Tuesday, January 19, 2016

Deceit and Ignorance

Deceit and ignorance are the primary themes I am left with from the Dan Ariely Ted Talk and assigned readings.  Fortunately, there appears to be a strong movement to mitigate their effects in science. 

The speaker discussed his research which showed a tendency of people to show a lack of integrity unless reminded to do so, and that people will practice more deception as their deceptive actions become further separated from the reward.  These results are not very surprising to me, but do strongly suggest that some scientists engage in deceptive practices (becoming a scientist does not automatically confer immunity from engaging in a behavior that seems to plague humans).  Moreover, as individuals whose livelihood depends on funding, which is directly linked to publications, scientists have incentive to practice deception.  Thus, despite the best intentions of peer review to eliminate deception, it should not be surprising to find faulty experimental designs, hand-picked results, and outright lies in scientific literature.

I was, however, struck by the alarmist attention given to irreproducible scientific results in the assigned readings, especially the Economist “Trouble at the lab” article.  I was stuck because I: 1) was not aware that there are so many examples of the inability to reproduce results of previously published heretofore important science; 2) tend to be laissez faire, assuming that nearly all canons of scientific discovery are vetted over time (and not single publications that gather dust) – so question whether what I perceive as alarmist rhetoric is really warranted, and 3) am encouraged that efforts are being made by journals such as PLOSone and Science to encourage replication of data.  Ignorance of proper application of statistical tools in designing and interpreting experiments (on the part of authors and reviewers) is given as a major contributing factor to publication of unreliable or unrepeatable data.  Seeking to produce the best science I can, I am keener than ever to learn how to properly apply statistical tools.  


One last observation: From the Horvath blog, it was stated, “Many of my colleagues worry that honesty and full disclosure will tarnish the reputation of science.”  Luckily, Horvath disagrees with his colleagues, but who are these colleagues?  It is hard to imagine that honesty and full disclosure do any more damage than the stance of his colleagues.  “Oh what a tangled web we weave unless we publish and get a grant?”  Is that the mantra?    

Horvath, J The Replication Myth: Shedding Light on One of Science’s Dirty Little Secrets  Scientific American Guest Blog, Dec 4, 2013

Trouble at the lab: Scientists like to think of science as self-correcting. To an alarming degree, it is not. The Economist, Oct 19, 2013

No comments:

Post a Comment