 Hello, everyone. Can you hear me? My name is Efsuna Naatsch. Currently I'm a PhD candidate. Actually, I'm at the end of my PhD. I'm working on visual attention and memory currently. So that's what I'm going to talk about, the relationship between visual attention and memory today to you. So think about yourself in PubQuiz. So they brought a paper to you and it is asking the capital cities of some countries. And one of them is Australia. You want beer at the same time and you want to remember the capital city of Australia. You cannot remember because you want to call for the waitress. And the beer blocks your memory retrieval or how you access to your long-term memory. Because you keep this information in your mind. So I'm going to talk about this today. Why we forget this information? But I am particularly interested in how it happens in our visual world. So that's why I'm going to talk about visual attention. So my question is do secondary working memory tasks play a role in long-term memory retrieval? What is working memory? Working memory is a cognitive system where we temporarily keep information. For example, I ask you calculate 15 times 4 and now you think about it. This is your active memory. And long-term memory is a storage to keep information over a long period of time, which can be your telephone number, the birthday of your girlfriend, boyfriend or the day you met with her or him. But I'm interested in this visually. Yes, so how we do it in visual tasks is this. So in visual working memory tasks, we ask participants to keep the location of these dots in their mind, for example. And then after a period of time, we give another window and ask them if the location of this dot matches with the locations that they are keeping in their mind. And then they should confirm if it matches or it doesn't match. And in long-term memory, in visual studies, I use especially a visual context memory task where we ask participants to find the target location T among L-distractor items. I'm going to talk about this more. So, do secondary working memory tasks play a role in long-term memory retrieval? What do I say when I say retrieval? So there are two phases when we learn or when we keep information. In learning phase, we encode the information. So it comes through our sensors system and then we encode this information in our environment and then we this information goes to our long-term memory and when we retrieve information, we actually successfully access our learned information. For example, I want to remember my phone number and then I retrieve this memory. But how does visual search work? I want to talk about it. Extracting statistical regularities is part of our, from a visual scene, is part of our visual system. And we are living in a rich environment. Lots of information coming or flowing through our system, but we cannot handle with all of this visual information. Therefore, our visual system is selective. And context information is important because this context information guides our attention to the relevant information for us. And it helps us to select the relevant information. For example, you are walking in a street or you want to find your house, basically. Every day you do this automatically, right? But for the first time, when you move to a new place, you have to learn all the information and that creates all the context for you. And this context information also helps us to recognize the relevant information, for example, the location of our house and to control our action. For example, this is your bedroom. Think about that. This is the door you entered to your room and you know what is where. Your bed is here, your television is here, your lamp, your sofa is here. Even if you close your eyes, close your eyes, you remember the context of your room, right? So this is... We remember this because the relationship between these items or the association between these items creates a map or creates context in our mind. Which we call visual context. But in lab environments, how we measure it is a little bit different. So we give participants a task like this, I mentioned to you, and we ask them to find the target T among all this tractor items as fast as possible. Again, the relationship between these items creates a context or a map in your mind. That we call contextual queuing. So with contextual queuing task, the task that I have just showed you, with this task, we investigate guidance of visual search by long-term memory and we record reaction times, which means whenever you find the target location, you press either mouse button or keyboard button, and that's your reaction time that we record. And we give participants two different display types or two different contexts. One, the context that you're familiar with, which is a repeated context, it can be your room context and another non-repeated context. It's a random context. It can be a new context. For example, I change the context of your room all the time. And when you enter your room, you have to learn where the items are again and again over the time. So it's very time-consuming for the brain. And the explanation is that the associative learning or the learning of the relationship between items guides our attention towards the target location. This is our lab environments. Participants come and sit in front of a computer and do the visual search tasks. So, we particularly in our task ask subjects to find rotated T among L-distractor items, while half of the displays are repeated and the other half is random. Let's do the task together, shall we? So raise your hand when you find the target T. Okay? You haven't found yet. Come on. Now fixate to the fixation cross and find the T as fast as possible. Awesome. Now fixate again and find the T again. Awesome reaction times. Now this is a repeated context. Your reaction times are faster. Find the T again. So this is a non-repeated new context for you, right? You haven't seen this before. This is, so, we record reaction times of this repeated and non-repeated displays and we subtract the reaction times of repeated displays from non-repeated displays and find contextual queuing effect. Basically, this is our reaction times. This is time. And the reaction times of repeated display is getting faster when compared to non-repeated display and this difference gives us the contextual queuing effect or how we learn context. So scientifically, this is how we measure it. Two secondary working memory tasks play a role in long-term memory retrieval. That's particularly what I'm interested. So for this, I combined a visual working memory task with a contextual queuing task or long-term contextual queuing task. So with this task, with this task. Let's do this experiment again. Two, six. Sure. So keep these two digits in your mind. Remember the locations of these dots, okay? And find the T now. Awesome. Now fixate. So does this location match with the locations that you have seen before? Correct. So how about these digits? Have you heard about these digits before? So this is how I combine a working memory task with a long-term context memory task. Okay? So I have done two experiments. I divided my experiments into two parts as learning and retrieval. So in experiment one, I applied this working memory task only in the learning phase, not in the retrieval phase. And in the second experiment, I applied this working memory task only in the retrieval phase where you recall that memory. Okay? And what we found in our lab, long-term context memory retrieval is influenced by working memory task. So it means that in order to retrieve information from my long-term memory, if I occupy that working memory space, let's say, I'm no longer able to do visual search or my visual attention decays. So I have a model like this. So we can not directly retrieve information to do visual search. We use working memory as a workspace to do that. Whenever I block this, you're no longer able to do search. So let's apply this to our first experiment. You want to retrieve the name of Australia's capital city. It doesn't come to your mind because you want to drink beer and you cannot express what is the capital city of Australia, even if you know it by heart. So this is the model. When you experience such a thing next time, you should maybe remember this. Thank you very much. If you're interested in more in this study, our study is published in 2013 in Journal of Vision. You can always ask me questions as well right now or later. Thank you very much for your attention and thank you very much for the 15 times four organizers for everything, actually. Thank you very much, Efsand, for the delightful talk. Your question, please. Yeah. Thanks a lot for a great talk. I have just a quick question that when you mention that if we block the working memory, according to your working memory is for the short term, right? It's part of the short term memory. It's a part of the short term memory. So as soon as you block the working memory, we cannot even retrieve the long term information which we retrieved with our closed eyes. So that was the conclusion. That's what we found, at least, in our experiments. And how did you find that out? Means what kind of, like in your studies, which I thought that they are happening in the room? Exactly. I mean, this is how psychophysical studies work, actually. We see a subject in front of a computer and give a task. And I showed you one trial. We give subjects usually 768 trials or sometimes even more. So they have to sit in front of a computer quite a long time. And after that, of course, we record all the reaction times and do all these calculations, basically. But how do you put the distraction during the study? Do you give them a beer or something like this in the middle of your study? No, actually, it's all visual. I'm interested in visual tasks. Actually, it can be an interesting task maybe in the future because we have the possibility of recording reaction times in virtual reality, for example. People are actually trying to do these studies in virtual reality right now. It can be an interesting study. I can be a volunteer and you give me a beer. Sure. Thanks a lot. Yeah, of course. Any other questions? Yes, please. And if a subject has to pass around 700 experiments, how do you account for the trials? Yeah, trial. How do you account for the fact that he will just be tired at the end of the experiment? He will be tired. Yes, this is a confounding factor, actually. We are aware of it. For that, actually, we give subjects, for example, this experiment takes two and a half hours. And it's really difficult for a person to sit in front of a computer and do this task for two and a half hours. That's why I have to give some breaks after every hour or like every 45 minutes and actually give them some chocolates. It works. I only talked about two studies here, but I have done seven of them. So I collected data from all in all 136 people. And for this particular two studies, it's 17 for each of them. So for psychophysical studies, it's usually a normal amount of people. And what builders? For example, drivers are known to smoke while talking on the phone while driving. Yeah, this is a very good question. Actually, this can be a real-world example. But there are people doing these studies in ergonomics department, actually. But not particularly this context study. So that would be interesting to apply, actually. That's one of my plans, maybe, in the future. Yeah, very good question. Hi, thanks. You talked about how working memory can make long-term memory retrieval worse. Is there any way in which you can improve long-term memory? Excellent, yes. That's my last paper. When you train participants from the beginning of the experiment until the end for two hours, two and a half hours, they actually improve in their retrieval. Yes, it is possible. But with training only. Okay, and last question, please. Does it mean that they can work simultaneously, that they have something in mind, or like in their working memory, and then they retrieve, or something? Yes, they actually... When you train participants, I cannot say they don't use their working memory. They, of course, use, but they are not so much dependent on that when you train them. But this is a really long training. Not like dividing the experiment in two phases. You give like 780 trials or something just for this experiment. So it is possible. All right. Please feel free to ask for the questions that I have sent or to mark as you in the brain. Thank you very much.