Data torturing

Data torturing is an abuse of scientific research by methodological flaw expressed by saying "Torture the data, and it will confess to anything." It is an attempt to manipulate the data in order to confirm the whishful thinking of the author and reach a apriori wanted result of the research. For example, a trial can be deliberately rigged to give the right answer to its sponsors. One systematic review perfomed in 2003 discovered that studies funded by a pharmaceutical company were found to be four time more likely to give results that were favourable to the company than independent studies. Data torturing is often associated with computer simulations.

Examples

COBE project

It has been claimed by NASA contractor that in COBE (Cosmic Background Explorer satellite) project, the sensitivity limits of one part in about 30,000 were "increased" to one part in 100,000 by means of data torturing. The data had to be ‘massaged’ through computer enhancement. The work was done at the University of California at Los Angeles, Berkeley, and Santa Barbara by researchers George Smoot, Charles Bennett, Edward Wright and Alan Kogut. With computer enhancement it is a prerequisite to know what is being searched for, which means that the data analyst brings a pronounced personal bias to the procedure. The discovery of anisotropy of one part in 100,000 is questionable to anyone who knows how the original data was processed and who is objective about cosmos origin theories.