Montag, November 24, 2008

How Science Can Work vs. How Science Should Work...

This is an interesting insight into how science can work.

Basically, two data sets that were in agreement turned out, as time went on, to be less and less in agreement. One pointed to cooling, the other pointed to warming.

So they "correct" one of the data sets.

Then they discover that reconstructed historical records didn't fit the model results.

So they fix the historical record. By using the "corrected" data sets.

Do you see the pattern here? I see an attempt to fit the data to models and to the expectations of the scientists involved.

This is an example of how science can work.

But it isn't an example of how science should work.

Keine Kommentare: