Interview with Pedro del Carpio.
What are the mechanisms behind behavioral biases in risk situations? How can we change them?
Decisions made in a worker's daily activities can be influenced by cognitive heuristics and biases. That is, people habitually choose a course of action based on processes that are not completely "rational", and that lead them to perform suboptimal behaviors.
To learn more about how to apply behavioral sciences and achieve behavioral changes in companies, we invite you to watch the interview or read a summary of the most important points that were discussed between Pedro Del Carpio, director of Heurística Lab, and Pablo Pinto from Ludo Prevención. You can also listen to the audio at Spotify o Apple Podcasts.
Cognitive biases in work environments | Interview with Pedro Del Carpio - YouTube
Bias of overconfidence in the area of occupational safety
Pablo Pinto: Let me tell you, in safety we have a problem and that is that workers sometimes behave in a certain unsafe way and when we investigate accidents, we state that the cause is overconfidence. I came across a paper of yours where you talk about biases, one of the biases you talk about is the overconfidence bias. Let's start there, what is that bias and what is overconfidence bias?
Pedro Del Carpio: What do you think if we go to the fundamentals of behavioral sciences, their view on rational decisions and how they eventually lead us to make mistakes? In a normative way, that is, how it should be in a theoretical way or in an ideal way, a behavior, a decision, or an action should be determined by the exercise of estimating and thinking about what are the consequences of our actions and the probability of occurrence of those consequences. Whether positive or negative. In economics, this is known as utility. So, you determine the utility of a path of action, i.e., I go to "A", then you determine the path of action "B", and you observe that you have different utilities. The utility of "A", the utility of "B" and the utility of "C". In theory, one should take the course of action that is going to give you more welfare, more utility, more profit, more happiness or more security. Now, evidently this exercise in practice, we don't do it. People don't do this deliberate exercise of calculating probabilities by consequences, or multiplying and adding and comparing data. No, we don't. Why don't we?
Because we do not have access to all the necessary information. It is impossible to have access to all the necessary information around an event. We can have a little or a lot of information, but not all of it, that's impossible. Especially in complex systems. Assuming that in our day-to-day activities we had access to that information, that we had the cognitive capacity or the mental capacity to make those necessary calculations, and we also had more practical circumstances, or even had the time to take out a calculator and make estimates like the ones I mentioned, we have to react quickly. So what do we humans actually do?
We use mental shortcuts or simple decision rules, known as heuristics, which is this name that we take and then it is the name of the company, but where does this name come from? Heuristics is that, it's those simple decision rules. For example, a very intuitive one, very well known, is that of social norms, in the work environment. In a hypothetical case, one could have rules established regarding the policy of putting on gloves because those are the norms. However, if it is that in that environment or in that workplace, even though there are rules, most people, that is, the others, do not wear gloves, one is going to have the tendency in the face of the possibility of not knowing what to do, to do what the others do is the ideal. "I'm going to do what others do". In other words, the social norm, the shortcut here is "I do what others do, that is expected, that is desirable and that is my shortcut, my simple decision rule". So there is a collection in the discipline of behavioral sciences, a collection of biases.
By the way I want to make a parenthesis, these mental shortcuts that I mentioned, are not always bad. In fact, that is key, in the vast majority of events, these mental shortcuts, these simple decision rules are good, they are desirable. Moreover, these simple decision rules have allowed us humans to get to where we are today. In complex systems, where we have no information, where we cannot make methodical and rigorous calculations, these simple decision rules allow us to make decisions that are not optimal, not ideal, not perfect, but good enough to achieve our objectives in the different contexts we face. However, sometimes these simple decision rules, these heuristics can be bad. And I think in the work environment they are one of them because the risk of being wrong is very high. That is why we try not to have simple decision rules, but processes, checklists, a step-by-step detail, where people do not have to rely on their intuition, on heuristics, but they already know what to do in each of these events, in each of these activities or circumstances.
Pablo Pinto: So, in short, to see if I got it right, bias is a shortcut my mind makes to make decisions when I do not have all the facts.
Pedro Del Carpio: This is a heuristic, which can sometimes have systematically negative results. And that's when it's called bias.
Pablo Pinto: Ah ok, so bias has a negative connotation?
Pedro Del Carpio: Exactly, exactly. It's like the bad side of heuristics, although shortcuts are heuristics, when they fail they are known as biases. So the bias of overconfidence is the product of a subjective misestimation of the probabilities of success of my performance.
Pablo Pinto: Just what you just mentioned is what we call risk perception. In the security field we say that risk is the probability and severity of something manifesting itself.
Pedro Del Carpio: And when you think that doesn't apply to you, it's because you're overconfident about your own abilities. Therein lies the mismatch. This could be because of personal, personality, circumstantial things depending on the difficulty of the task and also as you say because of practice, because of the years that people have.
Pablo Pinto: Yes. Imagine, "20 years doing the same thing and nothing has ever happened to me". Then someone comes along, some young security guy comes along and wants to change the way I'm doing things that have worked well for me for 15 years.
Pedro Del Carpio: What I am saying here is that it is reasonable. It is important that those who evaluate these phenomena, also put themselves in the shoes of the person who does it and not just say this person is behaving irrationally, when in fact there is a rationality and there is a logic behind it. That has to be respected and understood. Now what the safety experts have to do is to consider that even if the person is skilled and that certainly the probability of risk with them could go down depending on what task is being performed, the risk of an accident, for example, although it could go down, that does not mean that it is not going to happen. That is, if the probability of an accident happening in an activity is 5% or something like that, getting the experience down to 4% or 3%, to talk about reference numbers, is good, but it still exists. That is, there is not going to be any experience that eliminates the risk. And there, the bias is that people eliminate the risk by their experience, by the experience they have in their mind, they do not eliminate the real risk, I want to emphasize, rather the perception of risk is subjective.
Pablo Pinto: Great, great, yes, just in security we talk a lot about that phrase, about risk perception and what we want to do is in a way to raise people's risk perception, especially those people who have a lot of time.
Pedro Del Carpio: Let me ask you, how are such initiatives received by experienced people?
Pablo Pinto: Resistance of course, right away. They say something like this, "this guy is going to teach me, it's like the son is going to teach dad, please, I have 15, 20 years of experience". So, when I am being supervised, I comply with the procedure. When the supervision goes away, then I adopt the way I've been doing it for 15 years straight without anything bad happening to me.
Pedro Del Carpio: And there again, if you allow me, because this is complex, it is not only the biases, but also the incentives behind it. If a worker is asked for security and speed, for example, there is going to be a problem. The person asks himself which is the most important incentive. And the answer is, "I am not paid for being safer, I am paid for producing more". So this contradiction of incentives can make one in his rationality say "ok, I am taking a risk but the probability is very small. But what is certain is that if I do not produce, I may be voted out or I will not be promoted in my position". So you make a quick, intuitive calculation and say, "you know what, I'm taking a gamble." That's respectable logic. So what organizations have to review is what their incentives are. Maybe they could do a financial incentive campaign for safety. A lot of people don't want to do that, they give economic incentives because they want it to be intrinsic motivation, but they could do experiments, they could test to see which one works best.
Pablo Pinto: Yes, yes. I remind my audience that Pedro is a scientist, that's why he talks about experiments. Pedro is a scientist. The scientist is based on experimentation.
Pedro Del Carpio: For me it is important to say that, actually, what I can share with you is an opinion of what can be done. An opinion of what could be done, but the rigorous, methodical, scientific and safe way, speaking of safety, is to create a series of interventions based on behavioral science principles and to conduct experiments. That is, you have a control group, it can be "office a", where nothing changes, it's the status quo, nothing changes. Then you have "office b" where you invent a strategy, "office c", "office d". And, then you compare the results after a certain time to see if your interventions, the ideas, the concepts you have implemented, the strategies or the tactics, have really worked or not. You don't do what the specialist says, or in this case what I say. But what the data says.
Building solutions upon a diagnosis
Pablo Pinto: Let's talk about this case of overconfidence. Imagine that I am the one who says to you, "look, in accidents 30%, or 40% are due to workers who are overconfident". What ideas would you recommend to me to try interventions?
Pedro Del Carpio: Let me tell you how we do it at Heuristics, which is the company I manage. The ideal is to do research on the barriers and facilitators of the behavior in question. Before prescribing what the interventions should be, you need to know in detail.
Pablo Pinto: Sure, like a diagnosis. It's like I'm asking for a drug and you're a doctor and he says "no, I have to make a diagnosis first".
Pedro Del Carpio: Exactly, that is a good analogy. The work that Heuristica does is like what a specialist doctor might do, not a generalist, but a specialist, who is the one who must understand the pains, the discomforts of the patient, in this case the client. And if you need more information than usual, you do qualitative or quantitative research. Going back to the analogy of the medical world, this would be doing the tests. The doctor sends you to do some physical tests or some laboratory tests. Based on those results, you can then make a diagnosis. And then, based on this very particular case of the patient, we can give a series of medicines, a series of practices, strategies or tactics that in the behavioral science world are called nudges.
Pablo Pinto: Yes nudges, is it like a little push?
Pedro Del Carpio: Exactly right, nudge is like a little push. The translation would be like a little nudge because it's not like I take you by the collar and say, "Do it because if you don't, I vote for you," which is also a way to change behaviors, because anybody could do it that way. In organizations, you could say, "If you do it, reward. If they don't do it punishment, humiliation." Believe me that could work, but not always.
Pablo Pinto: Yes. Hard behaviorism.
Pedro Del Carpio: Exactly, it could work, but there are other ways. One is education, which we have not mentioned. Yes, education is important, but it is only one part of the equation. The other part is to create a context that allows people, that workers don't have to rely on constantly changing factors such as their intuition, their emotions or their day-to-day habits, but that is structured so that the whole process of behavior is routed the way the company wants it to be. These nudges, going back to that definition, are small changes in the decision context that influence people's behavior without forcing them to do so. The principle behind nudges is that ultimately people have the decision, they are free to do it because there is no coercion, but you have created a context that increases the likelihood that they will do what you want. Which, in reality, is not what you want, but what the organization wants, i.e., it's ultimately for them.
Pablo Pinto: Okay, so, going back to the medical analogy, imagine that it is no longer that I say, "I have a headache, give me a medicine", but normally if it is a headache, aspirin is used. Could you give me an example on the issue of overconfidence? Do you know of something that could work to make the worker perform the desired behavior prior to his work? Do you have an example?
Pedro Del Carpio: I don't have it in mind, but it occurs to me that one can use, for example, reminders of the probability of the accident risk, a reminder at a random random moment: "today, just in case, don't forget, the probability of accidents is X". And although it is low, because it probably is low, otherwise we would be talking about accidents every day, the consequence... Moreover, now that I think about it, it is not so much to remind them of the probability or only the probability. Many times we forget about the consequences. A reminder of the consequences, in a random way, where the worker is not expecting it, I think can reduce the probability that one falls into overconfidence.
Pablo Pinto: Great, great. See, we do something called a 5-minute talk, but the problem is that we do it in a set way, and you talk in a random way.
Pedro Del Carpio: Of course, because they already expect it, one thinks "I already know what they are going to tell me" and people pay attention to novelty. For it to be memorable, it has to be novel too, it has to attract our attention, it has to find a way to surprise, I leave that to the creatives. How can we transmit that message in a striking and memorable way and also that it is not expected so that it has that effect?
Pablo Pinto: Yes, yes, great. So to summarize, a bias is always going to be negative, it's a shortcut within the heuristics, of which there are a lot. There are positive ones, and there are negative ones, which makes them biases.
Pedro Del Carpio: I would say that even, up to what we consider that, in some circumstances or in many probably, are good. That is, overconfidence could be very good, in aspects where there is no risk. For example, in sports. I have to believe it or say "hey, I'm the boss".
Pablo Pinto: Sure. When it boosts your self-esteem and or in places like a competition.
Pedro Del Carpio: Exactly, exactly. These heuristics have arrived in us, men and women of 2021, by an evolutionary process. Over tens of thousands of years, these intuitive behaviors have been engraved in the DNA of the species. And under this theory, this evolutionary approach would not have reached us if they had been bad. By natural selection, what is useful remains, what is good comes to us, what allows us to face the challenges of the context. However, in some circumstances, these heuristics can lead us to make mistakes. And they are particularly notorious in risky situations, such as in dangerous work environments. There, any heuristic, poorly calibrated, is going to be a bias, a dangerous or risky bias.
Pablo Pinto: And rescuing the last part, sometimes we are very adventurous in providing solutions. For example, we have an accident investigation format and we say, "cause: overconfidence, solutions: and at once we start to propose corrective actions" such as more training, an event or a campaign. It's like we give a lot of pills, a lot of drugs, without having done a more in-depth analysis of overconfidence.
Pedro Del Carpio: Yes, yes, going back to the example we started with about social norms, which we talked about the use of gloves, one could prescribe or imagine tactics to get new employees used to wearing gloves in the workplace. But, they would probably not be very effective, if the real origin, which can only be determined through a process of analytical study, the origin of the problem is because the referent, the senior person or their boss does not do it. So, if the referents of the group or the people with more experience don't do it, why is this person going to do it? So, a good behavior change initiative should start by studying the phenomenon, to see where the problem is, rather than necessarily prescribing some packaging. There are packages that of course help, there are already best practices. But, a fine work requires that primary research study.
We hope you enjoyed reading the interview with Pedro Del Carpio. In conclusion, the interview with Pedro Del Carpio, director of Heuristics Lab, provides insight into how behavioral biases impact work environments. From overconfidence bias to behavioral science-based strategies, this conversation provides valuable insights for promoting safe behaviors at work. The interview makes it clear that addressing unsafe behavior requires an approach based on understanding motivations and contextual influences, as well as implementing effective interventions based on behavioral science principles.
If you would like to explore these topics further and discover how to apply effective interventions, we invite you to watch the full interview on our YouTube channel. Don't miss it and join the passionate community on this journey towards understanding human behavior!
Let's stay close! Join our community at
Comments