2

By 'trivial' error I mean one that's local and easy to confirm and easy to understand. Something that makes you look silly. For example:

  • Sign errors in mathematics.
  • Forgetting to do a unit conversion in physics.
  • Skipping over a number in accounting.
  • Off by one errors in programming.
  • Fencepost errors in programming.
  • Out of bounds errors in programming.
  • Null value errors in programming.
  • Rear-ending someone at a stoplight when driving.

Some historical examples:

Some anti-examples:

Anyways... I'm looking for studies that investigated how often these mistakes happen, and how much the rates vary from person to person. Do some people make lots of them? Do some people make almost none? How much does practice affect the error rate? Are there disorders that affect the rate?

Craig Gidney
  • 151
  • 2
  • For me, this seems very difficult to properly define and quantify. Maybe for a restricted set, e.g. programming errors, you could do it. – awakenting Sep 14 '16 at 17:26
  • @awakenting It's fine if the study is about a particular type of simple error. Mostly I want something I can point to and say "this seems to happen to everyone, your system has to deal with this happening" or alternatively "yes, training the users will work". – Craig Gidney Sep 14 '16 at 18:01
  • 1
    hm, did you look at the user experience SE ? http://ux.stackexchange.com/ I could imagine that their expertise is better suited – awakenting Sep 14 '16 at 19:05
  • I know there exist some studies to error rates in aviation, but these are not "trivial" errors sadly. Perhaps you do not want numbers per se, but instead give some examples of how people could make specific mistakes. Otherwise look for Rasmussen's taxonomy of behavior (rule, knowledge and skill based) and Reason's error taxonomy (slips, lapses, mistakes, and violations). – Robin Kramer-ten Have Sep 15 '16 at 16:48
  • Kano et al (2007) do this for typing errors. – Arnon Weinberg Nov 10 '21 at 03:48

0 Answers0