We, the body politic, has been slammed by a number of alarming "scientific" studies over the past few years. There was the "hockey stick" graph "proving" global warming, the Emory university prof book claiming that colonial Americans didn't own guns,and the Boston Fed study "showing" discrimination against blacks in the granting of mortgages, and a bunch of others. All of these "scientific" studies were later proved false, but only after serious damage was done.
Frazier Institute members B.D. McCullough and Ross McKitrick described these and other sham studies in "The Case for Due Diligence".
A key finding, in every case, the authors of the study refused to release their data and the computer program that analyzed said data, making it difficult to impossible to check the conclusions of the study. Only when critics made Freedom of Information requests, or sued, did the authors reluctantly release the data upon which their conclusions were based. In all cases, critics were able to show the data had been edited (contrary examples discarded) or the computer program had bugs in it.
The McCullough & McKitrick paper is good, but somewhat heavy going. Take away, don't place much trust in "scientific" studies unless other scientists have checked the data and programming. "Peer review" prior to publication in a scientific journal doesn't mean any one checked the work. If the conclusion is politically controversial, watch out and hang onto your hat.
No comments:
Post a Comment