Welcome everybody. Hi. Um So, um thank you very much for having me and thank you very much for joining this session. Um We will talk about um unconscious bias and for this audience, I think it's not going to be news.So, you know what it is, I imagine you have heard it in many contexts. Um But I will actually try to explain it um from a sort of a cognitive bias perspective um that also occurs in other conditions. Um When you talk about unconscious bias or even uh also conscious bias, sometimes people uh refuse or they don't want to um accept that this exists or that, that they might have it um themselves. Um So I'm gonna um just start my timer so I can keep an eye on the time. Um So we will um start by talking about some visual illusions. Um And you see that here on this uh chapter uh slide already, um many of you might be familiar with these um when the circle is surrounded by smaller circles, uh it looks larger. Um But actually, if you um would measure it, you would see that this is actually the same size as this one. So this is a visual illusion. It's one of the more famous ones. Uh So, but why should I be talking about visual illusions?

I study visualization um as my job. So as a part of that I do actually research on visualization including um vision inspired visualization concepts and some of that is also to do with um actually visual illusions. Uh Currently, I'm um a professor of human computer Interaction and extended reality and my background is in Geospatial Sciences. Um And one interesting little bit about me is that I'm actually the first woman to earn a phd in a field called photogrammetry, which is um obviously a small field, but um still very interesting. I grew up in Turkey and I went to Finland to do a phd and it turns out that I'm like the first woman in Finland, which is sort of a as, as a sidetrack, which is also kind of ironic because um if you look at the um OECD statistics, you see that Finland is ranked as the best place for women to work because there's a lot of um benefits and um and in terms of um like child support and maternity leave and, and such, um and um Turkey is actually ranking quite low just about um Japan and South Korea, but just below Switzerland, which is where I am currently.

Um So this is um sort of for me, it was kind of an ironic title to have. But OK, let's dive into the visual illusions because I studied visualizations. I ended up noticing some visual illusions. An interesting one is a three dimensional um illusion. This is called H mask illusion. Um And you see you are looking at the mask from front first and then when it rotates, you are looking at it from the back, I'll play it once again. So when you are looking at the mask from the back, this is it, you see the nose comes out as a convex shape, right? So this is because your mind is like a statistical machine, it says to you that OK, I've seen noses 1000 times, maybe a million times there are always convex. Lets let me correct this for you. But this is not, of course, this does not happen with physical masks. You might have also seen physical masks. And when you look from the back, you can actually see the noses in the concave. It also happens because the shadows are in a funny place. The lighting is slightly uh from the below right here. And this phenomenon also occurs with terrains. As I said, my background is in geospatial. So I have studied maps and what is called shaded relief maps. We use shadows to express the terrain heights and variation. Um And here um it also occurs in subtle light images.

This is an example of a subtle light image, the satellites fly around nine or 10 a.m. They are called sun synchronous uh satellites. And when they capture the images of the northern hemisphere due to the tilt of the earth, uh the northern slopes of the mountains are in shade because the sun is coming from below, right? And of course, when you place the map up, um most people, a majority of people will see this what's marked with yellow dots as a ridge, but it's actually a valley. How do we know it's a valley because you see this, there's a river flowing on it actually. And there's some snow on what appears to be valley floors here. And we have studied how prevalent this is how many people can bypass this illusion. And it is very prevalent. If you rotate the image, the spatial relations are kinda restored for majority. Now the snow appears on the hills and the river flows in a valley. Although this image is still a little bit odd because everything is not exactly in the right place, but this happens quite strongly in no sphere images. And a lot of people get this illusion. So to be to understand exactly how that occurs and how strong this illusion is.

We would take shaded relief maps and we would because here we can control the light source as opposed to the satellites, which is, you know, basically very expensive to um have them, have them fly for us in a different um uh sort of time of the day or a different drought, a different orbit.

So in this case, we took uh basically software that can render terrains using shadow shading our shadows. And we would uh systematically rotate the illumination angle when you do that. It's actually a non effective per perceptual psychology. It's called light from above prior or also known as overhead illumination bias. So it's in essence, it's a bias. What's interesting is we ask people, is this ABC a value or a ridge? And they could mark clearly a valley and clearly a rich, you see there's a certain level of confidence here. So you can say it's a valley, but I'm not like clearly, I'm not like 100% sure. You can say I'm unsure and this gives us some information about the strength of the illusion. So we have shown people shaded relief maps, but we have also shown them satellite images in black and white and in color. And you see how many people can bypass the illusion here. It's like 2% or so. And here are some more people and here similarly more although um surprisingly less people in color, but that's not the talk about this. So here, what what you're seeing is that maybe with the terrain features, snow and river or so more people can bypass the illusion.

But if you look at the confidence, how confident people are, whether they can tell is this a value or is this a ridge, the pattern reverses. So when the illusion is really strong and you have only 2% of the people who bypass the illusion, they are more confident because the illusion is so strong that they think that yes, I'm sure that it was a value. Now, you're gonna say, you know, what's that to do with the cognitive biases? Um There's a, maybe a sidetrack here, I will also take you and then I will go into the um social cultural biases. Um So here, we also looked at the differences between men and women in terms of accuracy. And here it includes the correct images, southern hemisphere images. That's why you see the higher accuracy. So between men and women participants, female and male participants, we have actually a very tiny difference um between the expert and non expert groups. So people with some experience, people with no experience, people with some experience do slightly better in both groups.

But there isn't actually a difference in terms of like men and women. But if you look at the confidence, you first see men are actually more confident than women. And if you split it by expertise, you see that under experienced men are most confident. In fact, men who have some experience, their confidence level is quite the same as women with some experience. So here in terms of um what happens with women, they, their experience, their confidence is not affected so much by the experience level, there's a tiny difference which is statistically insignificant in this case. Uh Of course, if we measured many more people, maybe we would see more of it. But here it's really pronounced. And if you take average male, an average female, you get like huge confidence gap based on inexperienced men. I don't know how to socialize men, but somehow they are overconfident. You might hear this in some talks that women are under confident and we must push women so they can they can be more confident. But I suspect that sometimes that it is actually women's confidence is just right where it should be in, at least in this study, it was uh men are overconfident and perhaps they need to compensate as you see with some education and or experience, they actually can compensate and come closer to their levels of accuracy.

All right. So this happens, this happens that our minds automatically categorize stuff, collect statistics at the kind of operating system level. So you don't really like with the visual illusions, you don't really rationally decide what it is. It just happens your mind over lifetime over evolutionary uh uh little e evolutionary and it starts building up a sort of sort of a cognitive experience and cognition then overrides the perception, your perception says this is what it should be but your cognition says, no, no, this is wrong.

I actually already have this information. I don't need to interpret the scene I'm gonna bring it from the brain. So this happens also with people uh for, for um psycho in psychology literature when I first started to notice the link between visual illusions and cognitive biases.

Um I didn't realize that this was already uh actually discovered in psychology and even got a Nobel Prize for these people. Uh Kan. Um And in this book, Thinking Fast and Slow, some of you might be familiar, they talk about system system one and system two, system one is this, you know, intuitive decision making. You look, you take one look and you know, and systems is rational where you put the together, you do pro con list and stuff like this and they talk about a hollow effect about people. So you look at someone who looks right for the job or who looks right or who looks somehow fancy or I don't know, you make some associations, unconscious associations. This happens with gender, race, age, ethnicity name. You know, we assume uh people to have more authority or intellectual competence based on these things, which cannot be far from the truth. But we, we, we learn these by seeing these people in these roles over the course of their lives. And then we start associating that like tall man, um uh tall white men. Um they tend to be in the leadership positions.

We have seen that over the years and your mind kind of picks up on that and learns uh in this one book, Gladwell, uh Malcolm Gladwell, you might also know that has taken a uh implicit bias test and he himself is a black Jamaican black, I think person and he kept, he kept marking black faces as potentially more criminal.

And he's like, OK, am I biased against myself? Yes, that happens. Women also discriminate against women unconsciously and that happens that you internalize something because that's about seeing that people in these roles um heavily featured. And there's a lot of evidence for this.

I'm gonna rush through a number of examples if you give Children, um just tell them to draw a scientist. The chances that they draw a woman are very low. Uh So in 1983 of 5000, only 28 girls and none of the boys have drawn a woman. They repeated this study recently. And now it comes close to 8 to 10% which might reflect actually the number of times people have seen women in these roles uh potentially in their lifetimes. And if you take that at the name level, you have examples. This is one example, people do this blinded trials. And in this case, it's like this group of essays that are written by a number of students. And one group of teachers receive these essays written uh with um Turkish names and German names. And then another group of teacher has the same essays, but the names are switched and Turkish name apparently makes you lose about 10 points on average or so. And this is a sort of a controversial study, some people try to replicate but there's others that actually see this effect. So similarly, you can see this in likability. Some of you may have also known as the famous study Heidi Howard study here. It's actually a real entrepreneur, uh Heidi Rosenberg and they took his, her CV and they just changed the name to Howard. And then they had like fairly progressive group of people. I believe it was Yale under graduates.

They asked them to uh rate uh likability and they liked Howard better because Howard was a leader and it was too much for, for herself. I think this, this beep might be warning me uh or maybe it was another beep. I'm sorry. Yeah, I'm going back and I'll rush through a couple of more examples and then I'll wrap it up. So, um the name again. And so if you have like John and Jennifer's study, um so you have, first of all, you have fewer running uh fewer women run big companies than men named John. And then also some, again, that was a Yale study. I believe the other one. Rosanne's study might be Harvard. I maybe mixing that. I'm sorry for that. But people have actually like exactly identical qualifications, both jury members, men and women. They actually judged John to be more competent. They were more likely to hire him and they were more likely to interview him and they actually wanted to pay him more. Same CV. Uh So in the same for teaching evaluations, uh similarly in uh so there's like three examples here from popular press, uh that was from Twitter. Um Like people do blinded. Yeah.

But you, you go for like an online teaching, you, you see everything the same, just change the name. You don't actually meet the teachers virtual, you interact with them um without actually having met them and you always get like a number of negative views. People even discuss whether there should be some compensation for that because it seems to be so well established and it's not like it's like people really rate a woman delivering exactly the same thing less competent and they, they use words like brilliant for men. It happens in, in grants. So to receive research grants, a woman's CV should be like five times more productive. And if you take it like this happens in Sweden where you have like a lot more egalitarianism and things like that. You know, there's a lot of evidence of that as well and invited talks. We have less invited talks, somebody tried to correct um for um like whether it was because they were not interested or they said no or something and they say that it, it really isn't about the pool composition. So like some sometimes you hear people saying that yeah, but we have only so many women in our field, right? So this is one of the things um that they corrected for and it pops up recommendation letters.

So when I'm a professor, when we write recommendation letters, apparently we write shorter for women uh for our female advises and apparently we use different words. Uh So one has to watch out and one of the most entertaining ones, it apparently happens also with space. So at mit they have looked into the spaces and senior professors, if they were female, they had on average, the same space as the junior faculty and that's like significantly less space. Um It wasn't, but somehow it wasn't about so. So it's like this is unconscious bias in my head. I can't imagine a group of people sitting there and saying that, yeah, like women does, they don't need as much room. So like, you know, I'm pretty sure this happens somehow with an association uh thinking that, you know, who is more deserving or who's more entitled. And now there's another very depressive study that was across 75 countries. It said that, you know, um uh people like across the board, people think that women shouldn't do some professions, they don't fit to some professions and some of them are like intellectual professions, they are not even like physical professions where you can justify about muscular differences.

You know, it is it justified at all intellectually, not at all. So, if you look at the academic achievement, females are actually completely surpassing males in school success everywhere in the world. Even when there is no gender equality uh in a school system where it's actually built for boys. Originally, women are completely going over and taking over and also in hard subjects. It's not like, yeah, no woman cannot do math. Don't buy that. It's bullshit. Sorry for the bad wording here. But it frustrates me. So now it's actually 56% of the university students that was, that might actually continue. This is already a couple of years ago. That was 2015. The statistics. Uh So what can we do? Obviously, if we are in a position to hire, we should reflect what we are doing. Uh If we are giving recommendation letters or advising, we should um just try to activate the system too. And um and if you are in the job market, I thought that you could maybe change your name to John. And if that doesn't work, just don't be discouraged. If you get refusals or if you don't get shortlisted quickly, just keep going because it may not be to do with your qualifications.

Of course, you know, don't blame everything on gender but just take it into account that it's possible that you are being overlooked or underestimated for the wrong reasons. Um Maybe one can do blind reviews institutionally or quotas, but this creates um a lot of discussion in both cases, I will leave that um because the time is running. So if there's some questions, I'm gonna try to talk to you about this. So hopefully we can remove the obstacles so we can see talent for what it is. Thank you very much for your attention. I'll stop the sharing so I can actually see you guys. All right, stop sharing. Thank you very much. All right. So I think, yeah, thank you. Thank you. Thank you. I think I'm actually maybe using all, all my time. Mm Thank you. So I think there was at least one from Rebecca um prejudice control. Yeah, that's so you want a specific trick? Mm Yeah, but I don't have a specific trick. So what I try to do is that to go from um some sort of like intuitive judgment to a more rational one as, as much as I can and ask myself, am I really being fair? Am I really being, am I really doing this? I use different words for my uh female and male students? I also catch that in myself and I tried to correct for it.

But uh there's, I guess some exercise at becoming aware just attending such events and so on so forth is probably the only way to create a cultural shift. All right, I think you guys have to shift to the others. Thank you very much. Everyone for the kind words. Yeah. And the impo imposter syndrome. Yeah, it's true. Um, it's not only women but it's more common among women. Um, so I guess you just have to sort of, um, tell yourself it's a psychological mechanism. You just have to tell yourself that the, um, that yes, there are, there are people who are a little bit better than you. Um, but it doesn't mean that people who are in these positions, who, who are occupying these positions are always better than you. You get disillusioned as you go up in the letter, that people whom you thought you put them on a pedestal, they are not, they are not so much different or better. In fact, you might be doing much more and much deeper. That's, um, there's, uh, this, uh, bias or imposter syndrome tends to, um, yeah, they tend to make you work harder. Uh So in many ways, I think that it's often not justified and it, it's worth reflecting. Um, and trying to just keep going. The, yeah, chat is super interesting. II I, so for the organizers, the next session is going on, right? Is, is this, is there any room for doing the discussion a little bit longer or? Yes, exactly. So what Daniel said is very correct.

Um I would say, um, in general, you pick up on the other's image of you. Um, I mean, I have, I have grown up in Turkey and I lived abroad for half of my life and I, my nationality and my gender is always a topic, always a topic. It brings up a lot of questions and I mean, when I lived home, not so much perhaps, but at least the next part. But this, you, you also subconsciously or consciously pick up how people see you. Uh they, they bring me bills to pay because they think I should be the secretary or, or they actively think that I should be in a, in a helping role. And this happens over and over again, they switch my name and put the guy's name. If I'm collaborating with the white man, the chances are somebody will actively switch and put the name their name first. It's very, it's very interesting. Um Gender neutral names. Yeah, thank you very much. So. Yeah, I, I haven't seen gender neutral names. I have seen a lot of blind cities that happened in, in France for the interviews that happened um in the US in a number of time. But the gender neutral names are interesting whether it gives you an advantage. Um That could be interesting and could be easily doable if you are interested, Morgan. Um we can run a study together. Yes, Rebecca. The that was repeated. Um And I think the rates went from 0% to uh around 8 to 10% today.

Uh That was, I think I found another study from um two years ago or so. So it, it has gotten better and I think that it's, it's actually, it shows it's very interesting, it shows that this role model approach actually works. So you um you have more women in these positions. I mean, at the moment, for example, University of Zurich, the number of female professors is around 13%. If the kids now draw 13% or 1010, 8 to 10% draw that way. The chances are they have um they have experienced it and, and now they are associating it also more with women. So the idea of the quotas, as I said, it leaves a bad taste, but it's, it's, it's important because it's um I mean, even if it is a little bit questionable, it is important to create these role models and put women in, you know, as pilots, as judges, as professors and, and make them visible.

And I know that it creates a lot of weird, uncomfortable feelings both like I don't want to be selected because I'm a woman. It feels like an insult. And I think that also creates uncomfortable feelings among men who feels that. Yeah, maybe a woman is, is selected. Um even if they are not qualified, so we should ensure that the qualification is that they just, what you need to do is to prevent discrimination. And then once you have these people in these positions highlight and make them visible um whether I confronted people with their bias.

Um Yes. Um to some degree. Yeah. Thank you, Rebecca. Uh This is a very sort of difficult, culturally, difficult subject politically. I think we are on the right track, more or less in most parts of the world. But it is indeed uncomfortable culturally because they feel discriminated or they, they feel like some of the privileges are being lost or I i it's, it's a hard conversation. Um Have I confronted others? Yes, I did. I tried to be not, like I tried to be not super um like combative. I try to like unconscious biases so that you don't realize it. Yeah, we, we do the same. I mean, we have um whether you are a woman or a man or, or you come from one part of the world or another, you have these biases and you, you have to start by confronting yourself and then you can bring it to the others. What is good about the unconscious bias versus conscious bias is that it doesn't have to be a hard confrontation. Uh People who identify egalitarian progressive, they want to be fair. Those people, they actually wanna be, want to do the right thing. You know, if you are able to show that their biases are not coming from a bad place, it's just an evolutionary mechanism that's supposed to help us function.

But at some point it goes over the board and then it fires and then you deny yourself the talent um or you hire the people for the wrong reasons, you give responsibility to people for the wrong reasons. You don't get the better person doing the better job and this is a disadvantage. So in if it causes that they want to become aware, they want to actually uh cooper. So I think that if you take it, um if you emphasize the unconsciousness, the accusation level are lower and you can have a conversation. Um the conscious bias is harder because if, if people actively want to say that, yeah, this is how it must be, you know, we must oppress and I mean, attending people who attend this conference probably never have these views, but even uh ourselves, we must check um whether we are actually suffering from this, potentially we are all right.

Um Good. Um So any, I don't know if there's more questions or I don't know, how am I doing with time? I believe our session is like six minutes over and um but I see still some people here and I don't see the organizers kicking me out. So, um the um woman tech team, can we continue a little bit more or should we wrap it up? Thank you. All right. I think we can maybe wrap it up. Um I'm um probably missing some instruction I might have been given um as to if I'm supposed to close the session, uh or if someone else, but if you need to be in another session or so, of course, we can um wrap it up here. Um Yeah, thank you, Rebecca. And um yeah, thank you everybody for all the nice kind words. Um Yeah, and um I wish everybody a very nice day, afternoon, evening, wherever you are. And I also find it fascinating. Hopefully we can discuss broadly in some other opportune, other events and yeah, be in touch and take care. Thank you, Rebecca by the way for nominating me to begin with. Yeah. All right. Bye.