Found this on Facebook.  Pretty much, Climate Science is the basis, the fundamental example, of how NOT to do science, of how to screw science up!

This entry was posted in Sophistry and tagged , , , . Bookmark the permalink.

### 8 Responses to Guide to Bad Science

1. James Rollins Jr says:

Super find, Joseph.

2. mitigatedsceptic says:

Well found!
Should be put up in every classroom – not just science.
My only worry – “be wary of correlation and causation… “this is the understatement of the millennium. For three hundred years, this warning has been ignored and is still being ignored. Indeed there may be a case for banning statistics from physical science classes. Mostly statistics are used to offer extensions of the findings of empirical science beyond what has been observed – breaching Newton’s admonition.

3. Bryan says:

mitigatedsceptic
Good point.
When doing experiments the old way of working out uncertanties was to assume all measurements could combine to give a worst case outcome.
Say for
A =BCD

Then A has 6% uncetanty and no ifs or buts

The modern method is instead to give confidence levels from statistics.

Fine but if there is a financial motive the statistics can be ‘spun’
Otherwise when Bankers in 2007 employing ‘rocket scientists’ predicted that there was only a one in a thousand chance of CDU default…….well you know the rest!

4. mitigatedsceptic says:

Once probability or uncertainty is included in a forecast, one thing is certain – the forecast MUST be correct whatever the outcome. No need to fiddle anything. All possible outcomes are covered. Probability and uncertainty ARE the spin! Proposition which include either of these cannot be falsified.

5. mitigatedsceptic says:

Did the experts try to pull a parlour trick on the Commons Committee On the IPCC AR5?

I have just noticed that on page 3 of the oral evidence Prof Allen is erported as saying –
“Just to illustrate this point, I hope you will forgive me, I did send in as some supplementary information the little two-time constant climate model that the IPCC uses. I would encourage Committee members—it is an Excel spreadsheet, it is easy to do—to put in some numbers, see how you can explain the data, and see whether or not your results differ from those of the IPCC’s. You will find it is very hard to plug numbers into that model and not get results out that are broadly consistent with the IPCC projections. Ultimately, it is the data that speaks for itself.”

In short, whatever data are fed into this IPCC model. the model predicts warming. Surely this tells us that the model must be biased? Surely It is not the data that (sic) speaks for itself but it is the model that speaks for itself (regardless of the data)? Is this not just the old schoolboy trick that starts with “Think of a number – any number will do. Now process it thus…” and the answer always turns out to be the number you first thought of. This conveys information, not about the numbers being processed, but about the processor itself.

Have the experts really stooped to playing tricks like this?

6. Yes, they have. They’re not experts.