Kim Mills: All day, every day, we make decisions. Some are so small, we barely think about them, such as what to have for breakfast or which route will avoid the most traffic on the way to work. Others are more consequential—whether to take a new job, or how to spend or save our money. And some decisions can have life or death consequences, such as whether to get a vaccine or evacuate ahead of a hurricane.

Over the past several decades, psychologists and other behavioral science researchers have become increasingly interested in understanding how people make decisions like these, why we so often make bad decisions, and how even seemingly small changes in the way that choices are explained and presented can make a big difference in the decisions that people make.

So what have researchers learned about decision making? Why do people make bad decisions? Do bad decisions happen when people don't have enough information or when they're overloaded with too much? How do behavioral scientists define a bad decision anyway? And how can decision researchers findings best be deployed in the real world to make a positive difference in people's lives?

Welcome to Speaking of Psychology, the flagship podcast of the American Psychological Association that examines the links between psychological science and everyday life.

I'm Kim Mills.

We have two guests today. First is Dr. Lace Padilla, an assistant professor of cognitive and information sciences at the University of California, Merced.

Dr. Padilla studies how people use data visualizations to make real-world decisions with life-or-death consequences in areas including hurricane evacuations and vaccine uptake. Her lab's mission is to help people make the best possible judgments about their health and safety by developing and testing new ways to visualize and communicate complex data.

Our second guest today is Dr. Hannah Perfecto. Dr. Perfecto is an assistant professor of marketing at the Olin School of Business at Washington University in St. Louis, where she studies how consumers make decisions. Much of her work also focuses on improving research methods, including designing decision making studies that are more likely to replicate in real-world settings.?

Thank you both for joining me.

Lace Padilla, PhD: Thank you for having us.

Hannah Perfecto, PhD: Thanks so much.

Mills: So I just mentioned about how behavioral science research can help people make better decisions, but the term “better” can be a subjective term. What makes something a bad decision? How do you define “good” versus “bad” in your research?

Dr. Padilla, let's start with you.

Padilla: I think different groups of people define good and bad decisions differently, and there are some people who might think that a good decision would be the best computational decision. If we had all of the information available, we could calculate the most optimal decision. But the truth is that humans don't function like computers, and from my standpoint in my work, I think of a good decision as someone using all the information available to them to the best of their ability.

And in contrast, we can compare that to a bad decision and those are a little bit easier to identify, and I often see bad decisions as clearly someone misinterpreting or misunderstanding information that was presented to them that would ultimately lead to some type of error in their reasoning.

Mills: You two come to the topic of decision making from very different angles. So let's talk a little bit about each of your research backgrounds.

So, Dr. Perfecto, in your research, what's the main thing you're trying to understand about how people make decisions?

Perfecto: That is a great question. So I would say the work that I do tackles this issue from a broader lens—for better or for worse—than Dr Padilla's.

I spend a lot of my time making sure that the work that we're all doing in this field of decision making research is able to actually have a useful impact out in the world. So the work that we're doing, we're not learning about how people make decisions just for the fun of it. The hope is that in the end, we're actually able to improve people's lives down the road.

One area that I'm focusing on right now in some ongoing work is realizing that a lot of the ways that we've previously studied decision making is pretty abstract. So we are using very hypothetical designs, sometimes a little bit strange, a little bit artificial, and that's done so that we can have a really clear understanding of what is contributing to somebody's decision.

But out in the world, the world is not clean. The world is not so tightly controlled, and so something that I'm looking at is that in designing those studies, we've used a lot of positive stimuli. We've asked people when we're trying to learn about risk preferences or dealing with uncertainty, we say, “Do you want to have this gamble with only upsides or this gamble with smaller upsides?” And so we can certainly see if someone's willing to go for a bigger riskier outcome versus play it safe with a smaller positive outcome.

But both are still great. You still get something, you still have a chance at something in the end.

And I have found in my work and others as well, that those types of more positive decisions we're more likely to make errors in. When things are going well and things are feeling great, you're more likely to just go with your gut and go with that initial response. Sometimes that's fine, but a lot of times that then leads us to, as Dr. Padilla said, misinterpret the information we're dealing with, makes errors in evaluating those outcomes, and lead us to an outcome that is suboptimal for our situation. So I'm assessing the magnitude of that issue going back, looking at some things we thought were super duper robust and different really big effects that may not actually be so robust out in the world so that when we're actually trying to implement our findings, we will see results.

Mills: And, Dr. Padilla, what's the main thing that you're trying to understand? How is your research different from Dr. Perfecto's?

Padilla: My research is maybe different from most people who study decision making, honestly, in the fact that I like to examine how we can change the information that people are presented with to help them make better decisions. And I specifically study data visual stations. So I'm looking at creating forecasts with uncertainty and visualizing that data to try to help people understand the risk that they might be under.

So that's a wholly different angle that I bring to it.

Mills: So what are some of the main reasons that people make bad decisions?

Perfecto: That is such a good question. It is what hundreds of people have spent decades trying to get a good grip on, and hundreds more will in the decades to come.

I think both of us can speak to one or two particular reasons where people might go wrong. I touched on one already where folks might go with their gut a little bit too readily. I think that aspect of the positivity or negativity of what you're dealing with, people might not realize the extent to which that can influence how likely you are to make a snap decision versus sit with it for a while.

So I would say one main root of the problem is that people just sometimes go too fast. It feels right, they go with what they see, and then they end up somewhere they don't want to be. Whereas if they had taken a little bit more time and sat with it, they wouldn't have made those errors.

And when I say more time, I don't mean sleep on it, come back to it in a few hours. I mean 30 seconds, 10 seconds. Even that amount of time we readily find that folks can—And by “we” I mean not just me, but decision making researchers more generally find that people can recognize that they might have misinterpreted something or they're coming at the problem from the wrong perspective or even just that they made a mistake. And so by just taking a second, checking yourself, people could be much better off.

Padilla: Yeah, I really agree with that. And I think that that ties nicely into the way that I think about it, which is that throughout human evolution we have been able to make some pretty fast snap decisions that have worked out fairly well for us. We are now in a very complicated world with lots of things that there's no reason that we would've have evolved the ability to calculate long term financial stock projections, or any of the very complicated long-term things that we have to make decisions with. So for that reason, our intuitions sometimes fail us.

So if we can slow down and try to activate a more analytical approach, that can sometimes be very useful.

And I think in addition to that, I'll say my general refrain of, the way that information has been presented to us isn't easy to understand either. It is hard for the average person to look at probabilities and forecasts, and understand what they mean, and make effective decisions and it's not their fault. And I think that there's a lot more that can be done to make that information easier for all people to understand. And a lot of people are working on that particular problem.

So I think it's a dual issue of we have some strategies that might not be working for us, including making these gut decisions, but also the information that we're working with. It's very complicated, probably more complicated than and necessarily needs to be.

Mills: But some of the problem that we as a society are not as math literate or probability literate as we should be. So for example, when people hear that a vaccine is 90% effective, then there are people out there who say, well, but that's not good enough. What about the other 10%? And then they don't get the vaccine. What is wrong with us that we can't understand that 90% is a pretty good number?

Padilla: People always mention that wouldn't it be great if we all could be more math literate? And I agree. However, I do think that that's not the only solution because the American population is fairly well educated and we're not the only group of people who need to interpret this information.

If you can think about the people who are most in need of decision support, they likely have the least amount of education. So I think in some ways, blaming people for not being more literate isn't fair. I think that people want to make their best decision and they might be in a circumstance where they didn't have access to certain types of education, and that shouldn't mean that they have poor health outcomes.

I think it's our responsibility as scientists and good citizens of the world to ensure that we're not negatively impacting people simply because they didn't have access to high quality higher education.

Mills: Well then let me ask, what are some of the effective strategies for helping people to make good decisions and what are some of the strategies that don't work and that might even backfire by driving people to do exactly the opposite of what we're trying to convince them to do?

And maybe this is the place where I ask you both to explain some terms that are used in your research, choice architecture and Libertarian paternalism, for example.

Perfecto: Yeah, I think choice architecture one could describe it—I don't know if Dr. Padilla sees herself this way, but you could imagine framing the information, visualizing the information differently as a form of choice architecture. So when you think of the choice architect, it is the person who is structuring how the decision is phrased. It doesn't fall out of the sky. Someone has to decide, “Are we going to frame the decision positively? Do you want to increase your contributions to retirement?” Or more neutrally, “What do you want to do with your upcoming elections? Do you want to increase or decrease? Should increase come first? Should increase come second? When should you mention how much it's going to increase?” All of these decisions have to be made by somebody.

That is the choice architect. And the way in which all of those approaches—I don't want to keep saying decisions, but the way in which all of these little decisions that the person framing the actual decision needs to make, how that impacts what people actually do in the end would be choice architecture.

I think Libertarian paternalism comes from a particular way of being that choice architect, of creating those decisions, saying, that's the choice architect, “I know what answer you should pick. I don't want to force you to pick it, but I'm going to make it really easy for you to pick it and hard for you to not pick it.”

Mills: And so are those methods then that help people make better decisions?

Perfecto: See that now that's a thorny question. It depends on the benevolence of the choice architect. Certainly, a controversial question these days of who has the best idea what is best for people? So this development of this term choice architecture and libertarian paternalism is a big celebration of getting more people to save more for retirement. And that's great if you have the money to save for retirement. And so some people who might not be in such a good financial position, they might be able to do better for themselves with that money now instead of being funneled through defaults, and/or options are pre-selected for you and you need to do work to do something else. Like what Dr. Padilla is talking about, they might not be able to figure out how the decision has already been made in some way for them, and so whether that's the right move or maybe we've gone too far, is certainly a hot topic these days.

Mills: Dr. Padilla, I want to ask you, because you study some very consequential decisions that people have to make, such as how to decide whether to evacuate when you're faced by the threat of a hurricane. And I know you've come up with some data visualizations that help people, but what are weather forecasters doing wrong that drive people to make bad decisions when they are perhaps in the path of a hurricane?

Padilla: This ties into what Dr. Perfecto was just mentioning about who is the decider of what the right decision is. And the approach that I take is I don't know if everyone should evacuate. I'm a cognitive scientist, but I do know if someone misinterprets something, misunderstands information being presented to them. So I can at least identify when there is a massive failing in how people are interpreting what they're presented with.

So I take the standpoint of if I can at least reduce some of the errors, then that won't be in the process. I'm not sure if I'm making the decision better, but I'm certainly removing some of the errors. And we can do that with data visualizations.

And part of what has happened historically is people didn't appreciate the importance of how data visualizations were displayed to people. And there's some type of visualizations like the cone of uncertainty, which kind of starts at a point and the width of the cone grows, and it is intended to show the mean path of the storm. And it is a 66% confidence interval around the mean predicted path.

Well, most people don't know what a 66% confidence interval is, and so instead they see it as the storm growing in size over time because they see a small point and it gets bigger, so reasonably they think the storm's growing in size.

That is, I think, a visualization that hadn't been tested for the last 30 years that it's been in use, and so no one really appreciated that it was causing massive confusion and misinterpretation of what the storm was going to do.

And more recently we've been studying these in a careful way and studying alternatives to those classic visualizations that at least don't have those same misinterpretations.

Again, I'm not trying to make someone evacuate or not, I'm just trying to help them understand where the storm is going to go successfully to empower their decision making.

Mills: Let's talk for a minute about the dichotomy between time perception and decision making. So if you're making a decision what you're going to wear to work tomorrow, that's one thing. But if you're trying to make a decision about how we save the planet, what do we do to stop global warming, people have a problem understanding something that is so far out in the distance that they don't know how to make any decision about. What can we do to get people to better understand how to make those kinds of decisions effectively at this point in time?

Perfecto: You were hinting at it a bit, Kim, of making that really far off abstract thing much more concrete and feeling real here.

There's some cool work by Hal Hirschfield that talks about imagining yourself farther along. And so he does cool stuff like aging people's faces in photos and being like, “This is you in 50 years now. What are your thoughts about what's going to happen in 50 years?” And you're like, “Oh, that's me. I'm going to be there.”

It feels more concrete, and so we think about it as feeling sooner and evaluating it like we would a more closer in time event. So like I said, like you hinted at, taking steps to make that feel real now could certainly be beneficial.

Mills: So it's personalizing it.

Perfecto: Yeah. So that is one way to do it. Certainly, you could think of time. So there's this whole literature on construal level theory that talks about how things can feel very abstract and concrete, and that abstract versus concrete continuum can emerge in time like things are far away in time versus close in time. But they can also emerge interpersonally, like things that matter to me versus things that matter to people farther away for me that are less related to me, concepts that are less related to me.

So psychology is very related to me. Economics is maybe next closer, but theoretical physics much farther away from me. I don't do anything related to theoretical physics.

And so by bringing it closer to me on any of those dimensions can help make it more concrete. So putting myself into that situation can help.

Mills: Dr. Padilla, any thoughts on how to help people visualize things that are so far in the future, but help them make the right decisions now?

Padilla: Yeah. I think one of the things that data visualizations can do is to remove some of that obstruction. We can show, similar to what Dr. Perfecto was indicating, what the floodplain of your home will look like in 50 years. And you can use that information, you can just see it, to help you make in decisions now about getting flood insurance, or we can show you what it would look like if we did forest restoration in this area and how we can reduce the risk over long term, and you can just see it with data visualizations in ways that you wouldn't be able to before.

So I think making things more concrete, less abstract, especially in these long-term future projections, I think that's really one of the powers of data visualizations that have just recently been considered as a resource.

Mills: So with the advent of the internet, we now have access to huge amounts of information in an instant, whereas in pre-internet days, if we wanted to make a decision regarding say, which car to buy or what college to attend, the research was very labor intensive. You might have to go to a bunch of dealerships, you'd talk to your friends and neighbors, maybe you'd go to the library and look at books. Now everything is immediately available on the internet. And I'm wondering is that making decision making easier for people or more difficult? Do we know?

Perfecto: I know Barry Schwartz, a longtime psychologist at Swarthmore and Berkeley talks about how when we have so many options available—the work on choice overload, whether having so many options and I feel I can't choose or feel less happy with that—is mixed, still an ongoing debate, something that's hard to study in the lab that probably does exist out in the field. But Barry Schwartz's work it says if we have a lot of options, it feels like we have a lot of potential, we should be able to find the perfect thing.

I'm in marketing, so I always couch it in products and whatnot. But if I'm looking for a college, I should be able to eliminate my uncertainty. I have all these forums with students I can talk to. I have lots of resources available at the click of a mouse to be able to learn about the school, so I should feel really confident.

But all that information doesn't perfectly solve the problem for us, and so we might feel less satisfied with what we come up with in the end, that I feel like I should know exactly what I want. I feel like I should find everything, but I don't, and so now I'm less happy with what I get.

Mills: Dr. Padilla, easier or harder?

Padilla: I think more complex, would be my answer to that. I think what happens partly is that we are information foragers. We go online and we search for information. And so the way we do that is biased. We're looking for things that affirm our own beliefs most of the time. So we have a tendency to find things that affirm our own beliefs.

And within that we're exposed to lots of misinformation and being able to identify what misinformation is very challenging, especially if it affirms your own beliefs. You're more likely to believe it's true if it matches your own beliefs. So we have a whole new set of problems that we have.

No, I haven't developed any skills at combating, which makes the whole thing very complicated. And I think for me, it really brings this new concept of trust to the forefront.

Now we have to evaluate the information that we're taking in and think about how much we trust it. And trust is a whole complicated area of exploration that doesn't necessarily correspond with how effective or useful information is.

Oftentimes, we trust things that don't always help us make the best possible decision. At least that's what my work has been finding, that they don't always go together. So I think that that's a new problem that the internet is presenting to us, is that we now have to evaluate trust and we have to forge this information for us. And there are certain people who are profiting off of trying to get us to trust things that aren't necessarily true.

Mills: So is artificial intelligence going to save us? If we can use artificial intelligence to sift through a lot of this data and maybe identify the things that are fake? We're confronted with a lot of information as you just said, Dr. Padilla, that isn't. It's just not accurate. What role will AI be playing going forward, do you think?

Padilla: It's really going to be playing a massive role. Maybe not as much in our own personal decisions, but certainly decisions that are made on institutional government level high impact decisions. And I feel like I'm an outlier in saying that AI can be very useful. Every time I talk with colleagues about this, the one who thinks AI could be good, but there's this general apprehension to trust AI to give our decisions over to artificial intelligence, partly because there's a whole history of AI also making biased decisions because they're made by people, and people have biased input data and have biased different algorithms and so forth. And I think that's a very real problem. And I'm very excited that there's many ethicists who are working on those particular issues. But I do think that we should start to get comfortable with a future that is not far off where many decisions will be curated in some capacity by AI, where rather than having thousands of things available to us, they'll be some type of decision scaffolding that AI will be doing for us. And that's the future that I see, which I think could be a good one if there's a lot of people involved in the AI development process and not just certain groups of people that don't necessarily represent all of the interests of the people making the decisions.

Mills: Dr. Perfecto, do you share those thoughts or do you come down on a different side of the question?

Perfecto: Yeah, I think it's that last if that is the most important component, because as Dr. Padilla said, and like we were talking about with choice architecture, the people are the ones making the decisions, whether it is in terms of choice architecture, actually forming the decision that is being made for you in that moment, the actual thing you're going to see when you're electing your retirement plans, or whether they are farther upstream, creating the thing that will eventually help create that decision scaffolding, as Dr. Padilla said, to then present you with a decision and options down the road. So I think that's a really big component to consider.

But I think everyone acknowledges Dr. Padilla's main point that it's happening. AI and algorithms are constantly increasing every day their presence in our lives and in our decision making, and so there's a lot of work ongoing in the field of decision making on the academic side, investigating how people evaluate algorithms, and how and when people shy away or shun completely algorithms, certain contexts, what happens when they make mistakes, how do people respond to that and how we can improve people's perceptions of them, and making sure that since oftentimes the algorithm is going to help you make a better decision, if, the big if, it's made properly and has the right inputs.

And we want people to be excited and willing to go with this good decision that the algorithm is helping you make. So a lot of cool work ongoing by people learning about ways to facilitate that adoption of algorithms in people's decision making.

Mills: So let me ask you both, what are the big questions you're working on now? Dr. Padilla, what's your research looking at in right at this moment?

Padilla: I'm really interested in the question of how to facilitate trust and what that means. I think partly because I study a lot of forecast visualizations, there's lots of interesting questions about how much uncertainty is useful to show, particularly in facilitating trust.

It's flooding in California right now. If there was a forecast that said every day this month there was some potential of flooding, that would be correct, but in some ways it would feel like a cry wolf effect. It's saying that there would be flooding too much going on. So what you really want is a forecast that gives you just the right amount of information to make your decision and not too much uncertainty because then you can't make your decision with it. So I'm looking at trade offs and trying to identify the sweet spot and how much uncertainty we want to show people and how that can support both trust and performance because I think there's certainly thresholds there that haven't been carefully identified.

Mills: And, Dr. Perfecto, what are you looking at these days?

Perfecto: I teased it a little bit earlier. I'm right now spending a lot of time getting a sense of weather and how we've been creating these more abstract studies in the lab, whether and how that influences the likelihood of success down the road. How closely is the lab matching the real world, since in the end that's what we really care about is helping out the real world? And so in doing that, I am having lots of fun going through lots of longstanding literature, identifying common mistakes made, rectifying those mistakes and getting a sense of the impact that that might have in terms of when this particular error or bias might be made out in the real world.

Mills: And for our listeners, any words of wisdom? I think, Dr. Perfecto, you talked about not rushing to make decisions, to even just take a few seconds to think.?

Dr. Padilla, any other ideas that you can offer to our listeners, more practical advice on how to make better decisions?

Padilla: Finding situations in which you can play around with the data can start to give people an intuitive sense of some of the uncertainties and variability in the data. I think what tends to happen is people in the general public, they just get shown a forecast on the news and they just have to decide, “Do I trust it? Do I not? What do I do with it?” And that's a very one-sided path of the information.

There are some places that you can go online that allow you to interact with the data and give you the opportunity to say, “What if I extend this forecast a little?” Or, “What if I change the parameters of this forecast?” And then you can kind of get a sense of what the different aspects are affecting this particular model for really any type of natural disaster or biological disaster.

They have these with COVID. They have it with fires and hurricanes, and so forth. So if you're wanting to make a better decision for those types of things and you sincerely care about it, consider getting a little bit more comfortable with the data yourself and finding those types of sites that empower people with the data that they can play around with, that that could help to give people this more intuitions about what the forecast will do and the uncertainty that goes into those forecasts.

Mills: And, Dr. Perfecto, last word. Any practical advice?

Perfecto: In addition to, hold on a second, hold on 10 seconds and think about it, I would say to make sure you are bringing the right tools to task, that if it's a very difficult decision, then you should really take some time and sit with it. If it's a really easy like, what cereal am I going to have, maybe don't. We only have so much time here on this earth. And just grab the Cocoa Puffs, whatever. You can take the special cake tomorrow.

And making sure that you are answering the question that is being asked. So if you are in the supermarket and you're trying to decide what to get, and the easiest question is what do I want to buy? A harder question is what should I be buying? What is the healthiest option? What did my doctor say that I should get? That's a harder question to answer maybe when you don't want to answer as much as, “What do I want?” And so making sure that you are actually not substituting an easier question for a harder one can also be helpful.

Mills: Well, I want to thank you both for joining me today. This has been really interesting and enlightening. I appreciate your thoughts and I'm going to grab the Cocoa Puffs. Thank you.

Perfecto: Enjoy.

Padilla: It was great talking with you.

Perfecto: Thanks so much for having us.

Mills: For previous episodes of Speaking of Psychology, you can visit us on our website at www.speakingofpsychology.org, or you can find us on Apple, Stitcher, or wherever you get your podcasts. And if you like what you hear, leave us a review.

If you have comments or ideas for future podcasts, you can email us at speakingofpsychology@apa.org. Speaking of Psychology is produced by Lea Winerman. Our sound editor is Chris Condayan.

Thank you for listening to the American Psychological Association. I'm Kim Mills.