Unlocking Behaviour Change Podcast Episode 6: How we assess options when making decisions
This podcast episode looks at how we sometimes analyse pros and cons and sometimes engage in pattern matching. And it looks into the potentially deadly 'confirmation bias' and how to avoid it.
JAMIE WEST
Hello and welcome to the Unlocking Behaviour Change podcast, the show all about psychology and changing behaviour. My name is Jamie West, and we have a very special guest today who is our guest every time we do the podcast! It is Professor Emeritus Robert West. You had a very special birthday not long ago.
ROBERT WEST
The big seven zero.
JAMIE WEST
Well, I'm no mathematician, but that's 70. How was it?
ROBERT WEST
It was very nice. We were in Portugal, in my pretty much favourite place on the planet, just outside Lisbon. And the weather was lovely. The people were lovely. The vibe was lovely. The food was lovely.
Slight issue, I have to say. Which was that I had a small abscess under a tooth, which meant that I had to take metronidazole. And metronidazole is a drug that they give to alcoholics to stop them drinking. Because if you drink any amount of alcohol while you're taking metronidazole, big tip here for you folks out there, it makes you feel unwell. So I couldn't drink alcohol.
JAMIE WEST
Did you resist drinking alcohol, or is that what made you feel ill?
ROBERT WEST
I didn’t feel ill because I didn’t drink alcohol! Anyway, all is well now.
JAMIE WEST
Well, I'm glad to hear that. Happy Birthday. And I do need to make a correction. In the last episode I got the name of the chocolates that I eat wrong. I called them Le Petit Prince or something like that, but they are actually called Le Petit Chocolat.
ROBERT WEST
That's not the only mistake that you and I have made in the course of these podcasts!
JAMIE WEST
I'm sure we can have a whole podcast dedicated to mistakes we've made.
Anyway, this first season is focused on decision-making. We've introduced the FACET framework to help us understand decisions and break them down into their different components. As a recap, you don't have to go back and listen to the previous episodes, but you can if you'd like.
So it's called FACET, and it stands for Framing, Assessing, Comparing, Enacting, and Timing or Time. In the last episode, we covered framing, and we went through a whole range of topics about how important it is to frame a decision in a helpful way and how often you have to reframe decisions as you go along. However, today we're going to move on to assessing. What is assessing?
ROBERT WEST
So, you've reached the point in the decision-making process where you've asked the decision question. That's your frame. What shall I do? Shall I do this or not? Which of these things shall I choose? And now, you're looking at the options that are in front of you, and you're trying to figure out what the best course of action is.
There are two main ways we do this. One is a kind of pros-cons analysis. For each of the options, we look at that option and we say, what are the good things about that option? What are the less good things? What are the bad things about that option? And we do that for each of the options under consideration. So this is the classic pros-cons, cost-benefit type analysis, which is the basis for so much of decision theory, including economic theory. And we'll come to this in a minute, but models such as subjective expected utility theory, which focuses on the pros and cons in terms of what might happen in the future that would be good or bad.
JAMIE WEST
Before you proceed, I'd like to get a concrete example that people can understand. You have a bunch of guitars - I have a bunch of guitars. Would you be applying this pros-cons approach to buying guitars?
ROBERT WEST
Generally, yes. So let's say that I am in a guitar shop and have decided to buy a guitar. I’m choosing from among the guitars in the shop. In a pros-cons type of analysis, for each of the guitars, I would say, what are the positive features about this and that guitar? What are the less good features? There will be features like, what's the ‘action’ (the height of the strings from the fretboard)? What does it look like? What is the sound from the different pickups? What is the feel, the weight, the balance of the guitar? What is the make of the guitar? Some of us who are a bit old-fashioned think of Fender and Gibson and guitars of that era as the classics, and that probably comes into the decision a bit.
And then on the other side, you've got things like cost and so on. You would weigh these up and then you would come to a decision based on some kind of comparison - which we will cover in the next episode - but in order to make the comparison, you have to look at these pros and cons.
Another example is when you're trying to decide on a candidate for a job. Each candidate has strengths and limitations. This candidate might have a lot of experience, but they might be weak in something like teamwork or whatever. So you're weighing up the good and the bad features of those options.
On the other hand, if you're focusing on a decision where you're forward-looking, you're thinking about what might happen as opposed to the attributes of the options you've got available to you. A gambling decision, for example, or an investment decision. You're looking at the stocks and you're trying to predict what is likely to happen to those stocks? What is the probability that the stock is going to grow or produce a particular dividend over a particular time period? And if it’s a startup, there's potentially high gain but also high risk. So you've got the costs and the benefits in terms of expected outcomes, positive and negative.
JAMIE WEST
So with one kind of pros-cons analysis, you are looking at known features of the options, and in the other kind, you are speculating about the future.
ROBERT WEST
The second of these approaches is labelled consequentialist or expectancy-value models, of which subjective expected utility theory is the most widely known example. The most widely known model of the feature-based approach is called multi-attribute utility theory. And those are two types of pros-cons analysis.
JAMIE WEST
This is important, so just to recap ....
ROBERT WEST
So, in expectancy value models you are looking at what the possible outcomes might be and evaluating the likelihood that each of those outcomes will occur and how much you like or don't like them. The likelihood and degrees of liking (value or utility) for all possible outcomes from an option are combined into a single value, which is the expected utility of that particular option.
And then you compare the expected utilities of the different options. And there are lots and lots of different ways of doing that, and lots of different theories about how we evaluate things and come up with different probabilities.
The other approach, the multi-attribute utility approach, involves assessing the known or believed features of a particular option. And then comparing those features according to how important you think they are.
Purchasing decisions are very often like this: you know, buying a car. So the features you're looking for might be fuel economy, the price of the car, colour, style, presence of particular features such as air conditioning, comfort of the seats … Those are all features that are more or less important for different people. And different cars have those features to differing degrees. You put them together so that you get a general sense for each car that you're looking at, and you make your comparison on that basis.
JAMIE WEST
And when shopping online, there are very often tables, aren't there? Comparing similar models, be it a car, be it a smartwatch … whatever … saying this one is waterproof, this one has a longer battery life ...
But we don't generally perform these as calculations in our minds, do we?
ROBERT WEST
That's why these are more ‘models’ than ‘theories’. They are not saying people actually do these calculations ... it's as though we do them.
JAMIE WEST
Okay, so what else?
ROBERT WEST
The other way in which we assess options that is very different from the pros-cons approach is pattern matching or if-then rules.
JAMIE WEST
What is pattern matching?
ROBERT WEST
So with pattern matching, we say: if this is true, that is true, and that's true, that's not true, then do this, or this would be a good thing to do. An example of this would be the diagnostic decisions that clinicians might make.
A patient comes in with a set of symptoms. And based on how well those symptoms fit a particular pattern for one or more different diseases and the history of the patient - what they've had in the past, possibly what their family history is - the doctor would come up with a diagnosis or a tentative diagnosis and treatment plan.
JAMIE WEST
They are using a whole body of research that has studied what diseases are linked to what symptoms and what treatments work better in what circumstances.
ROBERT WEST
One of the issues with the decision-making research literature is that people studying the pattern-matching type approach tend not to integrate this terribly well with the pros-cons approach. So they go down different pathways and the studies that are done fall into different silos.
And one of the things that we want to do, you and I, Jamie, in our book on decision-making ‘Reflect’ – due out at the end of this year - is to help people to integrate these, to figure out when people use the pattern matching approach, when they use the pros and cons approach, and when they go from one to the other.
JAMIE WEST
So I'm interested in this mixing and matching of pros and cons and pattern matching. Where might we do that?
ROBERT WEST
I think an example would be quite often if you start with a pattern matching approach and it's not delivering what you want in terms of a clear answer, or it's delivering an answer that you're not happy with.
Let's take an example of a pilot who gets a warning on their flight display saying that there's an engine fire. In that situation, they're not going to go through a pros-cons approach. They are trained to go through a checklist - the engine fire checklist - which is a pattern matching approach and so they say, ‘Where am I in the flight envelope? Have I just taken off? Am I in cruise? Am I coming into land?’ Whatever. Based on that, they could then start to make a decision about the best course of action.
However, there may be some exceptional circumstances, which would mean that if they follow the recommended approach, that could cause unwanted consequences or side-effects that they then have to weigh up against the fact that, yes, they have successfully put the engine fire out, but as a result of that, because of the situation they are in, they may well end up in a worse position than they would otherwise have been. So they then switch from a recommended approach to some other approach in which they apply a pros-cons analysis.
JAMIE WEST
So you first of all go through a checklist and then maybe you think, actually, if I follow this all the way through, it's going to cause unintended consequences. And you think, what are my other options? And then you move to a pros-cons analysis. Well, what's the pros of that? What's the cons of this? So I can see how you may be using both.
I'm very interested in pattern matching, partially because we love to spot patterns; we all know that human beings like to spot patterns where maybe we see faces in soups … and that's just the way our brains are wired. But I'm quite interested in how we can often make mistakes, falsely pattern match; we think, ‘Oh this is just like that’. We make a connection before we've really even challenged it or interrogated it.
ROBERT WEST
Yes, and that could be an example of confirmation bias. And it is so strong! I cannot overemphasise how strong confirmation bias is: to the extent where something could be right in front of your eyes, and you don't see it because you're expecting something else.
There's a good example of that in the aviation world, which, as you know, I study quite a lot, in which a pilot was expecting to see the lights of a runway. He saw lights - and thought that it was the runway. There was a lot of evidence that it wasn't the runway; they were the wrong colour lights, for example. But he was expecting to see a runway. The thought the lights were the runway and crashed the plane. And in our everyday life, confirmation bias is so prevalent I think that probably not a day goes by where we aren’t subject to some kind of error as a result of it.
JAMIE WEST
How can we avoid that?
ROBERT WEST
We have these expectations because of our limited processing capacity. We can't be analysing everything all the time. And so our brains fill in the gaps and jump forward. So it's not that the process that leads to confirmation bias isn't useful. It's just that you've got to be terribly careful and aware of its existence and be on the lookout for it.
JAMIE WEST
With events, such as political events, matching to what has gone before I think makes us feel secure. We think, it's happened before, and therefore we maybe can predict what will happen next.
ROBERT WEST
Confirmation bias arises from a number of factors, one of which, as you say, is the need to feel that the world is predictable. But part of it is simply the way our perceptual and cognitive apparatus works. It's what we do: filling in the gaps. It's not arising from any need to do anything. It's just that the brain fills in gaps.
One thing I would say: if there's some bit of information out there that doesn't fit, don't just ignore it. If you've got an uneasy feeling, pay attention to that because that is a cue that tells you that your expectation may be wrong.
JAMIE WEST
So don't ignore the outlier. It may mean that it just warrants a little bit of extra investigation, and you can carry on with what you were doing. However, it's worth examining the outlier.
So we're coming up on time for this episode. It feels like there could be more to say on assessment, maybe in another episode.
ROBERT WEST
There's much more to say. There are issues around comparing ‘chalk and cheese’ in the pros-cons analysis, for example. How on earth do you compare things when they're so different? What about what happens when there are no good options? That's a really common situation. How do you address that? And so I think we will come back to it. And there's lots of really good research to draw on.
JAMIE WEST
Well, I very much look forward to that. This has been very enlightening. And our first episode with you over the 70 mark. So all faculties seem to still be working, which is encouraging.
ROBERT WEST
Well, my Garmin watch says that my fitness age is 57, which frankly I find quite insulting because I do a lot of exercise. I think I'm actually fitter than that, but still …
JAMIE WEST
Well, I think 57 is pretty good. But anyway, thank you so much for listening, everybody. And remember to like and subscribe and review the podcast and links for our books will be in the description. And we very much look forward to being with you next time on the Unlocking Behaviour Change podcast.