Self-experimentation is a term the online Primal community regularly bandies about. I’ve been meaning to write a post on the subject, and I figured the first week of this year’s Primal challenge would be the perfect spot to drop it. Because, after all, those who accept and undertake the 30-day Primal Blueprint Challenge  will essentially be conducting a 30-day self-experiment on themselves. It won’t be your first self-experiment, nor will it be your last, but it may be your first chance at knowingly conducting one.
Yeah, we’re all lifetime self-experimenters, when you get down to it. From infancy onward, we conduct experiments – most of them totally informal – to understand how the world works and how to interact with it. A toddler trying avocado is testing whether it tastes good and nourishes, the teen using a cheesy pickup line is testing whether it gets the girl’s number, and the college freshman pulling an all-nighter before a midterm is testing whether she can party all quarter and still make grades. They’re all forays into the relative unknown, and they’re all crude, imperfect modes of self-experimentation, even though the experimenters probably aren’t consciously aware of any experiments being conducted. Life is full of these informal little tests.
So we have a legacy of experimentation.
But it’s hard to derive lasting value from informal, incidental, everyday self-experimentation. No, if you really want to step up your game and approach discovering cause-and-effect from self-experimenting, be a little more rigorous with your approach. Think of the experiment as an actual scientific experiment, with observation and hypothesis and testing and even recording of data, controlling of variables, and randomization. You’re probably familiar with Seth Roberts, who I’ve referenced before and who gave a great talk at the AHS about his self-experimentation with flax, pork fat , and butter  (check it out  if you haven’t). Seth quantifies the results of his tests. He measures, records, tracks, and graphs. It sounds rigorous but he keeps the design of his experiments very simple and because of that, his results are fairly conclusive. As a recent Chris Masterjohn post  explains, self-experiments conducted in this fashion are necessary to conclusively identify cause-and-effect, but they may not be required for the average Joe who wants to figure out how he feels trying something different.
There’s another way, one that doesn’t require that you bust out charts and stats and take measurements if you don’t want to. For the average person who just wants to play around with new stuff, I like what I call “soft self-experimentation” because it doesn’t require agonizing rigor to obtain useful information. You just do something different – add a food, remove a food, go for morning walks, lift using unilateral movements instead of bilateral movements, whatever it is – and “see how you feel” while trying to keep the other stuff constant (don’t change your sleep habits if you’re trying to test how you tolerate starch). My friend Richard Nikoley  of Free the Animal is an expert at this (and gave his AHS talk  on the subject) kind of intuitive experimentation. You won’t really hear him talk about numbers (besides maybe pounds lost and lifted), but he does account for confounding variables. Perhaps the biggest potentially confounding variable is bias. That is, your own bias. You have to be honest with yourself and be open to falsifying your hypothesis. In fact, you have to actively look for falsification. Sometimes (heck, many times) you will be wrong and it won’t work out. That’s okay, because that’s how you learn. “No failure, only feedback,” as Art Devany says.
I mean, you’re busy people with families, jobs, and hobbies, and the last you want to be doing is submitting your little self-study about cold showers and fat loss to peer review before you can draw conclusions, or run numbers and graphs. Heck, you barely have time to keep up with the blogs, let alone follow all the references to studies telling you how changing a variable affected a group of people who were not you. And that’s the rub, isn’t it? The scientific literature, however crucial for furthering our understanding of the way the world works, tells the story of other people. Other individuals. PubMed  abstracts or full texts you read on your laptop cannot and never will compete with the primacy of actual self experience, because that’s how biological organisms work. The thing that happens to you carries more weight than the thing that happens to another (let alone a nameless, faceless stranger).
And so we experiment on ourselves to figure out what works.
For guidance on how to set up a self-experiment and track your results, check out Seth’s advice .
A few more tips:
- Keep it simple. Stick to changing one thing at a time, and make it easy to follow.
- Be specific. Don’t test how “fruit” affects you post-workout. Test how “bananas” or “mangos” or “blueberries” affect you.
- Kill multiple birds with one stone, if possible. If you’re testing how post-workout bananas affect fat loss or gain, consider observing other possible effects, like improved performance or blood sugar readings. You don’t have to, but keep an open mind.
- Start with a sensible premise drawn from reliable sources. Like your own experience, another’s, or results from clinical studies. An example: you forget to eat breakfast and feel stronger during the afternoon lifting session. Was it the lack of breakfast? Maybe; test it.
- Eliminate outlandish premises. I wouldn’t advise testing whether the trans-fats in Crisco are actually harmful by eating a tablespoon every morning, and breatharianism  almost assuredly won’t work; the literature is quite consistent.
- Be prepared to discard a hypothesis. Things won’t always work. And things that seemed to work for a while might suddenly stop working. Or, things that seem to work won’t actually work, or they’ll be working for entirely different reasons than you supposed. In other words, you’re probably never going to know with absolute certainty that something is working for the reason you thought it was. Just be ready to ditch failed hypotheses and change them on the fly.
- Don’t extrapolate to others. Just as your experiences didn’t jibe with results from randomized controlled trials, the solutions you discovered from your own experiments may not always work for other people. They’ll have to test it on themselves.
- The beauty of the self-experiment is that it acknowledges the persuasive power of personal experience. Ultimately, we are all interested in leading healthy, happy, and productive lives, so we want whatever works. And while studies on other people are valuable, your own reaction to grass-fed butter  always trumps what your cardiologist says the studies about other people’s reactions to butter are. The working mother of three, for example, isn’t going to truly know what works for her until she actually tries some stuff out. Conducting a self-experiment, even if it’s totally casual and would never pass the muster of peer review, gives that mother valuable insight into what works and what does not. And it’s easy, free, and gets the mind working.
The possibilities for experiments are virtually endless…
- The effect of sleep duration on next day sprint/lifting/running performance.
- Tolerance of Primal starch source (choose one).
- The effect of evening laptop abstinence on various markers (sexual performance, sleep duration, next day mood).
- The effect of a daily “forest bathing ” hour-long session.
So, readers, tell me: what self-experiment are you going to conduct?