Today’s topic is elaborating on implications deriving from Yesterday’s, so you may want to read that (The important takeaway is that we’re often not properly calibrated to deal with many things, even though we may feel we are).
Calibration is the relationship between a measurement and reality. When talking about cognitive calibration, then, we refer to the relationship between our beliefs and reality. A good way to know about it is to systematically make your beliefs pay rent, as Eliezer Yudkowsky would say, in the form of predictions. Those who fail to make it should be evicted, and your belief system updated.
Here are some steps we will want to follow in order to be able to track our calibration – or lack thereof:
- Try to predict stuff you feel pretty certain about. Do it properly, as precisely as you think you can get it. It’s good to make sure to note as many measurable features as possible, such as timing and degree of certainty.
- Do this a lot. It is a good idea to carry a small organizer/planner just for this, and mark in the day you suppose something will be due, what it is and how certain you are. Another good idea is to use predictionbook.
- Tally up your results. You may notice miscalibration. Predictionbook sadly doesn’t help categorizing predictions, but try and do that.
- When you notice miscalibration around a particular topic, try and understand why. Examine your motivations for optimism or pessimism, cross-check possible biases you may be subject to, and, in line with the Dunning-Kruger study, learn more about the subject.
This reads easy, but it takes some honesty and work to do well. If you start on this path, you won’t have to take my word on whether you become better calibrated, because you will have your stats to tell you that. If you’re not moving forward – by all means, reach out to me (My e-mail is in the about me page) and at the very least we can bounce ideas on why.
There are two ways in which your beliefs can be adjusted: you may change the level of certainty in your predictions, or you may change the predictions you make. Both are hard to do, but practice makes perfect.