What ways can I test my mental models?

If the ST Loop is all about testing our mental models against reality and then getting feedback, what are some of the ways a person can do that?

Fantastic question!

So many ways. But I''ll start off by sharing a quote from Einstein who said, "The whole of science is nothing more than a refinement of everyday thinking." In order to answer this question, it is important to understand what that means. Because... all of science—every bit of it—is an attempt to build mental models that approximate reality, test them against reality, get feedback from the real world, and adapt our mental models. That is science!

And, what Einstein is saying is that all that science is just a slightly more refined version of everyday thinking. So, the answer to the question is: you can test your mental models in all of the same ways that science does. But the crux of the answer to this question is the degree to which you test them. 

For example, if you're Anthony Fauci or the President of the United States, you may want to test your mental models about COVID or some foreign policy decision in exactly the same way and to exactly the same degree that science does. 

But, what if you have a mental model about your kid, or your boss, your your self and you want to test it out?

  1. You may not be able to assemble a team of researchers (or spend five years investigating yourself) to identify all the available research on a topic, but you can "Google it" and pay attention to the reliability and veracity of the sources.
  2. You may not have the time that Jane Goodall had to sit in the bush and observe primates, but you could pay closer attention to behavior, especially when people don't know they're being watched.
  3. You may not assemble a team of survey design experts and poll a large swath of the population, but you could ask your son, self, or boss a well-thought-out question.
  4. You perhaps won't marshal the same resources and timelines (i.e., millions of dollars and months or years) to develop empirically valid and reliable research experiments, but you may still want to do a little thought-experiment (what Einstein called a Gedankenexperiment) or an actual mini-experiment.
  5. You may not have the entire edifice of science behind you or the funding to bring multiple research from multiple places but you could do a synthesis of all your findings in one bigger meta-analysis.

And, in those five items we have navigated the full continuum of scientific research as shown in this matrix of methods [1] below. 


These simple examples in the list of 5 above mimic the basic processes and methods of scientific research:

  1. Review the literature (i.e., Assess the current state of knowledge)
  2. Observe
  3. Question
  4. Experiment
  5. Synthesize (meta-analysis)

So, do the same thing, but to the degree necessary. Observe. Ask questions. Don't make assumptions. Devise mini-experiments or thought experiments. Synthesize what you have learned from many sources.

Each of these is an art form and a science.

  1. To observe without projecting yourself on the observations is not easy. To free yourself of all judgements and assumptions and just watch. Listen. Take in what the system is telling you about itself. See the system the way it is, not the way you want it to be. Learn to love reality.
  2. Practice the art and science of asking great questions. Just like a survey design expert, don't ask loaded questions. Keep your questions simple and direct. Don't ask double-barreled questions (questions that are actually two in one). And most of all, stop asking questions that are not actually questions but statements dressed up in question-clothing.
  3. With experiments, design little ones that are simple and have relatively conclusive results. For example, if I want to test whether my child doesn't take out the trash because they don't know how to or don't notice it, I can devise an experiment. First, thoroughly demonstrate how to take out the trash. Next, have them take the trash out with you. Then, see if the trash get's taken out (it won't). Next, test if it is a noticing problem by moving the trash to a place where they can't not-notice it. For example, put the can in a confined space where one would have to purposefully avoid it such as outside the door of their room. Or, come up with thought experiments like this: If the behavior I observe is X, what might be a few of the mental models that could be being enacted? Or, if a person's mental model is X, what might be the behavior I will expect to see?
  4. With synthesis or meta-analysis, bring all your data points together. The answer to a question might conflict or align with the results of observation and experiment, for example.

    Just like research, practice makes perfect. You will get better at it the more you do it. And, one area will support another. Increasing one's insight through observation will lead you to ask better questions. Getting answers to questions will help you to design better experiments. New insights from experiments will help you to know what to observe and what to ask next.

    Most of all, don't see testing your mental models as a chore that must be done but as a joyful experience that you get to have. You get to learn new things about reality! You get to get closer to the truth of the matter! Like science, it's fun!



    [1] Cabrera D. and Cabrera L. The Knowledge-Method Matching Matrix (KMMM): Considering the Condition of Knowledge When Choosing Research Methods. Journal of Applied Systems Thinking (20) 5. (2020)