Checklists are a rationality technique, mentioned previously on OB. Everyone knows this, but we don't hear about them as often as we should, possibly because they seem prosaic and boring.

In the context of doing something over and over, there is a checklist improvement cycle.

You try to make the thing (e.g. the blog post, the rational decision, the mathematical proof)
For each kind of error in your checklist, you search the thing for that kind of error, and fix it if it occurs.
When something that passed your checklist turns out to have had an error, you add that kind of error to your checklist.

There are many caveats to this description: Some checklists are not primarily lists of errors, but primarily ordered procedures. You may want to complicate the cycle to track the cost and benefit of the items on the checklist. We're assuming that errors are eventually discovered. I want to pass over those caveats and claim that this kind of checklist-of-errors is very successful. If you agree, my question is: What feature or features of our minds are checklists compensating for? If we understood that, then we would be able to use checklists even more effectively.

The act of considering the current checklist item and the post/decision/proof simultaneously reminds me of "Forced Association", a creativity technique. So one idea is that by putting one's mind into several different states via forced association with the items on the checklist, we gain more independent chances to detect an error.

Even if you haven't made any errors yet, and so your checklist is empty, conducting several searches for errors while wearing De Bono's hats (or another forced association list) might be a way to make fewer errors.

If you're consciously looking for a red minivan, then you will notice more red minivans. This "noticing" seems surprisingly spontaneous, unlike deliberately scanning a scene and considering "Is that a red minivan? Is that a red minivan?". Possibly this is because the noticing is being done by unconscious modules of our minds. A checklist breaks our search for errors into a sequence of searches for more specific kinds of errors. Possibly checklists are effective because the general concept "error" is too vague for those modules. By breaking it into easier chunks (e.g. "ad hominem fallacy", "missing semicolon"), we can start using those modules to get a more thorough search.

I admit, I'm not sure how to use this "modules" idea to use checklists more effectively.