Most inferential statistics trades error for bias. We get sample sizes large enough so that we can treat the uncertainties of estimates mathematically. This only works if the sample is unbiased. In measuring natural phenomena we can usually ignore variation and treat it statistically because we are playing against Nature which has no stake in the game. In financial auditing we have to worry about intentions. People unlike Nature have bad intentions – pennys from millions of accounts are not just random errors if they all accumulate in an employee’s account.

The reason this blog exists is because Steve decided to treat the Hockey Stick statistics not just as one would treat data from Nature but as data from a source with an agenda. That’s why it’s called Climate Audit not Climate Anaysis.

There may be additional sources of uncertainty as yet unquantified

…

It seems likely that different groups of observations may be measuring SST in different ways even in recent decades, and therefore there may be unresolved bias uncertainties in the modern data. Quantifying such effects will be a priority in future work on marine data.

Pl. hurry.

One wonders if anyone actually reads these papers

Apparently you do. I’d venture if you were on that committee, you’d have also been drinking some whiskey.

If they are using his quote as some sort of evidence that “yeah, you can never know everything,” I’ll bite, i.e. “we expect their to be unknowable uncertanties in data.” If, however, they expect this as validation that their methods are correct because of the unknowable uncertanties, they’re just joking, right?

Mark

The sad thing is that if you were to make that case explicitly, the referees might well reject your paper.

None of which matters; issues regarding availability of the underlying data so that you could do your own analysis seem more salient.
per

So in this case the question might be: Did the reviewers actually look at the references?

Scientific papers appearing in our journals are subject to rigorous scrutiny by scientific peers prior to acceptance….—John Orcutt, President of AGU [Letter to The Honorable Joe Barton, 08 August 2005]

I think that this could be really funny. If it didn’t show so much contempt for the peer review system and readership.