img_0255-bannerpng

N=1 experimentation is research. Research that focuses on one individual for health improvement purposes.

That’s what ‘n=1’ means: an experiment with a sample size of one.

This blog post comes with a warning label: it gets an 8 out of 10 on the official (but unvalidated) Biohack U nerdiness scale.

Validity & Bias

Whenever we do research, we need to consider validity and bias. So, let’s do that now:

Validity

Validity is about whether a tool measures what it is supposed to measure.

Bias 

Some bias is caused by errors in n=1 research methods, but not all.

It is impossible not to be biased.

We do everything we can to remove bias in our research, and then acknowledge what remains. Bias can occur at any stage of the n=1 process: planning, data collection, analysis or reporting. It results in inaccurate findings.

How to Address Validity

If we had a zillion dollars and could hire a team of hot-shot researchers to help us run your n=1, they could probably do away with all of the threats to validity in your experiment.

Those researchers could all wear white lab coats (or magenta ones, if you preferred) and they would ensure that all the instruments used for measurement were carefully constructed and tested, and that they were applied in a standardized way according to exacting pre-defined specifications.

That might be awesome. That experiment would probably generate precise results.

But most of us are just doing the best we can. We don’t wear lab coats when we conduct our n=1’s. Most of us are just happy if we can find socks that match that don’t have holes in them. Nevertheless, we can still be rigorous in our approach to n=1 research.

BE proactive

Get new posts by subscribing to our newsletter

Thank you! You're on your way to a better life

Please check your inbox for an email from us. Don't forget to click yes to confirm your subscription.

Take the Assessment

Thinking about validity is the first step in addressing it.

Threats to Validity

Some of the threats to validity we need to take into consideration include:

  1. History: Factors that may be influencing our results that we haven’t considered and aren’t measuring. In this case, we may incorrectly attribute results to the stuff we did in our n=1, when actually they are caused by ‘confounding variables’ (i.e. something else).
  2. Maturation: Sometimes things just happen because time has passed, not because of our experiment. Over time, babies learn to talk. People get hungry. People get tired. People get cold. People get bored.
  3. Instrumentation: If a measurement tool is changed during an experiment, the results lose validity, because we aren’t comparing an identical measure before and after the experiment.

It is important to acknowledge threats to validity. Most experiments have them. Eradicate as many as possible, and then just keep the rest in mind. When you report your results, tell people what they are and what you did about them.

Face Validity

In n=1 experimentation we can use what is called ‘face validity’ (admittedly the least rigorous kind).

Face validity is about using common sense. We think carefully about the tools we want to use, and if it seems to us like the instrument measures what it is supposed to measure, then we proceed.

In n=1 experimentation, we are free to use any measurement tools we like, including ones that we create, or ones that other people make and share like Danielle LaPorte's Soulful Habit Tracker, or ones that have already been validated through prior research like the Rosenberg Self-Esteem Scale or the Mindful Attention Awareness Scale.

How to Address Bias

In the old days, researchers were supposed to be objective and unbiased. That reflected the belief that it was possible to be unbiased and objective. It also suggested that there is an objective and verifiable reality that we can observe.

This is where we start straying into research philosophy (otherwise known as epistemology: one of my favourite things!), but I will not geek out on that in this post.

Instead I’ll just quote Michael Quinn Patton, who asserts that objectivity is “impossible to attain in practice and of questionable desirability”. He suggests that researchers seek "honest, meaningful, credible, and empirically supported findings” instead.

We can address bias by staying rigorously open-minded about what we observe during an n=1 experiment. We don’t set out to prove anything: we merely to test an idea.

Our goal is healing. To heal, we must learn about what works and what doesn’t work for us, as individuals.

To learn what works, we need to understand what is. Some bias is always present, Our job is to consider how it is affecting our perceptions.

A conscientious approach to n=1 experimentation enables us to address bias. This includes:

  1. Being systematic Matthew and I can help with that. Our first program, an n=1 workbook kit, is launching September 2nd and it will only cost $24.95.
  2. Using triangulation Seek a few (aim for three) separate, valid perspectives throughout your n=1. Triangulation counteracts bias by engaging multiple theories, multiple methods, multiple sources of data, and/or multiple investigators. Triangulation is a core element of the Biohack U n=1 method, and is covered in our soon-to-be-launched workbook kit.
  3. Considering rival explanations Sound research always considers alternate explanations for what is being observed, to help reduce the chance that our biases and assumptions are colouring our findings. Rival explanations may include the ‘confounding variables’, mentioned above. The point is to learn, not confirm.
  4. Understanding complexity We are complex systems living in complex systems. Disequilibrium in one system can affect other systems, and that imbalance can spread. This interconnectedness can make it difficult to discern the root cause of symptoms. It can be hard to figure out cause and effect. 'Ambiguous temporal precedence’ is the term for that,  and it’s another threat to validity.
  5. Acknowledging biases A core practice of the n=1 method is to be aware of our biases and how they may be affecting our experiments. Acknowledging bias when reporting results is important.

In n=1 experimentation, the researcher is usually also the subject of the research, leading to interesting considerations in the area of bias. I'll write more about that soon.

Featured Resource

Real Plans

Up your nutrient density: a perfect n=1 experiment.

Real Plans offers Gluten-Free, Dairy-Free, GAPS, Paleo, Primal, Keto and Autoimmune Protocol options. 

Start with a plan. Measure your results. 

Easy to use and completely customizable, Real Plans helps you fit your n=1 lifestyle into your life.

You May Also Like

Biohacking Update: 28 months on the AIP

Matthew returned to work this week. When he took disability leave in 2013 we thought he’d never work again...
Read More
17 Apr 0 comments

Biohacking Update: 3 months on a yeast & biofilm busting protocol (& 25 months on the AIP)

Within a few weeks, Matthew was able to reintroduce some carbohydrates. Broccoli. Broccoli is amazing when you've been eating only well-cooked spinach and yellow zucchini for months...
Read More
31 Jan 0 comments

Online Community (for healing)

When we commit to healing, we need to surround ourselves people who ‘get it’. Increasingly, we find those people online. Eight people who have used online community for healing share their experience in a series of super-short videos...
Read More
03 Feb 0 comments

How to find the answers to your health questions. Online.

Valuable health information, even life-saving information, can be found online. Being intentional and systematic can help you find it...
Read More
26 May 0 comments

Click to close

Subscribe to our mailing list