How We Learn Versus How We Think We Learn

this program is presented by university of california television like what you learn visit our website or follow us on Facebook and Twitter to keep up with the latest UC TV programs funding for this program was provided by the UCLA office of instructional development the thank you Scott for that generous introduction I'm delighted to have the opportunity to give the hundred and twentieth UCLA Research lecture UCLA is the youngest and in many ways the most vibrant of the world's great universities and as I watched people come in here I see so many of you who have

contributed to the rise of this university including Chuck Young who led this as Chancellor for so many years more than thirty years ago I was invited to be the after-dinner speaker on human memory for the Harvard Ratcliffe Club of Southern California it's a nice event at a resort in Santa Barbara so Lisbon's I'd and I decided to be fun to have our two young sons join us so before the banquet we went to see what the room was like at one point Elizabeth I heard the microphone turn on and our older son Olin about eight

years old then was at the podium doing an imitation of his father and he said ladies and gentlemen the memory is in the brain the brain is in the head any questions I later went on for about an hour saying more or less the same thing actually the truth is we have learned a lot as a consequence of behavioral research for search in the human brain that have fleshed out the picture of the kind of functional architecture of how humans learn and remember what's very new though since that time is research on metacognition and meta

memory what do people believe and think about how their memory works what sort of judgments do they make about how to learn whether they've learned and so on that's a burgeoning field that started about 30 years ago and that has to do with the how we think we learn now before I get more on the story I just want some to do some thank-yous quickly funding over the years students postdoctoral fellows visiting scholars who have enriched my life at Michigan first but then so many at UCLA mentors colleagues have been incredibly important particularly my UCLA

colleagues and among those colleagues especially my wife and colleague Elizabeth York so the human memory is characterized by what I refer to as a remarkable symbiosis of forgetting learning and remembering so forgetting in contrast to what we might tend to think where we might think that learning is building up something in memory forgetting is losing some of what you built up actually forgetting enables learning and it focuses remembering remembering using our memories creates learning as I'll talk about in a minute one of the keys to effective learning is to retrieve information which not only reveals

that it's in your memory but makes it much more accessible in the future and it produces forgetting as we retrieve some things we tend to forget things that are in competition with that and learning a course baguettes remembering it though contributes to forgetting but it also enables new learning it's a remarkable system but equally remarkable is it seems poorly understood by its users that is all of us we seem to carry around a kind of flawed mental model of how the system works how we learn and remember our judgments whether we learned and will remember

are unreliable we're subject to illusions of comprehension students are familiar with that where they think they're very well prepared before some exam we manage the decisions we make about managing our learning are far from optimal students for example go into a kind of court stenographer mode taking notes and that suppresses learning rather than creates learning and these misunderstandings are coupled with some counterproductive attitudes and assumptions that are prevalent in the society I'll turn to those at the end what's very surprising and keeps me fascinated about all this is we have a lifetime of learning starting

early on and why wouldn't we just by the trials and errors of everyday living learning learn how the system works but we don't seem to learn I could spend the rest of this talk showing these sort of slides where what's plotted here in some different experiments is the effect of something or other on actual learning in some case different experiments different paths there's the effect in I've got one graph wrong here but in any case showing actual learning and then in various ways having people judge whether they've learned choose some one way to learn or

not you can see that there's a complete contrast people remarkably misunderstand their own learning so why would this be well one factor is that the whole human learning memory system works in sort of a strange way it's characterized but what Elizabeth and I called important peculiarities their peculiarities because their aspects of human memory that make it very unlike any normal man-made recording device they're important because they have a lot to do with our using of the system for one thing a remarkable capacity for storing information that's virtually infinite is coupled as we're all well aware

with a highly fallible retrieval process enormous amount can and is stored but we're all experienced that only some of its available at any one time and what's available is determined heavily by just current environmental cues interpersonal cues emotional body state use what this does do is help us make most recallable what is most needed in our current state retrieving information from memory is far different than simply reading that information out it alters a subsequent state of the system every time you recall something it makes that information more recallable in the future things in competition with

it less recall and as I've already alluded to conditions that produce forgetting rather than undoing learning create opportunities for additional learning yesterday in my email I got this message from a brain doctor it said eat this never forget a single thing I can't actually tell you what it is because this is in my spam folder as afraid to open it up but you actually if it worked you would not want to eat this thing because as was pointed out by the great William James and late 1800s forgetting is a crucial in the practical use our

memory is as important as remembering and if we remembered everything if every phone number you ever had every street address you every event were memorable and in competition with each other you would have a difficult time using your memory successfully now some of the positive features of memory were of forgetting from memory where some of the earliest work I did all the way back to being a graduate student and that together with work on learning was celebrated in a Festschrift volume organized by former students and colleagues they managed to find this picture from the an

article in Wired magazine but this was put together and organized by Aaron Benjamin one of my UCLA students who's now prominent he's at University of Illinois as a professor there and when I gave a talk at Illinois similar to this one I both thanked Aaron publicly but I also raised this little question for this audience which was I noticed that Robert a Bjork's in kind of small font here and like Aaron Benjamin's in large fans so I asked him this and for those of you who know Aaron you'll appreciate his remark he said I should

have seen the first version it had his pitcher too so among the things that contribute to our not understanding how to learn how to manage your own learnings is this kind of problem we confront which is that conditions of instruction or practice that make performance improve rapidly often fail to support long term attention transfer whereas conditions of instruction that appear to create difficulties for the learner slowing the rate of apparent learning often optimize long term retention transfer I've come to call the conditions in that latter category desirable difficulties they're desirable because they enhance the very

goal of studying and practice their difficulties because they pose challenges they slow the rate of increase of performance one of those is varying the conditions of learning rather than keeping them constant and predictable another one is distributing or spacing practice sessions rather than massing or blocking those sessions using tests rather than presentations as learning events and providing contextual interference during learning what contextual interference refers to is if in some domain of knowledge or some domain of skills there's interference between possible components of that knowledge or skill then it turns out if you arrange practice conditions

learning conditions to maximize that possible interference you'll slow the rate of improvement but enhance and sometimes greatly enhance long term learning this is a broad kind of category I'm going to talk about one case interleaving rather than blocking so if you're learning several things deciding you're going to play tennis and getting instructed in tennis typically if you go there they'll work on your forehand for a while and then you're back in for a while then your serve this work says you should intermix those things in a random way and people will appear to be making

slower progress but then we'll learn better I always have to emphasize the word at Desirables important some early talks I given this I would have some ageing professors come up and say you know I've made things hard on my students for years and now you tell me it's a good thing it matters how you make it hard they're desirable because responding to them engages very processes that support comprehension learning remembering they become undesirable difficulties if the learner is not equipped so for example one thing we should all do as parents as teachers whatever whenever we're

teaching somebody and we can give some minimal information and have that person generate an answer generate a solution generate a procedure they will remember it far longer than if we just give it to them or present it to them so but they have to be able to succeed at the generation okay so varying the condition learning two quick examples the first is one of my favorite studies actually done in a recreation park in involve two age groups eight year olds and twelve year olds and here's what the basic tasks the researchers gave these kids were

they were in a kneeling position there was a box of sort of miniature beanbags and there was a target a black 4x4 square on the floor at a certain distance and they were to grab one of these and try to land it on the square pretty simple except that each time before they threw a screen gain on stopped their vision and then they could look and see how they did the interesting manipulation is for both groups one group did all our practice at one distance let's focus on the eight-year-olds three-foot distance in their case the

varied group had the same number of practice throws but at a mixture of two feet and four feet never three feet then after the training period I won't show you those results because it's sort of obvious the people throwing that constant distance did better a final test a week or so later was at the criterion distance namely it was at the three foot distance eight-year-olds practice and meaning it was at a distance that that group never practiced so what I'm showing you now is the errors measured from the center of the target in inches for

the two age groups that's the performance for the group that practiced at that very distance and there is a group that never practiced at that distance but always practice the foot shorter a foot longer a mixture so how can this be it would seem like we always want people to practice at the very conditions they're going to be tested in and I'll let you kind of ponder what possible advantage it is but I'll just say skills are kind of motor schemas and when you are having to just distance you exercise parameters of that and that's

worth more than the specific practice when I heard about these results and started to do related research Shaquille O'Neal was on the Lakers and he couldn't make free throws and so I actually started writing a letter to the Lakers saying maybe if you have him practice at a mixture of 14 feet and 16 feet not always 15 feet to do better I didn't send that letter but then it turned I was quoted in Time magazine after talking to reporter with that suggestion and I don't think the Lakers took advantage of it or the Clippers for

that matter with the Andre Jordan now as far as variation even varying the environmental context can have benefits this is a study we did years ago at Michigan we had people come to one there was two very different rooms one was a panel seminar room one was a basement room in the animal laboratories that had a faint smell of old rat cages and so on they we made them as different as we could and a given participant studied the material in one of the two three hours later came back and studied it again either in

the same place or a different place three hours later they came to a big classroom like this and they were tested now not surprisingly when people are put back in the very same place they studied they perform a bit better but half the time for half the participants instead of having them test to them we let them Reese Tutty the material as I mentioned and then tested them later now the people that studied it in two different rooms actually produce better recall this is only one instance where something that produces forgetting here a change of

context then actually enhances later recall this result is interesting because if you are a student on this and other campuses I haven't looked at recently but almost any campus if you pick up some little how-to study guide what's one piece advice find one good place to study in the library whatever do all your studying there it might help you to get you to study but it will not enhance long term Rico now the spacing distributing rather than massing repeated study sessions this is research is called the spacing effect it's one of the most robust of

all effects an experimental research on learning traces back 130 years and what it refers to is that if I going to study something multiple times and I could study it and then decide study it again studied again or I could study it go on and do other activities of some kind come back and study it go away and come back study it this would be the spacing case and then if at a delay I'm tested there's a benefit of spacing your study opportunities notice again incidentally that forgetting happens in these intervals but nonetheless the forgetting

enhances learning there these effects are very general in research with animals humans and skills verbal knowledge and sometimes a very large effect but it's a little bit more complicated picture in that if the retention interval is very short basically right away what we tend to find then is a slight advantage for the massing this corresponds to what the way a lot of us got through school cramming and it actually will produce good performance on that test I used to talk a lot in the freshman dorms about how to study and I've always get asked this

question my mother says I should get my sleep there's always her mother not their father and quite an but I'm sorry to contradict your dear sweet mother but it's midnight you don't know the material you won't make something out of nothing overnight but in any case the problem then is forgetting after mass study is very very rapid and there's many experimental demonstrations of it this is one recent one with mathematics learning basically where kids are cramming doing all the problems here or doing some there Waker we do others and you can see at the shorter

delay there's an advantage for this cramming but then at a four-week test a large advantage for the spacing the second thing there's many of these studies this one is kind of an interesting wrinkle because one of the first to ask people to rate their instruction this study was done by researchers at the time the British postal system was changing to new codes which had to be entered on a special machine and involved training lots of postal workers and they asked Alan Baddeley and and the students working with him to carry out some experiments on how

to conduct the training and one of the things they did was a simple spacing manipulation one group of postal workers got one our practice once a day another got one hour twice a day another got two hours once a day another got two hours twice a day so this is the most massed this is the most space these are in between now it's a very nice replication of spacing effects because hours to learn the keyboard show this nice relationship that this was the most efficient schedule the space the least efficient that was the most mass

the others in between but at the very end they asked the postal workers to rate their satisfaction with their training from one very satisfactory to one very unto five very unsatisfactory and looking at that you can see that people had the most efficient schedule liked at least the least efficient schedule liked it both best now using tests rather than presentations as learning events it's crucial given that testings got a bad word so it's it's such a bad word that when I talked to some audiences I switched to the phrase retrieval practice rather than testing and

retrieving and there are three virtues of tests from a pedagogical standpoint very important virtues one is retrieving information or procedures as I've alluded to is what I call a memory modifier the information you recall becomes more recallable would have been otherwise as a learning event it's more recallable than being presented that information I say this inflatable life vest example this is this a story where I'd rushed to an airplane some years ago and again they were trying to show me how to put on the on the vest and I thought oh how many times I

seen this so I kind of relaxed in my seat I imagined that the plane was down in the water those lights were on there were a few people screaming there were some faint hint of smoke do I know how to do this so where is that life preserver well it's under your seat but like how under your seat I mean I started get panicky like it's a hanging there is it in a velcro package it's whatever so then I imagined that I I got this out and and what do you do I it's obvious to

this to the flight attendant which is front and back but was that going to be obvious to me anyway I assumed I got it on and pulled these things then what do you do after that want to be on a plane with a lot of you I don't think ah anyway you try to pull this cord to inflate and if that doesn't work what then you find this tube and blow it and I assumed I did that and then got to the window and at least at that time you were supposed to blow it up

after you got out the window but as I once said in it talked at the air traffic control center Oklahoma City if there was one place in Airport where a single time you could do it just find it put it on do it go out that'll be worth more than these hundreds of exposures we're not sort of just recording devices I even suggested you might have people to give them a little pin I know how to go out the plane so other and then other information that's in competition there's a whole domain of research on

what's called retrieval induced forgetting that I won't be able to touch on and the basic point is that as we use our memories we alter or shape our memories in important and useful ways the second virtue is testing provides far better feedback to what's been learned than does Rhys tutting if you're a student preparing for a midterm exam and trying to scan the material to see what you still need to study when the material is right there in front of you you will be your ability to judge whether you'd be able to recall it is

very flawed very unactive you're with another student asking each other questions and trying to produce answers you'll get an accurate measure of what you know and understand the third thing this is the most recent line of this kind of research testing it turns out potentiates subsequent study even if you test people before they've had the material which something Elizabeth didn't one of the courses she teaches and they're mostly wrong and everything after that they study more efficiently three important benefits of retrieval practice one example the power of tests versus presentations in this experiment the participants

studied a to be learned passage it actually had to do with information on the Sun or at an information about the lives of sea otters they got one or the other two two different groups are of particular interest one of these just kept studying it over and over again that passage that's this SSS group consecutive 5-minute periods the other group studied at once and then tried four times to recall and notice after they recall there's no feedback of what they got right or what they got wrong so this diagrams it and among other things the

group that repeatedly studied it managed to go over this thing highlighting doing whatever they wanted fourteen point two times the group that read it only once three point four but then again they tried recalling it so they would recall as much as they could for five minutes the sheet was taken away and they were told try again so they did that three times now what happens at five minutes it looks like repeated studying all those exposures led to better recall but when they tested people to week all this turns around in the case where they

read it only once but try to recall it three times is better again on the sort of metacognition side they also asked people at the end of the study phase how much will you remember in one week's time there's the actual results that's what they said so we're we're shaped to think that studying is the best thing to do overall tests potentiate subsequent recall but what about failed tests suppose you produce errors maybe that makes those errors more likely to be remembered and you can seize things like this this was sent to the on the

internet awhile ago it brings up some considerations I hadn't even thought of namely that you may trigger emotional and aggressive behavior and put your career at risk if you produce tears and you don't want people to establish an error history well this has now been looked at at different levels and it actually looks like making errors is a critical component of effective learning I'm going to show you a simple procedure that we came up with but these kind of results are duplicated in more substantial cases in this procedure subjects are learning many pairs but for

half the pair's and they know they're going to be associated a cue like this and they know the response is going to be associated in half the pair's they have to try to predict what the associates going to be and I don't know you might predict big ocean whatever but then you see that the correct thing is mammal and that's what you'll need to recall later when you're given that particular cue where it has been studied lots of these pairs half of them that way half of them they just see it intact and have 13

seconds so in one case they spend eight of the 13 seconds coming up with a different response and I should clarify that we selected these pairs so only so 97% of the time people are wrong they come up with something else so you spend eight seconds coming up with a different thing than five seconds versus studying the right thing the whole time and therefore for different experiments is the results generating the wrong thing and then seeing the correct answer leads to better Rico now this is this is a topic of current research but one of

the interpretations of this is that when I see why wham-o when i see maybe that would be a way to encode that pair when i see whale activates that semantic network and I come up with ocean large maybe even that it's a mammal not a fish that activation appears then to facilitate associating the new response that's as I said this is very active the basic effect has been replicated many times including in this study at at Columbia University or they repeated our procedure but then they asked the participants which helped you remember better and they

said being able to read the pair of the whole time rather than generating error first this is even after the final test on which they've performed better on the pair's where they tried to generate the answer first okay this topic particular interleaving rather than blocking is a new domain and the implications are really pretty striking and we're doing research now trying to see how far is some of these kind of results extend but it all started with two graduate students in kinesiology at the University of Colorado who built this little apparatus and on using this

apparatus the participants had to learn three different movement patterns each pattern they would release the start button pick up this tennis ball knock over three of these hinge barriers in some order and put it there do that as fast as they can they had to learn three different patterns each of one involved knocking down a different sequence of three of those barriers now in one case they got 18 trials in a row on one of those patterns then 18 on the third then 18 18 the second 18 to the third in the other case they

got the same 54 trials 18 on each but which had to be executed on a given trial was determined and shown by those lights in any case if you just look at down as good here they're getting faster and faster if you just look at performance during their blocking is better the difference gets smaller but blocking appears to stay better the whole time that would be consistent with the fact that blocking is everywhere in the real world and that's what people involved in training and so on would tend to see however ten days later when

they tested people they either tested them under random conditions and you can see under random conditions the people who had the random training performing about the same level people that blocked look like they didn't even learn anything yet and when they tested them under block conditions even what little advantage there is favours the random training so again this has major implications how far can it be pushed in a domain of skills it again has the potential for leading you to misinterpret your current performance as learning so we've replicated that sort of experiment using keystroke patterns

that had to be executed in a certain target time the day - there was a criterion test the new feature is that periodically we ask people to predict how well they were going to do on the day to test a few times during the training and then right before the day to test we also ask them you can see here again down as good this just replicates the pattern that Shay and Morgan had with a different task and it replicates the fact that a day later this condition that doesn't look as good the random interleaved

the block that looks better is worse a day later and you're better with interleave but what did people predict basically they predict the opposite people who had the block practice thought they were going to perform better than the people who had the in early practice thought they would so this now one of the things that would be good for you to take away from this lecture is there's a fundamental distinction that has a kind of time-honored and research on human learning and that's the distinction between learning and performance at any one time what we can

observe is performance which may be a product of local conditions and other things what we're forced to infer is learning as it will be manifested later in another context or at a delay and current performance turns out to be very unreliable indicator that makes us vulnerable to choosing poor conditions of practice or learning over better conditions and this just reap lots those Simon and Bjork results the advantage of interleaving has been shown now with many other skills the serves in bad men pitching kayaking using different ATM machines even handwriting children's handwriting here I have to

show you a little example this is from penmanship booklet if you're beyond a certain age you may recognize this this is part of fundamental education way back and those of us of that age had to do many hours of this and also practice cursive which may be disappearing altogether but notice the designers of this probably didn't give much thought seem like obvious how to do it notice the child here's little diagram they have to make one W another W another W a fourth W 9ws then a bunch X's then a bunch of wise and Z's

what sort of practice blocked practice didn't probably incur to them to do differently but it did occur to these researchers to carried an experiment where they did compared this with one that alternated the letters and they found big benefits for the alternation particularly if the final test involved writing in which you alternated letters and you might think about that a bit because you don't have a look I don't have a lot of need in normal writing to make 9 WS in a row I sometimes flame that people of my era are kind of doubly punished

we had to spend some hours doing this and then about thirty years later had to read papers from students who didn't have to do it now it turns out a newer direction is these were all kind of skill learning is in kind of verbal conceptual learning as well there is new indications that interleaving has benefits as well this is study done by a colleague Douglas roar at the University of South Florida the participants here had to learn these formulas for things like a wedge a spheroid half cone so they worked many problems with these with

different dimensions as part of learning those formulas and the manipulation was that they worked the same number of problems but some participants had a mixture of working on the different ones the other ones did all the problems of one on one solid then all the problems on another and so on and a week after the second practice session like this they were tested and notice right away at the end of practice it looks like blocked was better but a week later there's a three to one advantage for in early a really interesting story having to

do with this paper it was rejected by for educational journals without review in a couple cases it's now heavily cited it's been replicated spewed as important once it got accepted by the fifth Journal and you might wonder with all this expenditure of money to try to increase educational outcomes in the public schools here's something that made a simple thing made a three to one difference but I think editors and reviewers of those journals this is maybe plain clinical psychologists too much or something but this is not something those people are doing in their own teaching

it's hard to believe the size of these effects so some of these things just take an open mind so to speak in the sense that they run so much against prevailing practices if their mathematics problems but they're kind of unrelated this shows some kinds of things students could be given and in this experiment they compared mostly interleave practice mostly barque practice then there was a final review and that pierced have been automated a final review and whether at one day or thirty days this case within early practice produces better performance on this surprise test not

too long ago got this email from William enemy who I said M&E didn't I not enemy okay a teacher in England who has done a series of wonderful things but he said after learning about this in particular space interleaving I started trying to put them into practice I did this in a number of ways and then he says first I realized we did interleaving as a standard practice but only during the last six months of the course I wanted to check the impact of this and so I looked at our assessment data we look at

them three times per year and so this led to the graph referred to an email which shows the impact of the interleaving so that's blocking up to that point and then last six months is interleaving this is not a controlled experiment but on the other hand it's done in actual classrooms now optimizing inductive learning this is a crucial kind of learning in which you extract a concept or category from induce it from examples so this might be a matter of trying to from these learning to categorize some sort of malignancy or possible problem another kind

of thing and I show this because we use these materials and experiment about to tell you about is we might develop an image of a painter's style from different paintings by that person you might be traveling in Europe somewhere and be able to say that a certain paintings is by Rubens even though you haven't seen that painting before this is inducing categories and concepts from examples just to give you one a little exercise here suppose I tell you that's a Gentoo penguin that's a Gentoo penguin that's a Gentoo that's a gem - where's the gen

- I can't see anybody so but oh I think I see somebody pointing up there now notice I did not show you that picture you you extracted some commonalities from the others that let you be right there and we in this case a postdoc who's now teaching Williams College Nate Cornell and I decided this has to be one case where blocking masking is advantageous and the argument was that when they're together like this like you had them you can see the features that sort of define the category maybe for example this little white band up

here whereas in ER leaving or spacing makes that very difficult to do you see a Gen 2 here then a lot Jesus then a Reinhard then a - and thinking back here - what's common is more difficult I think I have put an extra slide in here get rid of that so we had people learn the styles of twelve different paintings painters by seeing six examples of each painters paintings half of those six of the artists the participants saw their paintings one after another in a row block practice the other half they saw them interleaved

with the paintings by the other artists and I'm going to give you a feeling for this by just showing you an example this this would be an example of a mass block that's a painting by Luis another one by Luis another one by Luis are you getting Louis's style now other one by Luis another one by Luis they weren't selected to be good artists another one by Luis so an in early block on the other hand you'd see the six individual paintings by given artists intermixed with paintings for the other artists and it would be

like seeing that there's a Pisani there is a Wechsler there's a sloth there's scratch a lot Hawkins mill Raya so anyway they go through those 72 paintings what this would be a sample of what the first 24 would be like and then on the final test they see completely new paintings and they have to say pick who painted that new painting that's the task they haven't seen those before they have to extract the style of a given painter and say who painted it again we were expecting this is one case where blocking massing should help

and this is the results we got researchers are often not as smart as we think we are and this now has been replicated with many learning many other categories butterflies birds novel objects women's voices statistical rules a whole bunch of things now what had made this research have a big impact as something we kind of threw in almost as afterthought after they were tested and on this test had done better with the artists that were interleaved west's in this question which do you think help you learn more master space they could say massed about the

same space now these are the actual results over here and that's what they said now this has been replicated many times and actually dr. Veronica Yin who's here somewhere in our lab has been trying to explore what sort of experience do you have to give people what sort of instruction would lead them to come to realize the benefits of interleaving one of the early things she tried was she ran this same sort of experiment but right before she asked him his question they were told 90% of learners do as good or better with interleaving what

she found was that 80% of people think they're in the other 10% it's really compelling and I won't go into all the interpretations but what it does look like just in general that interleaving highlights a difference in relationships and that may be worth more with respect to the final test than seeing the commonalities why is that judgment so difficult to overcome well one reason among others is blocking provides a sense of fluency that's why I showed you a sample actually whereas interleaving provides a kind of sense of confusion or difficulty the participants actually come to

the experiment thinking that blockings better why would they they've had blocking their whole educational life every teacher they've had has tried to be well organized in the course outline or syllabus by blocking things in a clear way recent evidence for this is in this experiment down at Kent State University they're learning as families of birds and here they gave the participants control over it so these were this just a sample of the kind of pictures they got of the birds here would be 3 different jays over trials but they would get it you've just studied

a Jake click on the bird family you'd like to study next so they could ask for another Jay a Finch a sparrow whatever overwhelmingly they exhaust the Jay category basically they give themselves block practice when it's under their control maybe older adults maybe there's something about blocking older adults with these so we looked at this in a study at UCLA in one case the the people studied the same artists the same picture over and over again another case for different but in any case then later the older adults there was a advantage for spacing with

the same picture being shown over again an advantage for the induction test this just replicates the effect I've already showed you they'd actually had mixed feelings about seeing the same picture over and over again here but again when if the task was the induction induction task they too thought massing was far better then it's a question well what about kids is a critical role in kids coming to organize and learn about the rule this kind of categorization they're amazing at it their parents tell them something like this little thing is a boat or ship that

is the plastic thing in their bathtubs about this thing's a boat down at Harvard that's a boat and somehow kids extract the commonalities and come up with this concept so I research in our developmental area here hey leave LOC and Catherine's and offer had them learn categories of novel objects so they would be told like the thing on the left is a in the middle of a blicket and alaska's of dax so in one case they would say look at this and they'd be handed this thing for ten seconds look at the then they're handed

another one look at the wall then another one look at the wall the other case was just like this except each time between seen logs they played with a novel object this was to introduce spacing then finally they say can you hand me the walk one of the won something then on the name of cat one was this object they played with and one was novel so this just recaptures the design I'll just quickly show you the results basically whether again it was a memory task or an induction task you are better to have this

spaced learning which you might have thought would produce forgetting so what the students already know and not know about how to study we decided to take a look by doing a little survey of 472 intro psych students questions like this how do you decide what to study next whatever is 59% what Evers do soon as most overdue for prasit whatever I haven't studied for the longest time that would be what spacing that would be a good thing whatever I find interesting doesn't matter 22% whatever I feel I'm doing worst in and then 11% I planned

my study scheduled ahead of time and I study whatever I've scheduled there are actually liars another question if you quiz yourself while you study one way or another why do you do it here they show some a bit of an insight they don't know that it's a learning activity to produce things that retrieval will be a learning activity they do know that tests can help you identify what you've learned and not learn these people very few liked quizzing and these 9 percent that don't quiz themselves are probably in trouble one way or another do usually

return to the course material this would be a very good thing to do in terms of long term spacing effects 86% know when you study tip of the read an article other source material more than once yes I reread whole chapters articles 60% I reread sections I underlined or highlighted or mark not the original one not usually would you say and this is interesting would you say that you study the way you do because somebody taught you to do that this is on two accounts first of all the 20 we don't know what the 20%

who answered yes what were they taught probably to study in one place and focus on one thing at a time and so on but 86% now they just base to delve some habits somehow and that's what they do we worry at the university whether students are prepared in certain math topics English proficiency we don't actually tend to worry whether they're equipped in terms of learning skills to cope with these four or five more years of intensive learning now what are the implications of these desirable difficulties for how we would organize our courses as teachers and

we could spend a long time there thinking how we might change our courses this way how we might incorporate they would be quite different courses but what are the implications for the evaluation students evaluations drugs students expect to be taught the way they've been taught what's going to be the reactions to the immediate consequences of desirable difficulties which is sort of worst performance I've sometimes said that if you gave me a half for all these years of research and teaching you get me a new course to teach and you tell me do everything you know

how to do so six months after the course has ended students will have a maximal memory for the key concepts I feel like I know how to do that if you told me do everything you have learned how to do to get the highest possible student evaluations I feel like I know how to do that unfortunately they would be very different courses concluding comments there's some general attitudes and assumptions that make it difficult to optimize learning one of those is errors and mistakes need to be viewed as opportunity for learning you should be willing to

make them where as often they're reviewed to just reflect inadequacies of the learner the teacher or maybe both differences in performance are attributed to an eight differences ability there's an under appreciation of the power of training practice and experience some of you may know there's an important domain of work on people having either a fixed or growth mindset and all the benefits of having a growth mindset rather than a fixed mindset there's also a prevailing idea that efficient learning is easy learning if somebody will just present something to me in a way that fits with

my style of learning I won't really have to engage in much effort it will just sort of happen we did a you four of us of research on this very popular styles of learning idea and could find no support for that notion basically individual differences matter and they matter in a crucian way because new learning builds on depends on old learning our family cultural histories affect greatly motivation to learn degree to which learnings valued aspirations and so on one example this little experiment we carried out years ago in terms of individual differences we got wondering

about this which is optimal doing the readings then go the lecture or lecture then readings what our students told students are told that and we did a preliminary survey where we asked them which is more effective two-thirds say do the text in the lecture what do you do two-thirds do the lecture than the text which is more difficult people say doing the text in the lecture so we carried on experiment won't get in the detail it was had to do with material on one one thing had to do with brain aphasias and other things had

to do with real estate law new material to participants and they either read the text and saw video or they saw video than read the text and what we found overall sorry is that there was no significant difference but whether your language first language was English matter which way do you think it mattered if English was not your first language then reading the text better was better first is better otherwise if it was going to the lecture first was better so the bottom line is individual differences matter optimizing learning teaching rests on what we all

share a unique functional architectures learners a remarkable capacity to learn and one more little comment the importance of avoiding egocentrism in social communication what this refers to is we are all vulnerable to overestimating how much what we are saying our students friends and colleagues actually understand and this leads me to conclude with one of my all-time favorite experiments carried out as a doctoral dissertation at Stanford University and what Newton did is the participants were either tappers or listeners if you were a tapper you got a sheet of with 30 familiar melodies pick any one you

want and tap it out so it might be if if you were if you were a listener you didn't have a sheet of paper but you then tried to identify what that tune was and then the cappers predicted the likelihood the listeners would succeed and overall of the tappers just almost exactly half the time they thought the listener would get it here's what the Lister is actually got so I try to remind myself I refer to this study as sort of a a parable of teaching as teachers were so incidentally what did I tap out

Wow that's the highest response rate that's correct it's usually zero well I must have done a little better but in any case when I tapped it out what was happening in my head jingle bells jingle bells music lyrics what were you getting this is analogous to what happens when we're teaching often we're going the music lyrics in our head how wonderful this is and what the students are getting is blah blah blah blah blah and in 1962 a giant of developmental psychologists made basically this point every beginning instructor discovers sooner or later those first lectures

in comprehensible is talking to himself so to say mindful only of his point of view reveals only graduating with difficult it's not easy to place oneself in the shoes of students who do not yet know about the subject matter of the course I can only thank you and hope I didn't illustrate that excessively thank you very much you video copies of this program are available for purchase from the UCLA instructional media library call toll-free 1-800