Richard Harris puts scientific research in a new light in the book Rigor Mortis. Exploring how it isn’t as factual as some might think.
In the first chapter he introduces this idea with the example of the paper done mostly by C. Glenn Begley. Begly was a director at a pharmaceutical company. Part of his job was looking at promising research and trying to reproduce it to see if it could be used. Often he couldn’t reproduce the data. So, as an experiment before he left, he choose 53 of what he considered the most pioneering ideas and had them tested to see if they could get the same results. He even invited the original researchers to come help, which they often did, though he was careful to blind the tests. The end data was astonishing; “Of the fifty-three exciting studies that Begley put to the test, he could only reproduce just six.” – p.9. When these findings were published in March 2012 it brought the issue to center stage, where it was a much discussed topic.
Additionally, Leonard Freeman, the founder of the Global Biological Standards Institute, and two other economists estimated that untrustworthy designs, dubious ingredients (such as contaminated cells or antibodies), that aren’t nearly as selective and accurate as scientists assume them to be, and scientists mishandling their data analysis all contribute to this irreproducability. “In sum, Freedman figured that about half the preclinical research isn’t trustworthy. ” – p.14.
All in all, this chapter introduced the idea that scientific research isn’t always correct. In fact, it states more the opposite: that it is usually wrong. This contradicts how I previously thought of science articles. Before I would just assume that the article was correct.
Now I am always questioning. Recently I listened to the episode “Are the Rich Really Less Generous Than the Poor?” of NPR’s podcast Freakonomics. In the episode they reference a study done by Nikos Nikiforakis, Jan Stoop and Jim Andreoni where they dressed up as postal workers and put a see-through letter with cash in rich and poor peoples mailboxes to see who returned the money more often, and was thus more generous. Their results were that 80% of the rich people returned the envelope while 40% of the poor people did. Immediately, I started thinking about their experiment and what they might have done wrong. I realized that already Rigor Mortis is making me think deeper about whether or not an experiment is trustworthy, because of all the history saying it might not be.
Later in the episode they do address some of my concerns. They mention a time when they were presenting their research and another economist brought up that the reason the poor were returning less envelopes was because it was harder for them to do so. They later did more research that supports this theory. It was nice to see that they are accepting of new theories, but it still shows that experiments aren’t always perfect.