Tuesday, December 13, 2011

Heteros*edasticity

Earlier this year I had a post about the decline in the amount of attention paid to the concept of "multicollinearity" in econometrics texts over the years. It's a fully justifiable decline, in my view.



There was also a short follow-up post, prompted by some evidence provided by Boris Kaiser.

Today, I've heard from Alfredo Paloyo, who sent me a copy of his recent paper, When Did we Begin to Spell "Heteros*edasticity" Correctly? Most of you will guess that it's related in some way to Huston McCulloch's well-known 1985 piece in Econometrica. You'd be right.

Alfredo goes through an analysis of the choice between the spellings "heteroscedasticity" and "heteroskedasticity", in  way that's rather similar to, but much more thorough than, that in my first post about "multicollinearity". Among other things, he finds that:
  • The 'correct' spelling, "Heteroskedasticity", was first used in 1927.
  • "Heteroskedasticity" has dominated "heteroscedasticity" (with the exception of 2005), only since 2001.
  • "Homoskedasticity" has dominated "homoscedasticity" only since 2008.
  • Overall, the trend seems to be towards the "k" word(s), and away from the "c" words.
Of course, I'm simplifying Alfredo's conclusions - I''ll let you read the details for yourselves.

Nice paper, Alfredo! I'm sure that readers of this blog will enjoy it.


Note: The link to the following reference will be helpful only if your computer's IP address gives you access to the electronic version of the publication in question. That's why a written References section is provided.


References

McCulloch, J. Huston, 1985. On Heteros*edasticity. Econometrica, 53, 483.



© 2011, David E. Giles

2 comments:

  1. I always though the "c" spelling was British and the "k" spelling American.

    ReplyDelete
  2. Prof. Giles,

    Thanks for linking to the paper.

    Yesterday, I made the point about multicollinearity to first-year graduate students. I realized that, after people have been reared with econometrics textbooks which devote an inordinate amount of pages to this topic, it is much harder to convince them of multicollinearity's unimportance.

    mOOm -- For the most part, -c- did dominate in the UK but recent trends show that the entire profession is changing its preference.

    Alfredo

    ReplyDelete