I was trying to find a copy of this paper as it seems to get referenced whenever anyone talks about uncertainty, but all I could find was a pdf scan from JSTOR. So I indulged my compulsions and set the reading order in the original pdf (here), but the pixelation was just too much for my delicate brain. So of course I redid it. Thanks to the wonders of OCR and hard graft… I’ve got a simple pdf, and an open office writer(odt) file. these look a bit crusty, so there is a nice indd file on it’s way, but that’ll have to wait until another day.

The general gist of the paper is that the real world is too complex to model in all it’s infinite detail, and experimental bias will creep in, but if we are smart about it, and realise that it is happening then we can mitigate against it.

Don’t be put off by the bit with equations in the middle, that can be skipped, and the message still comes out clearly.

there is a preview after the break

<iframe src="http://docs.google.com/viewer?url=http%3A%2F%2Fwww.notionparallax.co.uk%2Fblog%2Fwp-content%2Fuploads%2F2009%2F12%2FEd-Leamer_Lets-Take-the-Con-Out-of-Econometrics-retyped.pdf&embedded=true" width="500" height="780" style="border: none;"> There really should be an iframe here :( </iframe>

If you are reading this on Facebook then that’s great, but if you want to make a comment then follow this link and make a comment at the source.