How to Deal With Impostor Syndrome
Table of Contents
Everyone else is so much more knowledgeable and skilled.
I’m a fraud and people will laugh at me and fire me when they find out.
Except for literally that one1 best data scientist in the world, we are all in the same situation: there are better data scientists out there. Being self-aware, we must all live with that knowledge.
But that does not mean you are not providing value. In fact, if you are applying the defining skill of a data scientist, the scientific method, then you are doing the single best thing that can be done to uncover whatever value is to be had in that situation.
As with most things, there is a difference between knowing something intellectually, for example, reading it on a blog, and knowing it in your heart. And the fastest way there is through your own experience.
What concrete actions can you take to alleviate feelings of not being good enough, not being ready, of being a fraud, sometimes summed up as impostor syndrome?
Here are those I have found useful.
Work with others (or not!) #
When you work with people with different backgrounds and skills, it is much easier to see how each person brings something to the table. While you may still know woefully little, what you do know is valuable to your colleagues. Getting to know them can in turn let you see how, while you may struggle with feelings of not knowing enough, it’s nothing compared to how they feel intimidated by you and your role.
Sometimes, your environment hurts rather than helps your ability to deal with these feelings. If you work in an environment where people are unsure of themselves, and, try to prove themselves by one-upmanship and competition, it might be best to seek out a better environment.
Don’t rely on other people’s standards #
One place where feelings of inadequacy can become acute and hamper you is applying for jobs. We have all seen the job postings requiring 10 years experience with LLMs, and heard the stories about library authors getting rejected for lack of experience with said library. Employers will ask for the sun and moon in job postings.
Don’t measure whether you should apply to a job based on the requirements, measure it on whether you would want the job2.
Practice practice practice #
Have you ever felt nervous taking a test, doing a technical interview, or making a speech in front of others? Now have you ever felt that nervousness dissolve into calm confidence as you realize that you have it all under control, that you are so supremely prepared you can handle anything thrown at you, that you can even have a little fun with it?
I’ll admit this is rare for me, but every time it has happened it is because I have been very well prepared.
In my experience, the most concrete action I have been able to most easily take myself to alleviate these feelings is to practice, practice and practice precisely those things I felt unsure about. If you’re worried about window functions (yes I practice those too every time I go on the market), pick a common engine (e.g. Postgres) and practice them until you know them fluently.
If you’re worried about a demo or presentation, practice it and hone it down to the word and second. Any time you stumble on a phrase or transition, hone and revise it until it flows smoothly. Then once the moment comes you’ll know it so well you’ll be able to get your message across in a loose, relaxed yet perfect way.
Do all practice without AI.
In my experience this type of practice is of course great for achieving your immediate goals such as passing the test or getting the jobs. But more importantly, nothing boosts my confidence more than that feeling of high performance under pressure that being well prepared provides.
Lastly, but just as importantly, recognize that chances are, most people have even less of a clue.
Notes #
Yes, it is of course possible that more than one data scientist could be exactly and equally best. And, since data science requires several different skills, there could be more than one dimension. I consider those edge cases that only data scientists would think of and go to the footnotes to look up :). ↩︎
Yes, of course, you should objectively assess how you measure against the requirements for jobs that you want, and, then work systematically to improve yourself until you fulfil those requirements. But if you are worried about feelings of inadequacy and impostor syndrome, you are almost certainly already doing that. ↩︎