Arriving at knowledge: Why we should seek to falsify and not confirm your beliefs
Confirmation bias is a term that describes our collective tendency to seek and interpret evidence in ways that are partial to our existing expectations, beliefs, and hypotheses. In other words, we love to confirm what we believe, while we hate to consider opposing viewpoints. Thousands of scholars have examined the phenomenon of confirmation bias and found it to be robust, prevalent and powerful (Nickerson 1998, Jonas et al. 2001, Kahneman et al. 2002). What’s more, the world of the internet and especially social media appears to drive polarization. Online users are very likely to “select claims that adhere to their system of beliefs and to ignore dissenting information. Confirmation bias, indeed, plays a pivotal role in viral phenomena. Furthermore, the wide availability of content on the web fosters the aggregation of like-minded people where debates tend to enforce group polarization” (Del Vicario, et al., 2017).
In this article I will not attempt to persuade you of a particular point of view. In fact, I am convinced that there are very few things of which I am sure, and nothing of which I am certain. Instead, I want to argue only that searching for evidence that falsifies a given claim is much more informational than evidence that confirms a claim. The scientific method is based on this premise, and it useful to apply it in everyday life, if you want to understand the world as it is.
Six years ago, Derek Muller (who is the YouTuber behind the science channel Veritasium) convinced me of this argument with a five-minute video entitled “Can you solve this?”. I recommend you watch it here. For those of you who did not watch the video, I am going to paraphrase Derek’s argument here. Imagine a sequence of numbers: 2, 4, 8 … and so on. Now, I have a rule in mind to which these numbers conform, and I am going to ask you to identify it. In order to determine the rule, you must provide me with a sequence of three numbers, and I will tell you either:
1. It conforms with my rule, or
2. It does not conform with my rule
Your goal is to identify my hidden rule. Most people will instinctively ask questions that confirm their hypotheses. For example, on first glance, I believe that the sequence should continue: 16, 32, 64 etc. In other words, my hypothesis was that the rule is doubling the number before and I sought to confirm this. However, this is not my rule. We tend to ask things like: “How about 100, 200 and 400? Or 10, 20, 40?” The answer? Yes, these follow my rule. But here’s the rub, by merely confirming a given hypothesis (in this case that the numbers double) gives you no new information. You know that these patterns follow the rule, but the rule is not to double the last number, so what is it?
At this point, you might start to think. What about 1,2,3, does this follow the rule? The answer is yes again. This, though, is a better question, because now you know that doubling isn’t the rule, it just fits within a broader rule. The best thing you could do to arrive an answer here is to disprove your initial hypothesis. For example, now we have established that 1,2,3 also follows the rule so we know:
1. The rule is not doubling and
2. The numbers are positive
3. The numbers are ascending
Let’s narrow down the rule further. Let’s try and disprove points 2 and 3. Does -3,-2,-1 fit the rule (negative numbers ascending). Yes they do. So we know the numbers do not have to be positive. Perhaps the rule has to do with ascending numbers. So let’s ask: how about 3,2,1? Do these fit the rule?
Eureka! The rule I was thinking of was that the numbers were ascending!
The moral here? If you are looking to understand something better, ask questions that may disprove your current theory. By doing this, you are going to be looking for the right information, and you may just outperform the herd, who tend to confirm their current beliefs repeatedly.
If you are convinced that office politics determines your career trajectory and there is nothing you can do about it, test this hypothesis by experimenting with different tactics. Simply ask your boss for a raise and if she refuses, ask “how can I earn a raise in the next 6 months?”
If you are convinced that working insane hours is the key to success, ask: “How could I work half-days and still make a good living? What would I have to do to achieve this?”
Looking to falsify, rather than confirm, claims is highly informational. It has the added benefit of generating solutions that most people aren’t looking for. Why? Because most people aren’t asking the right questions.
Del Vicario, M., Scala, A., Caldarelli, G. & Stanley, E. H. Q. W., 2017. Modeling confirmation bias and polarization. Scientific reports, Volume 7, p. 40391.
Jonas, E., Schulz-Hardt, S. F. D. & Thelen, N., 2001. Confirmation Bias in Sequential Information Search After Preliminary Decisions: An Expansion of Dissonance Theoretical Research on Selective Exposure to Information. Journal of Personality and Social Psychology, 80(4), pp. 557-571.
Kahneman, D., Griffin, D. & Golovich, T., 2002. Heuristics and Biases: The Psychology of Intuitive Judgement. 1 ed. Cambridge: Cambridge university press.
Nickerson, R. S., 1998. Confirmation Bias: A Ubiquitous Phenomenon in Many Guises. Review of General Psychology, 2(2), pp. 175-220.
Veritasium, 2014. Can You Solve This?. [Online] Available at: https://www.youtube.com/watch?v=vKA4w2O61Xo [Accessed 15 June 2020].