N-of-1 Self-Experiments: Using Daily Cognitive Scores to Test What Actually Affects Your Brain

The Problem with Population Averages

Most health advice is derived from randomized controlled trials studying hundreds or thousands of people. These studies are good at determining whether an intervention works on average. But they can't tell you whether it works for you. The average effect of caffeine on reaction time tells you nothing about how your third cup of coffee at 2pm affects your specific cognitive throughput at 4pm.

This is the fundamental limitation that N-of-1 experimentation addresses. An N-of-1 trial is a structured experiment conducted on a single person — you — where you systematically test an intervention against your own baseline. The method has roots in clinical medicine, where it's been used since the 1980s to determine optimal treatments for individual patients with chronic conditions. But with the rise of wearable sensors and the quantified self movement, the approach has expanded far beyond the clinic.

As a 2016 framework paper published in the Journal of the American Medical Informatics Association (Karkar et al.) put it, N-of-1 experimentation bridges the gap between average treatment effects for groups and specific treatment effects for individuals. You become both the researcher and the subject, and the data you collect is immediately actionable because it's about your body and your brain.

What You Need for a Valid Self-Experiment

Self-experimentation without structure is just guessing with extra steps. A valid N-of-1 experiment requires four things: a consistent outcome measure, a clearly defined intervention, a baseline period, and enough data points to distinguish signal from noise.

A consistent outcome measure is the foundation. If you're testing whether a supplement improves your cognition, you need a metric that captures cognitive performance reliably, at the same time each day, under comparable conditions. Subjective "I feel sharper" doesn't cut it — your perception of sharpness is contaminated by mood, expectations, and placebo effects. You need a number. Something like a daily Sharpness Score — a short, standardized assessment that measures your working memory and processing speed against your own rolling baseline — gives you the kind of quantified signal that makes self-experimentation meaningful.

A clearly defined intervention means changing exactly one variable at a time. If you simultaneously start a new sleep protocol, a nootropic stack, and a morning exercise routine, you'll have no idea which variable is driving any changes you observe. Isolate the variable. Test one thing.

A baseline period of at least two weeks gives you stable data to compare against. Without a baseline, any improvement you see could simply be the normal upward trend of getting better at the test through practice.

Enough data points means running the experiment long enough for patterns to emerge. Daily cognitive scores are noisy — you'll have good days and bad days regardless of the intervention. A minimum of 14 days on the intervention, compared to your 14-day baseline, gives you enough data to start making reasonable inferences. Longer is better.

A single day's score tells you almost nothing. Fourteen days of scores start to tell a story. Sixty days of scores can change what you believe about your own brain.

Five Experiments Worth Running

1. Sleep Duration

Track your Sharpness Score alongside sleep duration for 30 days. Don't change anything about your sleep — just observe. Most people discover their cognitive performance correlates more tightly with sleep than they expected, and that the relationship isn't linear. Six hours might produce a score indistinguishable from seven hours for you, while dropping below six creates a measurable cliff. The specific threshold is personal, which is exactly why you need your own data.

2. Caffeine Timing

For two weeks, take your Sharpness Score before your first coffee. For the next two weeks, take it 30 minutes after. Compare. Many people assume caffeine improves their cognitive performance, but the data sometimes tells a different story — particularly if you're testing in the morning, when cortisol is naturally high and caffeine's additive effect may be smaller than you think.

3. Exercise Timing

Compare your cognitive scores on days you exercise in the morning versus days you don't. Acute exercise has a well-documented positive effect on executive function in research, but the size and duration of the effect varies enormously between individuals. Your data will tell you whether morning exercise is cognitive fuel or fatigue for your specific brain.

4. Supplement Verification

If you're spending money on nootropics or cognitive supplements, run a structured test. Two weeks off the supplement (washout period), two weeks on, two weeks off again. Compare your average Sharpness Score across each period. If the supplement is actually affecting your cognitive throughput, you'll see it in the data. If it's not, you just saved yourself $50 to $200 per month.

5. Screen Time Before Bed

For two weeks, stop all screens one hour before sleep. For two weeks, use screens normally. Compare your next-morning Sharpness Scores across both periods. The research on screen time and cognitive function is mixed at the population level, but your personal data may show a clear pattern — or it may not. Either way, you'll have an answer instead of an assumption.

How to Read Your Results

The most common mistake in self-experimentation is over-interpreting small differences. A 2% change in your average Sharpness Score between conditions is probably noise. A 10% change sustained over two weeks is a meaningful signal worth paying attention to.

Look for consistency, not outliers. If your score is higher on 11 of 14 days during the intervention period compared to your baseline average, that's a pattern. If it's higher on 8 of 14 days, that's ambiguous. If it's higher on 6 of 14 days, the intervention likely isn't doing anything.

And be honest about confounders. If you started the supplement the same week you got a promotion, your improved scores might be driven by mood, not the supplement. The beauty of structured N-of-1 experiments — particularly the washout design where you alternate on and off — is that confounders tend to average out over time in a way that unstructured observation never achieves.

Common Pitfalls

The biggest threat to a self-experiment isn't bad data — it's premature conclusions. Running a supplement for five days, seeing a 3% Sharpness Score increase, and declaring it effective is not science. It's confirmation bias with a spreadsheet. Five days is noise. Fourteen is the minimum for a tentative signal. Thirty is where confidence begins.

The second pitfall is failing to control for the obvious. Starting a new supplement the same week you also changed your sleep schedule, began exercising, or had a major life event renders the experiment uninterpretable. This is why the single-variable rule exists. Change one thing. Measure the result. Then change the next thing.

The third pitfall is abandoning the experiment when the early results aren't exciting. If your data shows that a supplement has no effect, that's a finding — a valuable one, since it means you can stop spending money on it. Null results aren't failures. They're the whole point of running the experiment in the first place.

Your Data, Your Discovery

The quantified self community has a saying: "self-knowledge through numbers." But knowledge requires more than collection — it requires structure. Tracking your sleep, your supplements, and your exercise is useful. But tracking those variables alongside a daily cognitive output metric transforms raw data into personal experiments that can genuinely change how you manage your own brain.

The Sharpness Score was designed for exactly this kind of use. It's personal-baseline-relative, it's honest about day-to-day variation, and it doesn't inflate over time through gamification. That makes it a reliable dependent variable in your own N-of-1 experiments — the closest thing to a personal cognitive lab that fits in your pocket.