Does information help when you are anxious? Hell, no!
Hello, I make mistakes, nice to meet you!
I’m inherently an anxious person, so I do everything I can to get more information about a particular situation just to feel more in control. After all, we think that the more information (data) we have about a particular situation, the more informed our decision about it will be. For the longest time, I have been a person who tries to observe everything around me, gather as much information as I can about my surroundings and make very careful decisions, because there has been little room for mistakes. We’re part of a culture that loves to shame mistakes and bad decisions.
More information ⇒ the better is the decision ⇒ lesser chances of goofing up ⇒ false feelings of being in control of the situation ⇒ less anxiety
That’s why statistical tests fascinate me so much. When I started with data science, and specifically looked at statistical tests, I was like, “yay, something that can make me feel more certain about the decisions I am making and hopefully make me less anxious”. I was in for a surprise. Only recently, have I tried to put these thoughts into words, and now I’m putting it out here.
Disclaimer: I am not talking about any math in this post, it’s just ramblings.
I was watching a playlist on statistical thinking by Cassie Kozyrkov and came across some beautiful ways to make statistics comprehensive. But, the mind-boggling thing I realized is how statistics can never get rid of uncertainty for you. You’ll need just a tiny bit of context before you can understand what I’m saying later.
Statistics is the science of changing your mind.
— Cassie Kozyrkov
We use statistics to change our minds about something. We cannot change our minds unless we have a default first. In statistics, this is called a null hypothesis.
A hypothesis is some statement about how the world might work.
Null hypothesis: An action you will take by default or a belief you have by default.
Alternative hypothesis: An action you will take or a change in belief when you have enough evidence to reject your null hypothesis.
For instance, if I am choosing between ice cream and chocolate and I am inclined towards choosing ice cream, that is my null hypothesis. I won’t change my mind and reject the null hypothesis unless you give me enough proof that chocolate is a better option. Similarly, if I have no opinion whatsoever, any little bit of information should be enough for me to have a default — I don’t need statistics for that. Basically:
We are trying to find enough evidence that makes our null hypothesis sound ridiculous. [source]
Bayesian Brain
Bayes’ Theorem
There has been debate on whether our brains are Bayesian. Bayes’ theorem works by updating prior beliefs about something based on new information that you get. Here’s the math that you can choose to skip:
Some researchers believe that our brains might be working similar to Bayes’ theorem. If that is the case, there is a chance that our subconscious minds make decisions in this way. But a lot of studies also conclude that we might not have a Bayesian brain because we fail to comprehend certain probability problems which should be pretty clear to us had we been Bayesian — for instance, the Monty Hall problem.
Now, the point is, while Bayes’ theorem attempts to reduce entropy (uncertainty), you aren’t getting rid of it. So, even if your subconscious is making decisions that coincide with the Bayesian model, you will still not get rid of uncertainty. There is still probability involved. But, if my chances of goofing up reduce by even a tiny bit, I will bet on the test. I can get any amount of information at the moment, but I do not have enough information to predict the future with certainty. That is what I try to do by gathering all information I can. What do I end up with? Noise and information overload. So, how do I be more certain about the decisions I am making to reduce anxiety?
My big epiphany
Do NOT always seek more information (data).
I realize, that, in part, all I try to do when I am anxious is seek information so I can update my prior beliefs about the situation and make it as accurate as I can. I do this consciously. But the subconscious is much faster at doing this and decides for me before I know it. That’s gut feeling — you feel like something is right without knowing how.
Whether you’re doing a statistical test or are making decisions when you are anxious, gathering more information is not the best solution always. Sometimes, sure. But other times, the information I collect is just noise and is of no use or relevance to the decision I want to make. It does give a false sense of control and certainty — but for an anxious brain like mine, no amount suffices because that’s how the world works. You can never have ALL the information about anything because if you did, you would predict anything with certainty, in which case anxiety does not arise. But, with more information, I tend to join some dots that aren’t meant to be joined and ta-da! I conclude and have a false sense of control.
So, what do you do in such cases where information is not helping? You can always trust your gut! There’s loads of evidence about how your gut feeling is right a lot of times. It draws from all your past experiences, takes into account the current situation and before you know it, your subconscious has made a decision for you (again, this sounds similar to a Bayesian model). And now, imagine collecting all sorts of noise and trying to reject your default even when in this case the default is your best guess but you are trying to update the priors with what is seemingly noise! That sounds ridiculous. Yet, I have done that. I am guilty of not exercising the power of gut feelings. Though, it does take a lot of conscious effort to trust your intuition. But the more you trust it, the better it gets. Sometimes you have to let go and trust your intuition.
I have no sure shot ways of coping up with anxiety. But, one thing that helps is to be comfortable with the idea that you cannot know everything. You have the answers inside you a lot of the time — your seemingly Bayesian intuition.
Thoughts on the Bayesian brain debate
Humans could be Bayesian inference machines for the most part, as long as cognitive biases don’t overpower our statistical abilities to update our prior beliefs once we get new information. We combine information about our past experiences (priors) with an estimate for how probable the current circumstance you are trying to predict is. Your intuition is doing that for you already. As I mentioned, some studies show our brains might be Bayesian. At least that’s what researchers want to believe about how our brains should work because after all, Artificial Intelligence is modelled after human intelligence using Bayesian statistics. That should be a starting point in thinking that our subconscious decision making is Bayesian.
It’s okay: you can’t have all the information in the world because it’s impossible to have it!
I am just trying to make a case about being more comfortable with:
- Not knowing everything
- Knowing that it’s impossible to know everything
That’s why you should trust your gut more often, especially when you are anxious. More information will give you a false sense of reducing uncertainty and probably cognitive overload. Being more comfortable with the idea that your subconscious is making decisions for you when you don’t know how it is will make you better at trusting your intuition. Because, after all, if you are doing a statistical test outside, there’s a fair chance that so is your subconscious. It’ll help to leave decisions to your intuition sometimes.
