Fairytale about overcoming cognitive biases

Mariia Hrabovska at UXDX Community: Mastering AI Tools and Cognitive Insights for a Captivating UX Portfolio. Video: https://youtu.be/XqpSJlswSsk

Readable transcript: edited from the recording's captions for readability (fillers and false starts removed, punctuation and section headings added). Wording is the speaker's own. Timestamps are positions in the video. Names marked [?] could not be verified against the audio.

Three stories we can all recognize ourselves in

[00:00:00] Today we'll be talking about overcoming cognitive biases, stakeholder biases in particular. For the beginning, just a few words about me. For the last almost 10 years I've been working in different kinds of companies. I started as a freelancer, continued in a software development company, and recently I've been working in the financial giant Fidelity Investments. I was also a speaker at different conferences, a lecturer for design schools, and I've been writing articles for different platforms.

[00:00:35] As I mentioned, today I will focus on stakeholder biases, some stuff that can prevent them from considering new ideas or interfere with their perception, and also about biases that can curse us when we do our research. But there are also so many ways that they can be applicable directly to our decisions in UI elements, and if you're interested in those, here are my articles on Medium that shed some light on the topic. I don't know if there will be any sharing of materials, but you can just find me on LinkedIn, I'll gladly share it with you if you're interested in those.

[00:01:17] For today I'm going to start with three short stories, or rather situations, that I believe a majority of us can recognize ourselves in. First one: imagine that you come up with this amazing hypothesis for your product. You start testing, and a few people seem to contradict the idea. So do you continue seeking confirmation of your hypothesis? Do you feel like helping people to come to the right conclusion? Or do you accept this disproof easily?

[00:02:03] Here's another story. You are hired to redesign a legacy system. You did a UX audit and it looks awfully user-unfriendly, and everyone agrees with this. But you started interviewing employees and suddenly you faced confrontation. They do realize that the system is extremely insufficient, but they still don't want to cooperate when they understand the changes are coming. So would you do something about that, or just continue taking their questionable input?

[00:02:42] And the last story, and I believe it's a very common one. Last year I was at a Nielsen Norman training and I heard stories like that a great amount of times from our fellow designers of all levels and companies. So imagine that you and your team need some more budget or time for doing more research. You spoke with some executive person about the value of such research, you prepared a very professional presentation, you talked about all possible methods and frameworks, you used all the fancy designer words and terminology. You actually did a great job, right? But then your department still didn't get a budget and you didn't get support from executive level. What a disappointment, right?

What cognitive biases are, and the popular ones

[00:03:41] I'll be back to those stories shortly, but before that I want to quickly read a definition of cognitive bias. A cognitive bias is a systematic pattern of deviation from norm or rationality in judgment. The human brain has a tendency to simplify information processing through a filter of personal experience, and the filtering process is a coping mechanism that enables the brain to prioritize and process large amounts of information quickly.

[00:04:18] There are almost 200 known cognitive biases so far, and I can assume that the majority of you heard about such popular biases as confirmation bias, which is the tendency to search for or favor information in a way that confirms or supports your prior beliefs or values. Or anchoring, the bias that causes us to rely too heavily on information that we received early in the decision making process. The most famous example of that is how we show that something is on sale: they always show the new price compared to a bigger one.

[00:05:01] Or sunk cost fallacy, that forces us to finish an uninteresting book or film. It's a phenomenon when a person is reluctant to abandon a course of action because they have invested heavily in it, even when it is clear that abandonment would be more beneficial. Or of course the Dunning-Kruger effect, when people with limited competence in a particular field overestimate their real abilities.

[00:05:32] So basically a lot of them now are really popular and repeatedly discussed, but today I want to talk about a few less known biases that are extremely widespread in our work. I can tell that I notice them so often in my and my colleagues' behavior. Just a little disclaimer: I'm not a psychologist and I don't have a degree in this particular field, so I won't talk about any neuroscience particularities, but I fell in love with this topic after reading Thinking, Fast and Slow by Daniel Kahneman, and since then I started connecting the dots between biases and my work.

The biases hiding in the three stories

[00:06:14] Now let's go back to our stories I told you at the beginning and try to recognize at least a few cognitive biases here. In our first story, about a product hypothesis and testing it, we might have faced congruence bias. It is the tendency of people to over-rely on testing their initial hypothesis as the most congruent one, while neglecting to test alternative ones.

[00:06:48] And when testing is already going on, another bias comes into place, which is observer expectancy bias, when a researcher expects a given result and therefore subconsciously manipulates an experiment and influences the participant of this experiment. This happens when you ask leading or closed questions in testing sessions, when you ask people, will you use this product, or will you use this function. And in this case actually another bias might turn on, which is courtesy bias, when people don't want to offend us or hurt our feelings and therefore don't tell us the truth or the whole truth. So we really try to avoid this.

[00:07:36] In our second story, about redesigning a legacy system, we definitely recognize ambiguity aversion, or uncertainty aversion. It is the tendency to favor the known over the unknown, including known risks over unknown risks. And actually a very similar one, status quo bias. It is a preference for the maintenance of the current or previous state of affairs. It is a preference to not take any action to change this current state, no matter how bad things are going.

[00:08:16] When changes come, most people tend to freak out, freeze and stay where they are rather than taking the slightest risk. So despite realizing all the insufficiency that the current system might have, people are extremely lazy to change their habits. And it's even worse when people are afraid that the new system will replace them.

[00:08:43] And the third story is my favorite, because curse of knowledge is definitely one of my favorite biases. This bias occurs when a person communicating with other people assumes that these people have the background to understand what you are talking about. Let's remember about this bias when speaking with users, stakeholders or any person that is outside of our professional field.

[00:09:12] The importance and value of design research may be so obvious to us, right? All the methodology sounds so familiar and easy. But the truth is, nothing of it is obvious to our business stakeholders, or sometimes company leadership. And actually it shouldn't be. They have enough on their plate that we don't understand, right? That's why it should be a part of our job to explain everything in plain, ordinary language, to ensure that they are on the same page with us. So instead of using some pretentious smart words that can impress our colleagues, let's try and use simple ones.

Coping techniques for each bias

[00:10:06] Here I'd like to summarize some coping techniques for these specific biases. So with congruence bias, continuous discovery will help. Let's ensure that we have time to learn from our iterations in each sprint, for example, if you are working in an agile environment. Traditionally, product discovery is the first stage of the product development journey, so once you gather users' needs, pains, et cetera, you move on to the delivery stage and you kind of feel confident that you're building something that users actually want. But a digital product is never finished, so discovery shouldn't stop once you start delivering your validated product.

[00:11:05] This approach has been popularized by the author of Continuous Discovery Habits, Teresa Torres, and according to this book you should be gathering insights from customers on at least a weekly basis. These insights can be gathered through prototype tests, customer interviews, surveys, et cetera. In fact, any research method can be adapted to this continuous discovery process.

[00:11:36] For avoiding observer expectancy bias, we must review all our questions from the point of objectivization. A really great resource for this is a book named The Mom Test by Rob Fitzpatrick, and a lot of great advice for building scripts and testing hypotheses is there. The Mom Test has three rules. First one: talk about their life instead of your idea. Second: ask about specifics in the past instead of generics or opinions about the future. And third is to remember that you should talk less and listen more.

[00:12:34] Ask open-ended questions. Instead of "do you like this feature better," ask "what feature have you used the most lately." Try to get to know users' real behavior rather than hypothetical. Ask "describe a situation when you needed this last month," rather than "when do you think it will be beneficial for you." And of course don't forget about clarifying questions. They can reveal the whole truth, or sometimes users' real motivation.

[00:13:14] For reducing ambiguity aversion, try to get acquainted with your stakeholders. If you're collaborating via a workshop, you have an amazing opportunity to do ice breaker exercises to get acquainted with the participants. You can find a great amount of them on the internet, just find what fits your particular situation. For example, I like to use the game Two Truths One Lie, when people are supposed to write down two real facts about themselves and one lie, and then everyone tries to guess what is what. And that one lightens the atmosphere and breaks borders. It's hard to emphasize the difference before and after this exercise. People are so much more open and relaxed, and so they're prone to be productive and give meaningful insights.

[00:14:20] And when you are facing status quo bias, try and frame your message the way that reduces all possible risk and negativity. Try to leverage upcoming benefits and avoid mentioning possible losses and change. We can fight this bias by using another one, the framing effect. It is a cognitive bias where people decide on options based on positive or negative connotations. So for example, the sentence "the survival rate one month after surgery is 90%" is more reassuring than the equivalent statement that "mortality within one month after surgery is 10%," right? So when we face people's resistance to change, let's focus on positive aspects those changes will bring to their work and maybe personal life.

[00:15:25] And curse of knowledge. For this one, identify differences in background between you and your audience. Doing your homework by researching your stakeholders' background is always a good idea, not only because of the insights that you might find out, but also for you to realize your differences. Accept that you might be living in totally different worlds and they don't necessarily have to understand what you are talking about. Don't be afraid to sound too simple. It can never hurt a conversation. On the contrary, it can ease cognitive load, and therefore your audience will understand your point better and take your side.

Five general ways to avoid biases

[00:16:20] And just a few pieces of advice that are fair, I believe, for avoiding most of the biases in general. First one: learn about possible biases. Recent research reveals it is in fact possible to reduce biases through training. I will not tell you the details of this research, but in that study researchers provided one group with information that helps participants understand the topic of cognitive biases, and only after that they were to make a decision on some case. And the second study group wasn't provided with such training. The results showed that such training reduced the effects of those biases by 29%. So we are not total prisoners of our subconsciousness.

[00:17:29] Second one, cognitive ease, which is the ease with which our brain processes information. Don't overkill with it, as we remember that biases are just our coping mechanism for simplification, and when you overload with tons of details, people's biases are thriving.

[00:17:52] Third one: use processes and checklists. We can't just switch off our biases, no matter how much we know about them, so use predefined and approved processes, especially when we are conducting research with all different people. Never go to an interview or usability testing without a script. Therefore you will have less chance to fall into our own mind tricks.

[00:18:27] Fourth: frame your message. Again, framing is a huge bias that affects us. We tend to see risk when a negative connotation is presented and don't see them with a positive one. So review your message from this point before exposing it to people.

[00:18:48] And fifth, of course, give it time. As cognitive biases are just shortcuts for our brain, let's not push decision making right away, otherwise they will be biased the most.

Ethical design and wrapping up

[00:19:06] And just a few words about ethical design. This topic becomes more and more popular nowadays. Some companies start realizing how much influence they have on people and their lives. Maybe you remember a film named The Social Dilemma that emphasizes manipulating technologies and how they step by step distort our perception of everything. And the fact is that the lion's share in that belongs to designers. Knowing human psychology helps us not only build a better product, but also gives us some kind of superpower to point the user to the directions we need.

[00:19:59] So here are just a few questions that we can ask ourselves. Are users still in my focus? Am I considering the long-time effect of my design decisions? Am I ready to reveal what's behind the curtains? It's very easy to forget about social duty when trying to achieve some business goals. That's why I think every designer should think about the line they will never cross, especially regarding products for children and teenagers, as this audience is super easy to manipulate and our impact on their lives can be enormous.

[00:20:46] So the bad news is that we can't get rid of cognitive biases completely, but the good one is that the better we understand them, the more we can do to leverage them for our own benefit or for the benefit of our users. And just accept that we are human after all. It's our inevitable part, so let's embrace them rather than fighting some unnecessary battles. So that's all for me. Thank you for your time and let's stay connected.

Q&A

[00:21:28] Host: Excellent, thank you very much, Mariia. I guess the first thing that I can say is, when you put up that graph of the hundreds and hundreds of different biases, it's kind of a little overwhelming in ways. So I love the examples that you gave on focusing on five of them. But do you have any advice for people on how you try to keep on track of all of the different types of biases, to make sure that when you're putting together your questions and you're asking, that you're not trying to lead, but maybe you're making five other mistakes? Do you have any advice?

[00:22:10] Mariia: I don't think that it's possible for a human to keep up with all those biases, except if you're maybe a psychologist or a neuroscientist. But I don't think you need to, because actually not all of them are our enemy. It's just some shortcuts that help us to move more quickly, and we don't need all the time to overthink each piece of information. But I think just to remember some basic ones, like confirmation or this framing effect, it's helping a lot already to build our scripts for interviews and testings.

[00:23:08] Mariia: It's really helpful to just have a few of them open on the one side and the script on the other, and just to compare whether all your questions are objectivized, let's put it this way. Because it's super easy to fall into our own biases. We are human, and we can see easily some biases in other people, but it's really much harder to see them in ourselves. So yeah, just keep with a few main ones, a few powerful ones, and I think you'll be good.

[00:23:55] Host: Excellent. And yeah, reminder to everybody out there, we have three more minutes for questions, so please do post any questions. You're getting a lot of thank yous, which is positive, in the comments, so that's a good response. And Erin's asked about recording: yes, we will be recording this and sharing the recordings afterwards, Erin, so you'll be able to catch up.

[00:24:19] Host: While we're waiting, one of the other questions. Because you said avoid leading the person, and talk about the last features you've used rather than "do you like my feature." But what if your area of focus is on a particular part and you're trying to do research on that part, and people just keep talking about something else that's not in your remit? What should a person do there? Should they just follow that thread and share it internally with other teams, but then they lose out on the insight in their part of the product? Is there any kind of way that you can nudge without influencing?

[00:24:56] Mariia: Yeah, good question, thanks. I think that with such a situation, additional questions usually help to reveal the motivation, why they're talking about this particular thing and not about what you're expecting. Maybe what you're expecting is your own bias and that's not what we're actually looking for. If people keep talking about something else, maybe that's what we should focus our attention on too.

[00:25:40] Mariia: And I guess, even if we are talking about some new products, there are no features, no products, we can always ask about how they deal with this situation without any products on the market, some other way they dealt with it before. Because definitely, if there is a problem, then there is some coping mechanism they had before. So I think like that. I don't know if I answered.

[00:26:06] Host: No, it's a good one. It's kind of indicating where we should focus. We have more questions coming in, but I'll just ask Cindy's question because she asked it first, and then what we'll do is we'll post some more questions on our Slack group, so if anybody wants to join in there, I'll invite Mariia to answer the other ones. But Cindy's question is, would keeping a checklist of all the criteria and seeing if our questions match the criteria be good, or is it overdoing it?

[00:26:36] Mariia: Wow, it's a good question, but I think that depends on how much time you have. And again, we are human and there will always be some kind of mistakes. That's okay. But if we are already making an effort and doing this whole checklist, and not that deep maybe, but at least some kind of, we understand that we did it and we try to avoid subjectivization, and we try to keep ourselves outside this particular interview or testings, then I think that will be enough. I think if all designers will do at least this, we'll be living in a better world.

[00:27:28] Host: Yeah, it's kind of balancing. You need to do your best without becoming very rigid and a robot as well.

[00:27:35] Mariia: Yeah, there's never an end to the research and doing digital products.

[00:27:40] Host: Brilliant. Well, that brings us to time. So Stacy, we'll ask your question on the Slack channel, and if anybody has any other questions please keep posting them and we'll just post those through to Mariia on our Slack channel after the event. Thank you, and on behalf of everybody else, thank you very much, Mariia.

[00:27:58] Mariia: Thank you.