04 February 2019

My Learning - Part 12 - Experiments


  • Experiments are great at highlighting incorrect assumptions I've made.
  • The experiences/learnings I get from experiments "stick".
  • Experiments tend to be great conversation starters, which leads to more people, which leads to... well, I think you've heard me say that enough times now.
  • I love exploration and experiments are certainly an act of exploration which makes them fun!


I'd say there are two categories of experiments I conduct:

Experiments testing a hypothesis and experiments where I don't have a clear expectation, the activity I test simply seems useful and I want to learn more about it.

Example with hypothesis:
I had the hypothesis that when I'm programming very little time is actually spent on writing the final production code. So I picked a specific feature I had programmed under pretty optimal conditions (easy, clear purpose, minimal dependencies etc.) and simply rewrote that code, character by character, into a new file and compared the time between the two activities.

Example without hypothesis:
What happens if I start to get up at 5 AM every morning? How will my body feel after a while? I tried different activities in the morning, I tinkered with different times to go to bed, I altered the time to 4 AM and 6 AM etc. all this while monitoring my mood, activities performed etc.

A few more examples:
  • I've experimented with several different ways to take notes
    • Sketchnoting
    • Mind maps
    • Different pens
    • Digital notes
    • Using colors in different ways
    • Different notebook sizes
    • Different types of paper
  • Different ways to read
    • Speed reading
    • Various memory techniques connected to reading
    • Writing summaries for each chapter I complete
    • Jumping back and forth between a couple of different books hoping that the variety would make reading more fun (it didn't work...)
    • Read super carefully by looking up any word I didn't fully understand, reread sections until I could correctly recall what was in there etc. (worked even worse)
    • Only read books with 99 pages or less. The idea being they would better match the way I like to read.
    • Basically everything mentioned in Brandon Hakim's excellent Udemy course about reading.

What makes it an experiment

At this point it may sound like I call any variation to my approach an "experiment" and to some degree that's true. However, what makes me call these "experiments" is there's a very deliberate method and target with each of them. Before I conduct an experiment I carefully decide what it is I want to try (an hypothesis or an approach that seems reasonable), how I can isolate that activity and how I will evaluate the outcome(s).

So for instance I wouldn't call an exercise someone gives me an "experiment" because I'm not thinking and designing that activity for myself, I simply follow instructions (well, I don't but still...). The same goes if I've forgotten my note taking material and "just take notes" without really having a reason to why I do it the way I do it. This can still teach me something but I wouldn't call it an experiment.

How they're initiated

Often when I learn something I want to test how well that thing works for me. Rather than "just doing it" I often try to prepare it like an experiment. For instance before I coached using the NOPRA model I asked myself: "What is it I want to achieve? Why? And how can I evaluate the result?". I then design the actual activity and decided how I would evaluate it.

Another scenario is when I'm in the middle of doing something and realize I'm doing it slightly different than usual. If I'm in the right mood I can stop myself and say "this is interesting, how can I monitor and evaluate this?" (like the dish washing experience described in part 2).


There's no formal process where I write things down, remove all external distractions etc. so few, if any, of my experiment would survive an academic review.

I'm also well aware of e.g. confirmation bias, the Hawthorne effect and sampling bias and that all these will skew my results. So apart from considering this when I design my experiments, which is obviously not enough, I try to also be careful about what I make of the results, especially when I present them to others... I may have gotten carried away more than a few times though.

That being said; the experiments are usually quite harmless and their goal is mostly to help me understand how something works for me, and for that I think my approach is perfectly sufficient to be honest (ignorance is bliss).


No comments:

Post a Comment