Matthew Cornell

Recently I've been having trouble falling asleep, and it's making me tired during the day. I'm trying some experiments in Edison to figure out what the causes might be (improve sleep hygiene and try melatonin), but it's difficult pinning it down. Why is that? Here are some thoughts. Please tell me what you think!

The human body is a complex system, and complex systems have many variables. Our goal is to understand the system by teasing out the variables that matter and then identifying patterns. Following are some possibilities for sleep in particular. (Note that the default hypothesis when starting it is to ask for each variable, "Is there a significant relationship between this and the problem?")

• Taking naps
• Exercising right before bedtime
• Eating right before bedtime
• A disorder like Bipolar (I have type II)
• Doing a stimulating activity before bedtime (e.g., working, watching a thriller, or spending time on the computer)
• Taking a substance with insomniatic side-effects (including alcohol and caffeine)
• Stress from worrisome problems or situations

There are a couple of problems at work here. First is knowing where to start. In some ways this is straightforward to remedy; simply tap into the existing body of knowledge by doing research. However, since each of us is unique, it's not clear how to narrow the variable search space. Also, much possibly useful anecdotal knowledge is not captured, and resides unstructured in discussion groups, individual blogs, etc. I think providing this expertise is the ideal role of an authority like a medical doctor. But of course that has its limitations, e.g., your physician might not be an expert, he might not know relevant details about you, or your problem might lie outside accepted medical state of the art. (My Think, Try, Learn colleague Liza is a poster child - literally - for stepping out and finding solutions.)

Second, it is hard to have the discipline to do good science, such as controlling for one variable at a time or being adequately patient to let enough data come before forming a conclusion and moving on.

Finally, as a practical matter I think existing capture tools aren't up to the challenge of capturing just the right information in a quick manner. Entering data must be extremely fast, or compliance drops way off. (I looked at a bunch of iPhone apps before deciding this is true. We'll be addressing this with the Edison data layer and tools.)

I'd love to hear your thoughts on this. I'll leave with some relevant passages I found in my knowledge repository, i.e, my Big-Arse Text File.

Related resources

Benchmarks Online ~ Project 2061 ~ AAAS: Collaboration among investigators can often lead to research designs that are able to deal with situations where it is not possible to control all of the variables. (This is our goal with the second TTL module that will capture and generalized results - code name DaVinci.)

The election isn't just a referendum on ideology: ...when people are given choices with many variables ... they tend to make the best decision when relying on their unconscious. In contrast, people who consciously deliberate ... tend to fixate on extraneous facts, leading them to bad choices. (More in Jonah Lehrer's How We Decide, which I confess I haven't rea yet.)

[continues] ... the kind of situations that are best suited for each mode of decision-making ... simple problems - those involving a limited number of variables - are best suited for deliberate thought, so that people don't make any obvious mistakes ... complex problems seem to benefit from the processing powers of the unconscious, as long as people first take the time to carefully, deliberately assimilate all the relevant facts.

Info-Clutter: ...a study out of Australia indicates that the human mind can only process four variables at a time.

Understanding Complexity; Thought and Behavior: Thought Leader George Miller's experiments indicate that individuals cannot rely only on mental activity to inquire into relationships among large sets of variables. If learning about complexity is a goal, external adjuncts to learning are necessary to supplement mental activity.