Monday, July 21, 2025

Is it right to call such results "scientific investigation"?

Cooley and LeRoy (1981) have argued that a close correspondence tends to exist between the advocacy of a theory and the results of scientific investigation.

From "The Monetary-Fiscal Policy Debate and the Andersen-Jordan Equation"
by Dallas S. Batten and Daniel L. Thornton (1986).

2 comments:

Jerry said...

Maybe I should look more into the context to understand what they are saying, but I think it could go in either direction. If you discover scientifically that you can split the atom to release 100x more energy than you can get from chemistry, you might then decide to start advocating to build a power plant or a weapon. The science there is good. The advocacy is still a judgement call that could be right or wrong depending on other factors, but it is at least a result of being more informed about what the possibilities are. Another example might be if you discover that certain kinds of pollution are causing an increase in natural disasters and then start advocating that society should try to reduce those kinds of pollution. You could turn out to be wrong about the advocacy -- it's complicated, and maybe it turns out that you need more industry to get to renewables or something, instead of less industry to reduce pollution. But your science is still good and your advocacy is coming from the new (though perhaps still incomplete) information you gained.

But it could also go the other direction, like say if you are being paid by the tobacco company to not find a strong cancer correlation or something, and you do bad science to get to that conclusion.

It seems like you should be able to tell which case you're in by looking at whether the science or the advocacy happened first? But, it's probably complicated and hard to tell, most of the time.

Jerry said...

I think that in science, this kind of error is supposed to be corrected over time by having other scientists (with no, or competing advocacy) repeat the experiment or analysis and see if they get the same conclusions, and by the peer review process to filter out obviously-bad cases, etc. But, it can be a slow and inconsistent process.