1
$\begingroup$

A long time ago I surfed the web and I look a piece of information about the inventors of dummy variables. I recall they were two american economists (father and son). I have tried to find that particular webpage but without any success. Do you know who invented the dummy variables as we use it in econometric and statistical models today? Can any of you provide a web address with that particular piece of information? Thanks in advance to everyone!

$\endgroup$
  • $\begingroup$ Are you speaking of the ANOVA type context or the context of information-losing binning of continuous variables? $\endgroup$ – Frank Harrell Aug 3 '15 at 17:49
  • 3
    $\begingroup$ I suspect that "the" inventor cannot be found. I suspect that this idea was re-invented many times in diffrent disciplines, but if I had to guess then I would say that the first one was in all likelihood not an economist but more likely someone in biology/medicine. One could even argue that the first person to compare two or more means was the first one. $\endgroup$ – Maarten Buis Aug 3 '15 at 18:45
  • $\begingroup$ When I say dummies I mean categorical variables, i.e. variables that can take only two values, 0 and 1. I refer also to the use of dummies in econometrics. The webpage I read referred to two american economists (father and son) that were working in agricultural economics in the mid 1920s. For the father, that was the highest peak of his academic career. The son became an authority in agricultural economics. $\endgroup$ – libercaptor Aug 3 '15 at 19:53
  • 2
    $\begingroup$ I'm not sure anyone can take credit for this, as it's just a way of writing down the common sense notion of a main effect. I'm sure a caveman once realized that antelope are bigger on average in one area compared to another, so I guess he would be the inventor of the dummy variable. $\endgroup$ – dsaxton Aug 4 '15 at 0:08
  • $\begingroup$ I actually suspect that it was some cavewomen that realized some berries were larger on average in one area compared to another, so she would be the inventor of the dummy variable... $\endgroup$ – Maarten Buis Aug 4 '15 at 8:39
2
$\begingroup$

The inventor of dummy variables was George Boole in mid XIX century. On his book "An investigation of the laws of thought: on which are founded the mathematical theories of logic and probabilities" published on 1854 he proposes 0 and 1 as a mean to represent a class. You can look at his fascinating work on page 47 of his book on this link. http://www.archive.org/stream/investigationofl00boolrich#page/46/mode/2up

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.