Monday, 22 January 2018

Let the author beware

This is just a short note about something that I encountered during the past year.

Academic researchers often repeat very similar experimental work under different circumstances, often several years apart. They then then publish these works separately. When writing the later manuscripts, it is quite common to copy the description of the experiment from the earlier manuscripts. After all, to do otherwise would be to re-invent the wheel — if you got the description right the first time then you cannot do better by re-writing it.


However, one does have to remember to change the description of those parts that differ between the experiments. The nightmare that we authors have is that we will forget to do this!

These are the titles of two papers by the same authors, which were published in the same economics journal five years apart, but which I happened to read on the same day:
  • What explains Bordeaux wine prices? (2004)
  • Estimation of a hedonic price equation for Alsace, Beaujolais and Provence wines. (2009)
The data collection and analysis were the same in both cases, and so the published descriptions are almost identical. Unfortunately, they are a bit too similar.

Paper 1:
"The experts were then asked to give marks for the quality of the wine and an indication of the extent to which it can be laid down (for red and Sauternes wines only)."
Paper 2:
"The experts were then asked to give notes for the quality of the wine and an indication of the extent to which it can be laid down (for red and Sauternes wines only)."

The reference to Sauternes makes sense when you are writing about Bordeaux wines, as in Paper 1; but it does seem out of place in a paper about wines from Alsace, Beaujolais or Provence, as in Paper 2.

Almost all authors will have done this at some time — the trick is to catch the error before anyone else does.

Caveat auctor!