Most Popular

    Sorry. No data so far.

Winner of Amgen Patients | Choices | Empowerment Competition Emerging Star of HealthCare Engagement Award
Mayo Clinic Award - LeftA winner of the Mayo Clinic iSpot Competition for Ideas that will Transform HealthcareMayo Clinic Award - R

How To Run a Successful Self-Experiment

New relationship = more sex, new baby = more weight gain (Photo credit: juhansonin)

“Whatever course you decide upon, there is always someone to tell you that you are wrong. There are always difficulties arising which tempt you to believe that your critics are right. To map out a course of action and follow it to an end requires courage.” - Ralph Waldo Emerson

Say you’ve chosen to try a new treatment or new intervention for your health. How do you really know how well it is working?

The key is to run a self-experiment, to have the numbers tell you what’s happening. This is easier to do with some treatments than others, but here are 5 general principles that will help you do a successful self-experiment:

1. One test at a time. Keep it simple. Don’t try to test too many new variables at once, or your results will not be meaningful. Change one thing and one thing only, trying to keep everything else the same, and see what happens.

2. Define your outcome variables. What do you want to measure? If you want to see if something is affecting your mood or sleep, track how happy you are or how many hours you sleep. If your new treatment is to help with pain, track pain levels. Try to keep the things you measure to a minimum too – identify the one or two things you really want to change and see if they are changing.

3. Track these over time (longitudinally). One day of data isn’t enough to see how well something is working. Depending on what you’re measuring, you might need a few weeks of data to start seeing patterns. Or you might notice a change sooner than that. Ideally, start measuring yourself BEFORE you start the treatment, so you’ll have a baseline measurement to compare your new results to. For example, if you want to lower your blood pressure, start by knowing what your blood pressure is, then you’ll know if it’s improving or not.

4. Do an N=1 randomized controlled trial on yourself. Here’s where you can start to get fancy if you want more scientific results. CureTogether will soon be able to set up a randomization schedule for you, where you take a treatment some days and a placebo on other days. This might not work for some medications where not taking them would mean severe withdrawal or other side effects, but would probably work fine for things like lifestyle or dietary interventions

5. Self-blind your study. For uber-geeks, you could even go so far as to create your experiment so that you don’t know whether you’re taking a treatment or not. This would eliminate any placebo effect and give you the most definitive results from your self-experiment.

On the other hand, if this is getting complicated, watch Professor Seth Roberts’ “Stop Worrying and Start Experimenting” talk on experiment design and keeping it simple.

Seth Roberts on experiment design from Kevin Kelly on Vimeo.

Once you have your data and have analyzed your results, now it’s the time to make decisions on how to continue your course of treatment.

If it’s working, keep going!

If it’s not, rinse and repeat your self-experiment with a new intervention.

At the same time, keep in mind that our decisions can often be biased, which will be the topic of our next blog post, “Are You Biased? How To Make The Right Decisions”.

Related Posts

Feel Better – How To Find A Treatment That Works

Do You Have A Condition? How To Find Out

Getting Diagnosed – How To Choose The Right Test

You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

AddThis Social Bookmark Button

7 Responses to “How To Run a Successful Self-Experiment”

  1. Knowing whether or not you’re taking a placebo is kind of cheating, isn’t it? Are there any studies on the influences of placebo knowledge versus relative unawareness of taking a placebo?

  2. Great point, Bryan! You could test both ways – knowing when you’re taking a placebo or having a loved one prepare your treatment for you so you wouldn’t know if it was a placebo or not.

    This kind of keen inquiry is essential to both designing a self-experiment and to analyzing its results. Bravo!


  3. [...] it is time to talk about the power of conversations informed by data, the self-experimentation that people engage in, the impact of real-world experience on drug safety, and of course, [...]

  4. [...] it is time to talk about the power of health conversations informed by data, the self-experimentation that people engage in using information often found online, the impact of real-world experience on [...]

  5. [...] self-experimentation to aug­ment and sup­ple­ment his lab research. Addi­tion­ally, check out How to run a suc­cess­ful self exper­i­ment from the clever, crowd­sourced health site,, which is a great exam­ple of a [...]

  6. [...] CureTogether Blog » Blog Archive » How To Run a Successful Self-Experiment (tags: health science self-experiment self-experimentation) [...]

  7. > having a loved one prepare your treatment for you so you wouldn’t know if it was a placebo or not.

    It’s pretty easy to randomize for yourself (I always wonder why people don’t do it by default, it’s not much work at all); for example, you can put 2 pills into a cup, shake it, and take one blindly. After the final measurement, you look into the cup and obviously whichever one is missing is the one you took.

    (Or you could put all the pills into a cup or jar or bag, but then it gets a little more complex – )

Leave a comment or question