scared of technology.

Um, so before I do that, just think briefly back to the sort of thing that you can do to make things memorable.

At least the concepts that that you're hearing from me and other lecturers.

So, um. I don't know if you're on the pole everywhere but the pole everywhere is open.

Um, it was on the first slide, and you can grab it off the slides again.

So does anyone here want to vote for repeating the exact words that I say as being a way to remember what I say?

31. Um. Want to kind of vote for a Zen approach of letting the words wash over you so you can absorb them?

What about number three, where you sit and think, does what she's saying actually make sense?

Thinking about why it makes sense? I'm getting a few hands here.

And what about thinking about how the content relates to what you learned last year?

I'm getting hands there, too. Okay. Right. Not suggesting that some people, um, understand, um, and remember what was said.

Someone says it's easier to ignore the lecturer and just copy paste the slides if that works for you.

Good. I'm totally happy for you to ignore me if it works for you.

But that's not what the site says. So what? The site says anything that's basically copying and pasting the information,

whether you're doing it by rote repetition or whether you're doing it in writing, is basically shallow encoding.

It doesn't allow you to do much processing for the meaning of the information.

And generally speaking, processing for the meaning is one of the things that really helps you remember stuff.

So thinking about whether I'm talking sense, there's one thing that's useful.

I'm thinking about how it relates to what you already know is another thing that's useful.

Unlike in most multiple. Well, unlike in all your assessment, multiple choices, there are two that are correct here.

So let's think a bit naughty there, but seem to serve the topic.

So the laptop thing is partly about that, but it's also partly about something else.

It's partly about distraction. As I hinted when I was talking about the idea that people might get out of bed

in the morning and come to a lecture and then sit there and do something else.

I kind of do understand it. I go to meetings sometimes, and if I do something else at the same time, even though I know it's not terribly helpful.

Um, so, um, is it a good idea to to switch in between listening to me and going and looking at social media that we do know the answer to?

That's definitely not a good idea. Um, unless you know that you don't want to do the assessment.

Um, because people learn, I would say it's surprising they only learn 11% less.

Um, but they learn in that study by saying, alright, so, uh, 11% less than people that are not multitasking.

You don't know what the people are not multitasking doing. There's a blog about that.

So I've linking I'm linking in this lecture to a couple of blogs because partly because there's a group called The Learning Scientists from

Dundee who've done a lot of practical work trying to get information out to people about how science can tell us better or worse ways to learn.

Um, and also just because it potentially makes it a bit more accessible.

And you can go back to the textbook essentially.

Um, so multitasking can also affect people near you.

Um, I should just say about blogs, a little caveat that if you're writing an essay,

then you need to cite the studies that the blog is referring to, as well as the studies that you, um, learned about in class.

Just because we ask you to try and cite the evidence as in the published evidence, but that can really help you understand.

So in terms of laptops, um,

there was a really famous study that showed that people that took notes on a laptop

tended to take more verbatim notes and more exact words and less in terms of a summary.

And this, again, is the shallow processing. Um, but there have been some replications and some meta analysis.

Well, at least those two that I've cites here that make it look a bit less clear cut.

Um, one of the reasons for this is that when you're looking at notetaking, um, on a laptop,

and when you're looking at different people who take notes in different ways, it's an observational study.

So it's not a study where you're sort of randomising people into one condition or another.

So it may be that the people that choose to take verbatim notes are different people, but it's been a bit controversial.

Um, but we do know that notetaking can help you avoid mind wandering.

Um, as in the study by Wong and Lim. Um, and the other thing that complicates the picture, and it's.

This is why I'm not frowning at people who are sitting there with a laptop, is.

It's very likely that there are ways to use a laptop. Well, to take notes.

You just have to resist the temptation to just write everything down and think, oh, well, I'll understand it later.

Um, and a student who has dysgraphia has written a blog, um, about what he's learned, um,

as, um, someone who has to use a laptop in order to take clear and sufficient notes.

So what's it for? So why take notes as you do it, so that you think more about the meaning,

and you elaborate more on the material, more about what elaboration means in a minute.

And you can do it to provide this external storage. So you don't have to remember everything all the time in your brain.

And you can have this nice interactive relationship with what's out there in the world.

Um, and you can use what's in the world to help you remember and then to help you implement the testing effect and get it into your,

your brain later in a more long lasting way.

So, um, external storage can help with all of that. Um, and they can interact together, these two factors.

So good notes can be useful for later. They don't all have to be made in lecture, obviously.

We'll come back to that too. Um, but practising making good notes is worth doing.

So tips for how to do that. I'm sure that quite a few of you do use electron slides.

Using the lecture slides to organise organise the notes is good.

And again, um, rather than typing things sort of verbatim, you can just put little notes on it.

So you can do that if it's on a PDF version that you've made or that I've made.

Um, or you can do it on paper where you actually scribble things down.

Um, and you're basically elaborating the actual writing on the slides.

But also think about the main messages, which we do try and help you with, but it may be that your framing of it can be more helpful than my framing,

because it relates it to what you know in the specific way that you know it and you understand it.

Um, so if you look things up later, explore readings, or ask me and the others what is a clear there's some questions on the discussion board.

I still have to get to. I hope that today's the day. It's been one of those weeks, but I've seen some really good questions lining up there.

So thank you for those. Um, so you can also think during revision about how you make the argument yourself, like the argument that I'm making.

How could you explain it in your own words? Um, so how could you implement it in an essay?

Um, and one of the ways, certainly, that I thought was very useful myself when I was a student,

which I did as a mature student as well as, uh, um, just sort of have an argument in your head with it.

Think, well, has that really been shown? Is there any conflict here?

Can I put these things together neatly, or do they maybe not quite fit?

And that helps to generate questions. My usual memory.

Care for the pin? Wants me to log in again.

Okay, we've got a pin. It's 28 eight.

2088.

2088.

But if you hand or put something on pole everywhere, if you use it again.

Someone was mentioning rewriting what you saw in your own way on the, um, poll everywhere, which is a good way to do things.

So what do I mean by elaboration?

Well, to some degree people talk about elaboration, but they've only really understood it in research in relation to specific strategies.

I mean, you can imagine why because, um, you can't just read what's going on in someone's mind,

but you can say, well, I'm thinking about what it means. I'm relating it to what I know, but what does that actually mean?

And so quite a lot of the research that's been done on elaboration has been done in relation to specific mnemonic strategies.

Um, and these strategies mean particular ways of actively relating incoming information to what you already know.

So it's basically a deliberate strategy. Um, how to say and sometimes hard to use.

But they can be useful, particularly for certain kinds of information.

There were lots of mnemonics, um, when I was a medical student in a different, um, life.

And, um, they were always really annoying mnemonics related to things I wasn't interested in,

like sport and sexist banter, but, um, you can make up your own, um, more on that in a minute.

Elaborative interrogation is a mnemonic that allows you that, where you generate an explanation for why something is true.

So you basically just say, okay, why is that? You say that in my own words.

Um, self explanation is quite similar, but it's explaining how the new information is related to what you know,

but making that explanation explicit rather than just thinking, oh yeah, I can see how that fits.

Um, and then there's a particular one called the key word mnemonic.

Um, which is where you're using, um, this um, well, you're giving yourself keywords in,

you're learning it and you're using imagery as well as meaning elaboration to associate verbal material.

So it's very good for situations where you need to not learn like, um, a word and the thing that it's associated with or like,

like I say medical students, a drug and a treatment or something like that.

Um, so this is an example of, um, use of the keyword mnemonic.

Uh, if you want to learn the hippocampus and learn that that is the name of a brain structure that's important for memory.

You could come up with an annoying GIF. Um, you could think about a hippo and a campus hippocampus.

Is the hippo visiting our old university campus.

And this bringing back lots of memories. So I see it's a bit of work, right?

It's a bit work, isn't this?

But if you've just got facts that don't have any intrinsic link, at least to each other, then it may be that you've got to use something like this.

Um, and when you do it, surprisingly, um, using this kind of elaboration because it seems a bit random does help, um, generalise outside the test.

Um, and that was the study by Richmond.

Detail, the elaboration interrogation one where you've got to generate an explanation for why something's true.

Um, is a bit easier to use. And, um, it's reviewed by Dan Lawsky.

This is a very long paper on the further reading list that I'll give you.

So nobody's expecting anyone, even the best student in the world not gonna read the whole thing,

but it is useful sometimes to dip into an they've got all of the applied research up to 2013.

And what they basically found was that this mnemonic, um, it was effective for people who knew less about the topic.

Yeah. Um, a little bit of an increase in, um, final test performance, but it was more effective for people that already knew more about the topic.

So there is an interaction between your ability to do elaboration and your prior knowledge.

Obviously, the more you think about things, the more you understand, the more you know.

And so there's a potential iteration there over time, but it does require you to know something about it.

Um, and this reminds us, hopefully you as well as me, of another aspect of relating prior knowledge to, um, what you're learning.

And that's the idea of a schema. So.

Think back to the bank historian study. The information's sort of summarised on this slide, but see if you can remember for yourself.

So this is one of the ones that I think is quite important. Um, in assessments we never ask you like, okay, what was the year?

I've just updated it actually because I was a typo. We never say, well, what was the year that bank published a study about schemas?

Because that would be a ridiculous memory test, not a test of understanding.

But we might ask you about studies giving you some information about the study.

So you might want to say what did they actually do. Um, what did the behavioural results show.

Did the manipulation of actually affect memory and how did it, um, reflect in the fMRI brain activity that was related to cause performance?

So giving you a question here that tells you a bit about the study,

this is just an infographic that summarises the results that basically says the medial temporal cortex activity,

um, was um, that that was related to the schema related learning.

And the hippocampus activity was related to the non schema learning.

And so the schema related learning was new facts that related to what people had done in the first year.

So like you guys doing psychology and the um and the non schema the hippocampus

was needed more for was the incoming information that was completely novel.

And the fMRI brain activity, um did relate to close performance, particularly the medial prefrontal cortex.

So they did a follow up study that was also an applied study looking at schemas in education.

You can see that some sort of medical kind of stuff here. So the example.

Yeah. So the principle of this task is that people study pairs of words and then they have to generalise between the two pairs.

So the way it works is you get um a pair A and B and you learn they go together and then you learn and see.

So the same as something else. But you're not explicitly taught B and C.

The idea is you've got to generalise and make the connection yourself.

The B and C go together from that information and you think, oh, maybe that's hard, I don't know.

But you do it all the time. We all do it all the time in the world outside.

So the example here is trying to learn about, um, ADHD and medicines that can help with ADHD.

That artwork is pretty terrible.

There's a picture of a child being hyperactive, and there's the name of methylphenidate, the drug that is, um, used in some people with ADHD.

So people also learn those two together A and B, then the US to later learn um and C so these are new pairings, the same drug name.

And here they're learning the fact that it was that it is a medicine for ADHD.

So um, the task is to generalise and then A, B and C and they're manipulating how relate A, B and C all.

So if all of them all fit together with what you know already,

but you hadn't actually got the name of the drug before, for example, then that's a schema related bit of learning.

But if the, um, other thing that went together with methylphenidate was something that you didn't know anything about before,

like it was discovered by so-and-so and such and such a or something a bit random, then that would be non schema related.

So they did quite a lot of stuff in the study.

But the main message is that in the final memory tests, after they are people just to recognise, um, what they studied.

Um, in C they asked them to try and recall given um the, the, the Q of, of C here what the original B was.

So that giving them medicine for ADHD and asking to try and draw the picture that went with it was quite difficult.

And then they were also asked to judge the um, um, the A as well as a test of this generalisation of learning.

Um, and uh, I'm obviously telling you about this because it relates to schemas.

So you won't be surprised to hear that memory was better for these non studied pairings,

these inferred pairings, if they were congruent with people's prior knowledge.

And you can see that in the distance between those two lines.

Um but the interesting thing as well was the people's idea of what they were going to be able to remember.

They thought that, um, doing the interim memory tests, which were also part of this um, study, would help them.

But they didn't think they didn't. They weren't aware consciously that the schemas were going to help.

They underestimated the importance of prior knowledge here. It's been replicated as well as result.

So in terms of using it at university, the essential reading paper, which is very short and full of good advice by Putnam et al.

That involves quite a lot of good advice about how to use this principle.

So some of it's my job, of course, but looking at readings before lectures can help us.

Can revision quick revision of a few key ideas, which you can do in like ten minutes.

If you're sitting here waiting for me to get the computer set up.

Um, reading over slides that have been sent out can also help because you can think, what do I already know?

What what what sort of relates back to year one and what don't I know?

Um, but you can also do it later.

That's the nice thing about these kind of memory effects is they do, um, uh, work basically if you do some relating it later on afterwards.

Now this is just a summary slide to summarise the things that I've said just now in a more applied setting,

and as well some of the things that I mentioned only back in lecture one put them all together

because these are strategies that can help you basically catch that incoming information,

catching that moment, as it were. So you're catching it and getting it into your mind.

These strategies, which I'm not going to read through now, they're all about encoding.

So whether it's preparation, attention, distinctiveness, processing, processing for meaning,

organisation and notetaking, they're all about getting the information into memory.

But we are going to think also about what gets it to stay there.

We looked at it a little bit in the important, although rather complicated, Burkman and Roediger study,

where they were looking at it in relation to true versus distorted memory.

But we're going to look a bit more about what happens after encoding in a minute.

Um, and this is really a limit of improving encoding.

So, um, it's going to be about the fact that getting and getting information into your brain.

Is necessary, but it's not sufficient.

And so in the long term, it's the stuff that you do in between that can make at least as much, if not more, difference.

And there's a quote from Los Cartel saying that basically, yeah, elaboration, interrogation,

this laboratory of interrogation, this strategy mnemonic isn't isn't brilliant.

It's not quite as good as you might think. Um, so it's better than just verbatim rehearsal, but, um, it it's not the absolute best thing.

Um,

and I don't know if you're going to be able to guess what I'm going to say is the absolute best thing for memory when you need it for the longer term.

Anyone already guess what I'm going to say. You put it on the pole everywhere if you want.

So for long term learning, um. Also put questions on the poll everywhere if you got them.

But do you have a guess? Well I'm good. Well, I've got to turn to you next if you don't know already.

So we're basically moving on to the memory test. Um, and the fact that testing is important as well as studying.

Um, so you get memory cues that are external, uh, different things that matter during testing.

You get memory cues that are external. So I showed you those pictures of scenes in lecture two from the Smith style study,

or going back to the same lecture theatre and having memories triggered of what happened there.

Um. You can also self generate views.

And I covered that also lecture two and says that this idea of mental reinstatement.

They found brain activity suggestive of mental rate a statement in the study by Perlin et al.

But you can just do it yourself by thinking back.

So basically thinking back, does that fool you? Thinking back, oh, what did I learn last week?

You think about something that you know about last week to help trigger your memories of what was going on.

Um, but the other interesting thing in terms of moving from study to test,

is that what you do during studying can actually help you have more possible ways to retrieve information and test the people.

Find this a little bit difficult to understand, because it is about the way that memory is an interaction,

or a point of contact between a studying episode and a test that you somehow later did.

So self-generated cues, um, these are thought to be a second way that elaboration works are.

Elaboration works partly as processing for meaning, but it also works, um,

because it helps create cues that you can use later that connect with your personal knowledge.

But it's all about the cognitive context, so it's about what you connect it to at the time of learning.

Um, so here is a kind of mnemonic strategy, um, example that is given by Tullis and Finney, um, who say that, uh,

somebody who wants to remember that Rutherford was the first person to show that each atom has a nucleus at its centre.

Um, this person studying chemistry can create a cue for themselves, so they can create the link that says,

Ruth is my grandma's name, and grandma is the centre of the family.

And the idea is that that will help them to remember it later.

Now, of course, you got to remember the cue will be given the cue, but that is part of it.

Okay, is semantically related to the target fact.

Um, because they're talking about being the centre as well as having a cue that's part of partial key to the name.

Um, and here's where the this was my reference to annoying mnemonics that that used to happen back in the day by medical students.

Um, you can use your own interests, basically your own interest, your own perspective, your hobbies, your way of seeing the world.

You can use anything, um, as a mnemonic to link things together.

So one way that people do it is by having the first letter. If you got a lot of disembodied facts to learn, then you, you make up a word, um,

a sort of acronym, and then you learn the acronym, um, people that can make rhymes.

And the reason I'm not going into this in detail is because it's not that useful in psychology, because most of it is not disembodied fact.

It's mostly stuff we want you to learn that it's stuff that you can include more semantically,

but you can use your own interest in the same way that person did in the example.

And you'll find probably that those are easier. You should find that easier to remember.

So this is just a reminder of what the encoding specificity principle is.

This is a principle that says that memory is better when the context of a memory matches at encoding and retrieval.

Context can be something simple like time and place. So it can be the location, as I've already given you the example of that.

Um, but it can also be your thoughts. And here it's your thoughts, your internal context that's most important.

So if you think about something in relation to something else when you first encounter it,

then the two are linked in your memory, and later on one can be used as a cue to remember the other.

And so when you generate stuff in your mind, when you're studying information that helps you have a cue later on,

potentially more different cues that can help you remember.

So self-generated cues work better than cues that somebody like me has made for some lecturer or teacher or your friend,

even, um, and they can be external, you can write them down and use them to remember things later that way.

Well, if you're in an exam when you don't have the external care so much, although you might have them in the question.

Um, but you can also remember the cues better if they relate to things that are personal and important to you.

And these keys will match better with your personal cognitive context.

So the thoughts that were stored at the time of encoding.

And they will probably, although I don't know how well this has been looked at experimentally so far.

They will probably be more diagnostic because you should aim to make your cues sort of specific for that particular memory or that particular fact.

Um, and if you do this, then an encoding specificity will help you remember things for later.

Okay. Annoying gif the most annoying gif of all of them. Does this remind you of anything for lectures?

Little to. I have to go like this after a certain time.

Does anyone remember what that is? Why was she amazed?

Someone says is testing better for long term? Yes, testing is better for long term.

So the testing effect, the largeness of the testing effect, um, this is just a reminder of the result I showed you before.

Um, and it was about memory for ideas, units and prose passages.

So is life like ish? And there was 10%, um, forgetting, um, after a week,

people who had had multiple interim memory tests versus 52% for people who had

just restarted the information even though they restarted it several times.

Um, so you might think, okay, well, this is all very well, but the stuff I've got to remember is a lot more,

um, than it's it's more sort of general and less specific.

I don't have to remember a whole prose passage, but I do have to remember a lot more information than that.

So could testing be relevant to you? Um, here in your psychology course?

Well, Ayanna Thomas and her colleagues did a study doing exactly that.

And here, um, looking at testing in the form of quizzes that take place during a cognitive psychology course, just like the one you're doing now.

Um, and you're going to guess the punchline, because otherwise I wouldn't be standing up and waving my hands about about the testing effect.

Uh, but you're going to maybe now understand why we're giving these quizzes for each topic in this course, even before the memory section came along.

So what they got was people to do quizzes after lectures at different points in the course.

There's a diagram of the whole study design that I put at the end of the lecture slides that you can refer to if you find it interesting and useful.

Um, but basically the outcomes were multiple.

So they were unit exams for separate sections within the course.

And they had multiple choices on those that were not the same as the ones that were done.

Um, um, and the practice tests.

And they had short answer questions, and then they had multiple choices that were old and new, and they had essays in the final exam.

And so what they were looking at here was not just does the testing effect work in practice and in the class, because we know it does.

There's lots of research that already shows that. But can you transfer the learning to related concepts.

So a bit like that a b a c? Um, Vagn question.

I tell a study that I just ran through. Um, can you, can you do that in this context?

And can you also get transfer of learning to a different test format?

So if you practice learning by using memory tests with an MC,

does that allow you to use that same knowledge and remember it in the context of an essay or show answer?

Um, a the answer is yes and yes.

So I got a 12 to 14%, um sized effect.

So here we've got quizzes which is testing quizzes with feedback.

A great and just studying is um, is white.

So we've got about uh, 83%.

This is on the conceptually related multiple choices, 83% here in exam one versus 71%, um, that do very well in America.

Everything's really easy. And so we will get 70%.

Um, but. And it's the same principle.

So, um. Wisdom was better than studying.

Feedback did significantly improve. There's a sort of predation, you can say.

But it was nonsignificant. We'll come back to that.

Um, then for the conceptually related short answer question.

So this was a different format. It's actually, if anything, a bigger effect.

The 13 to 21% bigger. Might be surprising.

You might think that you'd see a greater impact for the the same test as for a different one.

Um, so just to summarise, what the study found was that testing memory benefits not just the,

um, the same type of quiz, um, for the same type of content.

It transfers to related material and it transfers to different test formats.

So this supports this idea that I've mentioned, um, in relation to other findings,

that testing actually supports concept development and learning and a much more general and interesting way than you might think.

So it's probably aiding the development of semantic memories that are more general and end up being context free,

as well as memories for instances like a given lecture.

So this is all very much of a piece with work that's been done on memory updating.

Um, and there's another blog if you want to see, uh, an explanation by somebody other than me in a different context to help you understand.

So. You might wonder why people do exams.

I bet you've wondered why I do exams. They are pretty stressful.

Um, but there is a case to be made. I'm personally not making the case that you should just have exams.

Well, there is a case to be made for exams. Uh, something that helps people to learn in a way that is more general and more useful later on.

And that does therefore have some benefit, other than being easy for the people running the assessments or it's not easy marking time.

It does take work. But that's Diana Thomas and she's got a YouTube video explaining.

And so you can see if you agree with her or not. Um, so in terms of evidence based revision.

This is a summary of what I tell fans as of 2013.

So lots more has been done since then. But. Things to know are definitely not useful.

So highlighting highlight happens. If it ends up like that, it's probably not useful for anything.

Highlighters.

Personally, I use them so when I go back to a paper or read a lot of research papers, I can pick out the bits that I found to be important last time.

So it's good for that. It doesn't help you remember for later. It's terrible for memory.

Um, summarising, um, as you read is also pretty terrible.

Rereading. Really quite bad. So just if you're sort of sitting there like in the summer, if you got summer exams,

I remember sitting out and people pretending to be revising and you take your revision notes outside and sort of half, half look at them.

So it's just no point. Take a break, go outside, sit and enjoy this up, take a break and then get back to studying in a more active way.

These elaborative, um, strategies do help and moderately useful, but they're not, um, as good as the best thing, which is testing yourself.

It just isn't anything else that is as good in applied settings as well as in, um, you know, face settings.

So. So, yeah, I mentioned these, um. These issues.

So I think I just want to make it really clear to you that not all the things that help the encoding help a memory test that happens a lot later,

say, more than a day or so. Really important to know that getting them in versus getting them to stay.

Necessary, but not sufficient. Um, and so out of the things I covered in lecture one that we talk about the encoding and the one short memories,

the ones that directly benefit longer term learning, are really only the, um, elaboration and the strategies to help yourself generate cues for later.

They do need to be combined with testing, and we do need more research over the long term.

And I'll come to the timing in a second. Um, but briefly, I just wanted to return to feedback.

So why do we bother giving feedback on practice?

Multiple choices. Given that Thomas arsehole found that it had no significant effect.

Well, the reason is. Well, the principal reason is that, um, finding a non-significant result doesn't mean that the effect isn't actually there.

There can be lots of reasons why a result is not significant. So finding no evidence for something doesn't mean the thing is false.

Um. But the other really important reason to use feedback.

And the reason that most people think it probably does matter, um,

just hasn't been studied enough in the right way, is that we need to avoid testing introducing misinformation.

So in the Berkman and Rodrigo study, having an immediate memory test dramatically improved memory later, even at six months.

But it also dramatically increased the number of, um, distortions that people remembered.

So that may be less important when the material is schema relevant, is easy to understand,

or at least comprehensible, unlike the story to the people, um, that that were given it by Bartlett.

Um, but it seems wise to get feedback just to be sure that we're not introducing misinformation.

But yeah, no, uh, more research is needed. There's a systematic review being done by Agarwal et al.

And, um, there's also more research needed in non, um, white western um, um, university college settings.

Um, but yeah, the retrieval practice effect still comes through in that matter analysis, even, um, when it goes across a wide range of contexts.

So revision timing is the other thing I wanted to just mention to you.

There is a lot of work on this that we don't have time to cover. You can if you want to look for more, you can look at this matter analysis.

You can see how much has been published on it. In 2006.

There were 184 articles. So a lot is known about this timing issue.

So this is referred to as the spacing effect. And it's basically about how you space out your revision.

If you've got to learn for an exam that's a bit later.

Um, you can do it all at one point in time, or you can space this out over time in multiple sessions.

And the the short message here is that spacing it out is better than bunching it up.

So that's spacing as opposed to what they call massed learning.

But I don't need you to learn that term. But spacing is good. So just conceptually be aware that spacing out revision is good.

This was a study where they varied different kinds of spacing basically.

Um a long lag is particularly good if, um, you're needing to know it for more than about six months later.

Um, but if it's in between amount of time, then doing a more immediate revision and then doing something later.

So even if you just do two is a lot better than one, and if you space them out, it's better than bunching them together.

So. So yeah. So how given all of this and what I've been saying about memory updating,

how just the episodic memory for the unique event relate to what you need for the longer term?

It's complicated, but it has to do with memory updating.

And this is kind of a nice study that was done in students back in the 90s by Como et al.

And one of my PhD supervisors, Tim. Perfect. Um, and there's been a replication recently that that does replicate the main findings.

So what they did was they tested psychology students memory for information in a lecture,

both immediately after the lecture and in a delayed test exam.

And they found that students did better in the exam, were more likely to remember the facts in the particular lecture at the immediate memory test.

So that's a context dependent memory in the sense that it's a memory that includes where it was learned and when.

So that's an episodic memory. So the better performing students tended to remember the actual study episodes.

But by the time of the exam, at the delayed test,

they were more likely to just know the facts and not necessarily remember whether you've gotten from the lecture,

whether you've gotten them from the reading, or whether you'd kind of inferred it somehow yourself.

So there was a transition between the episodic and the semantic memory.

Um, and a shift, um, that allows people to generalise beyond the initial context.

And we can guess, but they didn't look at it that this was attributable to revision.

But the other stuff that I've just shown you, um, does suggest that that is the case.

Um, yeah. Methods, courses. It's a bit different.

Just knowing things is more important, and you probably know that from doing your own and just knowing it,

and you're doing more learning but doing and is more procedural as well as semantic.

But that's not part of today. So episodic memories do get cemented over time.

Um, you can find reviews if you want to understand that better, but we're not going to test you on those mechanisms themselves.

Just be aware that repeated retrieval is probably really important to not.

Um, sleep helps to. So if you can't study, then take a break, have a nap, and then study again later.

So repeated retrieval or testing and spacing are really good for the long term.

Um, but do make sure the tested material is accurate.

So, um, give yourself feedback. So monitor your memory, but never completely trust it.

I don't trust mine. Um, you've seen me make mistakes.

Hopefully small ones. But, um, everyone does it, so checking is really important.

And the episodic quality of memories in the longer term is not always lost.

Quite when it's lost and when it's not lost is a bit controversial. I know Chris Bird mentioned that.

But again, beyond the scope of today.

Main point of today was try and try to understand how, um, the principles that I taught you before apply outside the lab and to see how they,

um, apply to learning at university and to be able to potentially discuss that.

And also, it's a life skill, getting to know your memory, knowing how to get the most out of your memory with at least effort.

You know, it's hard work at the time, doing testing. And there's some evidence that harder tests are slightly better.

But, um, doing it is worth a much longer period spent studying will do less good.

So that's just a summary slide on active learning at university.

Um, that I hope you found useful. I'm going to look at the example short answer question multiple choices now.

So if you're not already on the poll everywhere this will be good for your memory if you have a go.

Um, I should have probably put the, um. You all code all this slide, but you have got it on the lecture slides.

So what kind of information does a recollection typically contain.

So this is sorry this is a question with 31. questions.

So each one just needs one thing. So pretty brief things.

What kind of information does a recollection an episodic memory typically contain that's different from a semantic memory.

And then. What's a mechanism by which an external cue, such as a photo, can trigger recollection?

And what theory explains how processing during memory encoding is related to processing during memory retrieval.

We're getting lots of right answers here, which is great. Anyone else?

Well, I guess you can see it on there. So people are saying for one context, context of the event.

Yeah, you get more if you just say context. What about the mechanism?

I which the external queue. Can work.

Slightly harder. The theory.

Anyone want to hazard a guess is to eat with us. Bring them in now.

Totally anonymous. Yeah.

So somebody said patent recreation in the hippocampus.

I'll give you that. It's patent completion is proper term.

Um, that's one of the mechanisms. Almost a flashback.

I think that's too general. Um, what are you trying to achieve?

In a sense, is a flashback, which is retrieval. Thomas Hood turned the right thing for the third question, encouraging specificity.

Well done. That person takes a bit to be confident about these things.

Um, so I will post the answers when I update the slides after the lecture.

But yeah.

So, um, context reinstatement or pattern completion for the second one and encouraging specificity or transfer appropriate processing will also do.

So. What's the right answer here? A longer delay between encoding and retrieval of a memory one reduces false memory,

two reduces stereotyped memory errors, three improves recall of specific details, or four none of the above.

Who wants to say, um, three. Who wants to say full?

Got some volunteers. I know it's after a whole hour of hearing about memory, but that's correct.

Four is correct. None of them, because there's the opposite, basically.

Um, in all cases. Um, so.

This lets him. This is a little bit out there because we didn't talk that much about stereotype.

Harris. But you need to remember that they're related to schemas, bias and stereotype related to schemas and prior knowledge.

Question is stereotype memory errors? Can they occur a long time after the original event?

Do they reflect people's semantic memory and can they be induced by training or all of the above?

So again, that one's number four. That's all of the above.

Again, I'm doing the thing that you think I wouldn't do, which is to have the same answer.

And then this third one is what's the best way to remember new information for test weeks or months later?

Using imagery. Writing it all down at the time. Thinking about its distinctiveness or testing your memory.

Who wants to say, well. He wants to say to.

Who wants to say three? People in the poll everywhere saying the same thing as each other.

Who wants to say for? For the good work.

Thank you. That's a lot. So, um. See you on the discussion board or in office hours.