How I Use Cognitive Frameworks to Navigate My Research Career
I did not plan the career I have. I trained as an electronics engineer in Chennai, expecting to spend my life near circuits and signals. Then I moved into audio signal processing, then into speech enhancement for hearing aids at Fraunhofer, and now I study the human brain with multi-parametric MRI in Erlangen. Each of those moves looked, from the inside, less like a strategy and more like a series of decisions made with too little information, some fear, and a stubborn willingness to be wrong in public.
What held the moves together was not a plan. It was a small set of thinking tools I kept reaching for, in the lab and away from it, until I stopped noticing I was reaching. This post is my attempt to name them honestly and show how I actually use them — not as a productivity system, but as the closest thing I have to a compass for a career that refuses to run in a straight line.
I have argued elsewhere that the thinking skills a machine cannot run are the ones appreciating fastest right now, in how AI makes critical thinkers more valuable, not less. That piece was about the market for judgment in the abstract. This one is more private. It is the same three or four frameworks, turned inward, applied to the one experiment I cannot hand to anyone else: my own working life.
The career as an experiment, not a ladder
The single most useful reframe I have found is to stop treating a career as a ladder and start treating it as an experiment. A ladder has a fixed shape; you either climb it or you fail. An experiment has a hypothesis, a method, an honest reading of the result, and — this is the part people skip — permission to be wrong without the whole thing collapsing.
When I left engineering for neuroscience, I did not know it would work. I could not know. What I could do was frame the move as a testable hypothesis: *the reasoning I learned about noise, systems, and evidence will transfer to a field I have never worked in, and if it does, I will be able to contribute within eighteen months.* That is a claim you can actually check. It has a timeframe, a success condition, and a way of being falsified. A vaguer version — *maybe I’ll be happier in science* — cannot be tested, so it cannot teach you anything.
I want to be careful here, because this framing is easy to oversell. Treating your career as an experiment does not make the outcomes controllable; it makes them *legible*. You still get rejected, funded, ignored, and surprised. What changes is that each of those results becomes data rather than verdict. A failed grant is not a statement about your worth. It is one data point about one method under one set of conditions, and the correct response to a single data point is rarely to burn the whole apparatus down.
EVIDENCE GRADE: MODERATE
I grade the career-as-experiment frame Moderate, and I want to say why plainly. There is solid evidence that construing setbacks as informative rather than final protects against the kind of rumination that derails people — the broader literature on how we appraise stress supports it. But the leap from “this helps in controlled studies” to “this is how you should run a fifteen-year career” is mine, not the literature’s. It is a frame I trust because it has held up in my own life, graded honestly as experience rather than proof.
Signal versus noise: what deserves your attention
The framework I use most, by a wide margin, is the one I brought whole from my engineering years: separating signal from noise. In signal processing this is literal. You have a recording, most of it is the thing you care about, and some fraction of it is interference you need to suppress without destroying what matters underneath. I spent a master’s degree doing exactly this for hearing-aid audio, and the instinct never left.
A research career throws an enormous amount of noise at you. Every week there is a new method being celebrated, a preprint that supposedly changes everything, a metric your institution suddenly cares about, a rejection that stings, a piece of praise that flatters. Almost none of it carries information about the questions I am actually trying to answer. The skill is not consuming more of it faster. The skill is building a filter that lets the small amount of genuine signal through and attenuates the rest — which is exactly the problem I unpack in signal vs. noise: a guide to information overload at work.
Concretely, my filter is a set of questions I ask before something earns my attention. Does this change what I would do next, or only how I feel about what I am doing? Is this a durable shift in the field or a fashionable spike that will decay in a year? Would I still care about this if no one else were watching? Most inputs fail these questions quietly, and letting them fail is the whole point. Attention is the scarcest resource I have, and a career is, in large part, the cumulative record of what you chose to pay attention to.
FIELD NOTE — ERLANGEN
My studies are small by design — often thirty to forty participants — because the cohorts I work with, chronic-pain-adjacent conditions imaged with MRI, are genuinely hard to recruit. Early on, that felt like a limitation I should apologise for. Over time I have come to treat it as a discipline that clarifies everything. With a small sample you cannot afford to chase every fashionable, data-hungry method; the noise would swallow you. So we choose simpler, interpretable analyses matched to the evidence we actually have, and we spend our real effort on design and sanity checks. That constraint taught me the career lesson more than any success did: match your confidence to your evidence, and most of the panic about keeping up with everyone else simply dissolves.
First principles: reasoning from the fundamentals when the map runs out
The second framework earns its keep at exactly the moments the signal-versus-noise filter cannot help — when there is no established path to filter, because the situation is new. For that I reason from first principles: I strip a problem back to the things I am confident are true, and rebuild from there rather than from analogy to what other people did.
Changing fields is the purest example I have. There was no template for “electronics engineer becomes brain-MRI researcher,” and the templates that did exist — get a second bachelor’s, start over from zero — assumed my previous training was worthless. First-principles thinking let me ask a better question: what is actually required to do this work, and which of those requirements do I already meet? Statistics, signal theory, programming, the discipline of designing a comparison that could prove you wrong — I had those. What I lacked was domain knowledge of the brain and a specific imaging toolkit, both of which are learnable. Reasoning from fundamentals turned an impossible-looking leap into a defined list of gaps to close. I have written the full method out in first-principles thinking: the engineer’s tool for future-proofing your career, because it is the tool I would hand a younger version of myself first.
EVIDENCE GRADE: STRONG
The narrow claim here is one I will grade Strong: reasoning from fundamentals rather than from analogy produces more robust conclusions when the analogy is weak or the situation is genuinely novel. That is well established across problem-solving and engineering-design research, and it matches everything I have watched in my own transitions. The caveat is equally strong — first-principles reasoning is slow and expensive, and using it for routine decisions where a good analogy already exists is just a way to waste your best hours. Reserve it for the moments the map runs out.
Hypothesis testing: how I actually make a decision
The third framework is the one closest to my daily work, and the one I am least willing to romanticise: hypothesis testing. In the lab, I do not decide whether an effect is real by how convincing it looks. I state what I expect, specify in advance what would count as evidence against me, run the test, and then read the result even when it disappoints me. The order matters. Deciding what would change your mind *before* you look is the only reliable defence against seeing what you wanted to see.
I run career decisions through the same loop, imperfectly. When I consider a new direction — a collaboration, a method, a project — I try to write down what I expect to be true if it is a good idea, and what I would observe if it is not, before I am emotionally committed. Then I treat the early months as the experiment and I hold myself to reading the result. The discipline is not in generating the hypothesis; anyone can have a hunch. The discipline is in specifying the disconfirming evidence in advance and then not flinching from it. This is simply thinking like a scientist applied to your own life, and it is the same habit I use every day at the workstation.
I fail at this regularly, which is exactly why I keep it explicit. The natural human move is to form a belief, invest in it, and then unconsciously reinterpret every result as confirmation. The pre-registered version of my own life — deciding the success condition before I start caring about the outcome — is the only thing that reliably catches me doing it.
Where the frameworks meet, and where they end
These are not separate tools I select from a menu. In practice they interlock. Signal-versus-noise decides what is worth thinking about at all. First principles handles the genuinely new problems where no filter yet exists. Hypothesis testing turns a chosen direction into something I can learn from instead of merely hope about. And underneath all three sits the same quiet value I brought from small-sample science: match your confidence to your evidence, and say so out loud.
I also want to be honest about their limits, because a post that sold you frameworks as a solution to a life would be exactly the kind of noise I am telling you to filter out. None of this removes uncertainty; it makes uncertainty workable. None of it guarantees the outcomes; it makes the outcomes informative. And no framework substitutes for the things that actually carry a research career — patience, colleagues who are generous with their time, and a genuine interest in the question that survives the years it takes to answer it. The frameworks are the compass. They are not the terrain, and they are not the reason to walk.
Try this today
Take one decision you are currently sitting on — a role, a project, a direction — and write it as a hypothesis instead of a worry. One sentence: *if this is the right move, then within [a specific timeframe] I will observe [a specific, checkable result].* Then add the harder line underneath: *and if instead I observe [this], that is my signal it was wrong.* You do not have to act today. You have just converted a vague anxiety into an experiment with a readable result, which is the difference between deciding and merely dreading.
What a decade of moving taught me
If I compress fifteen years of not planning into one sentence, it is this: a career you cannot control can still be a career you can reason about. The engineer who left circuits, the signal processor who left audio, and the researcher now reading brain scans are the same person running the same small set of frameworks on progressively stranger problems. The tools transferred because they live at a level the field changes never reached — which, not coincidentally, is the same reason I am not afraid of what better AI tools will make cheap.
I did not always trust this. There were years when the moves felt reckless and the uncertainty felt like failure. What I would tell the version of me standing at each of those thresholds is not “it works out” — I could not have known that, and neither can you. It is smaller and truer: you already own the instruments you need to navigate this. Point them at the right problem, read the results honestly, and let the ladder be a rumour. The experiment is the career.
If this way of thinking is useful to you, the natural next step is the toolkit this whole site is built around: [5 Mental Models to Future-Proof Your Career](/newsletter/) — the same frameworks I have narrated here, laid out cleanly, each with an honest grade of the evidence behind it. It is where the compass becomes something you can actually hold.
If this way of thinking is useful to you, the natural next step is the free guide this site is built around: [5 Mental Models to Future-Proof Your Career](/newsletter/) — five models chosen and stress-tested by a brain researcher, each with an honest grade of the evidence behind it. You’ll also get *Signal*, my monthly email: one idea from neuroscience you can use at work. No productivity spam, no AI panic.
*Mageshwar Selvakumar is a doctoral researcher in neuroscience in Erlangen, Germany, studying how chronic pain reshapes the brain using multi-parametric MRI.*