As someone who has spent a significant amount of time conducting primary and secondary research, it’s my natural inclination to always want to define and isolate discrete variables. I approach most problems with the following 2-step problem solving technique:

Step 1: Define all variables associated with the problem

Step 2: Weight all the variables according to their relative importance

Let’s take a simple example.

PROBLEM: (in honor of March Madness) How to jump high.

Step 1 (variables): Upper leg strength, calf strength, body weight, height.

Step 2 (weights): Upper leg strength 30%, calf strength 45%, body weight 20%, height 5%.

Although it’s unlikely that I will grow taller, the above “working” solution presents an actionable exercise regiment should I want to maximize my jump height. I would spend most of my time on my upper leg and calf strength, and some time on losing body weight. It’s a “working “ solution because it’s almost certainly wrong (as in – not the definite solution), but the variable weighting method helps break the problem down into manageable bits (by isolating variables) and eventually through re-tweaking will lead to the definitive solution.

This works very nicely for a simple problem like jump height, but what about more complicated problems like “how do I plan a successful online advertising campaign” or – even more complicated – “how do I live a successful life”? Can the same problem solving convention be applied?

Despite my best efforts (and at one point I did scrawl out 44 index cards, each containing one variable involved in online advertising), I’ve started to realize that really big problems have so many variables that it would be impossible to test and isolate them all. So what do you do when the definitive solution is practically un-findable?

I certainly don’t claim to have the solution to this problem – but so far I’ve done pretty much what I think most people do: embrace the uncertainty – and guess. Because after all, if you spend all your time writing variables on note cards – you’ll never get anything done.

Learning to be Comfortable with Uncertainty
Tagged on:         
  • I wouldn’t call that uncertainty so much as complexity. Sure, a certain amount of uncertainty exists for you, but that seems due to computational difficulty more than anything else.

    Some things, though, are by definition uncertain – complex adaptive systems, for instance, can have different outputs for the same set of inputs. That’s truly uncertain – you can’t predict the output, even if you know all the inputs.

    But it’s good practical advice – and I prefer the term “analysis paralysis.” Getting stuff done frequently trumps being right.

  • Dadio

    Emotional intelligence is the phrase I would use instead of “guess”. To learn to trust your gut is as important as developing strategies to break down and analyze complex problems. This is another example of an area where that elusive balance, we have discussed so often, applies. One should not be too dependent on logic or on intuition in decision making. Try this; next time you open a cabinet and something falls out and you react immediately to catch whatever it is, if you had waited until logic kicks in you will miss catching it every time. Your emotional reaction is 20,000 times faster than your logic controlled physical response. When you are in a presentation situation and you are asked a question for which you did not prepare, what should you do? There is more than one correct answer but the choice you make defines who you are to your listeners. Logical and physical reaction act as two elephants to keep your emotions from running away with any given situation. Embrace the dynamic and learn to become an observer of yourself in these situations. You are not any… thing.

  • Well, I have to disagree with the characterization of that as “emotional intelligence.” Emotional intelligence refers to the ability to (i) identify emotions in the self and other, and (ii) predict changes in those emotions given certain situations.

    While trusting your gut has been popularized by Malcom Gladwell’s Blink, the pop psychology he advocates has, well, some flaws. It’s an interesting argument with a grain of truth: that the brain does not operate in a sequential logical fashion, but rather by association and activation. If you’ve seen something 10,000 times, recognizing another instance will be trivial – since it only requires identifying which previously experienced events are most similar to the new one, and comparing the significance of the differences. The brain is very, very good at that: it’s less good at logical thought, because that’s limited by working (short-term) memory.

    I think what Dadio is referring to is analagous to an area of research in psychology that studies the problem of “free will.”

    The following paper has a good explanation of the research, and gives a nice overview of some problems with it (his argument is that the evidence is inconclusive, which it is).

    But in short, a number of tests have found that the brain “prepares actions” prior to conscious choice, from anywhere to a few hundred milliseconds to 8 seconds before conscious choice. However, the brain retains a “veto” power, e.g. you feel an urge to do something and choose not to. In essence: your brain was ready to do something before you decided to do it, which seems to put a dent in the “free will” argument.

    It’s a little more complicated, and I’d suggest reading the paper.

    http://philpapers.org/archive/BATMCA.1.pdf

    However, the research does support the contention that experience is valuable (which it is: cognitive psychologists have demonstrated that (i) experts exist, and (ii) expertise is _highly_ domain-specific; which means that experience is the determining factor, not intelligence or anything else. And expertise is simply not transferable.

    On a side note: in my earlier comment, I pointed out the complexity problem. Another issue is the “isolation” problem, which economics and sociology (amongst others) experiences. You cannot isolate variables which only exist in the “open system;” e.g. interactions between a large number of agents. Some phenomena change not only in magnitude but kind as the expand – so laboratory tests and the like which can actually isolate variables are methodologically impossible.

    Not to mention the fact that interaction effects only exist when you have multiple interacting variables; so isolating a 10-variable interaction is bloody fucking difficult, and interpreting it (if you can even find it) is bound to be near-impossible (even 3-variable interactions are tricky to interpret correctly; each additional variable adds an order of magnitude of complexity).

  • Well, I have to disagree with the characterization of that as “emotional intelligence.” Emotional intelligence refers to the ability to (i) identify emotions in the self and other, and (ii) predict changes in those emotions given certain situations.

    While trusting your gut has been popularized by Malcom Gladwell’s Blink, the pop psychology he advocates has, well, some flaws. It’s an interesting argument with a grain of truth: that the brain does not operate in a sequential logical fashion, but rather by association and activation. If you’ve seen something 10,000 times, recognizing another instance will be trivial – since it only requires identifying which previously experienced events are most similar to the new one, and comparing the significance of the differences. The brain is very, very good at that: it’s less good at logical thought, because that’s limited by working (short-term) memory.

    I think what Dadio is referring to is analagous to an area of research in psychology that studies the problem of “free will.”

    The following paper has a good explanation of the research, and gives a nice overview of some problems with it (his argument is that the evidence is inconclusive, which it is).

    But in short, a number of tests have found that the brain “prepares actions” prior to conscious choice, from anywhere to a few hundred milliseconds to 8 seconds before conscious choice. However, the brain retains a “veto” power, e.g. you feel an urge to do something and choose not to. In essence: your brain was ready to do something before you decided to do it, which seems to put a dent in the “free will” argument.

    It’s a little more complicated, and I’d suggest reading the paper.

    http://philpapers.org/archive/BATMCA.1.pdf

    However, the research does support the contention that experience is valuable (which it is: cognitive psychologists have demonstrated that (i) experts exist, and (ii) expertise is _highly_ domain-specific; which means that experience is the determining factor, not intelligence or anything else. And expertise is simply not transferable.

    On a side note: in my earlier comment, I pointed out the complexity problem. Another issue is the “isolation” problem, which economics and sociology (amongst others) experiences. You cannot isolate variables which only exist in the “open system;” e.g. interactions between a large number of agents. Some phenomena change not only in magnitude but kind as the expand – so laboratory tests and the like which can actually isolate variables are methodologically impossible.

    Not to mention the fact that interaction effects only exist when you have multiple interacting variables; so isolating a 10-variable interaction is bloody fucking difficult, and interpreting it (if you can even find it) is bound to be near-impossible (even 3-variable interactions are tricky to interpret correctly; each additional variable adds an order of magnitude of complexity).