Many of the assigned articles point out that bias is inherent
and impossible to completely erase. It is refreshing to hear this standpoint as
I have only taken one class previously that has addressed bias in this way. In
this class on how sex and gender are reflected in the brain, we discussed the
myth of objectivity: the idea that it is impossible to look from a disembodied view
at natural phenomena and discern some kind of absolute truth. The professor of
this class pointed out that our cultural ideas of a “neutral,” objective standpoint
effectively has meant through the eyes of white males. This fact is reflected
in several trends throughout scientific research, for example, the assumption
that results gained using male organisms would be true for females as well.
Recently, as this
article describes, policies have been implemented by the NIH to increase experimentation
on both sexes through realizations that including both is critical for the validity
of scientific results. This is an example of basic, pervading bias that
counters the idea of pure objectivity. This idea was also described in the
article by Jared Horvath. I appreciate the way Jared Horvath frames his
discussion of this topic in his article, because it does not privilege complete
truth and objectivity over other kinds of knowledge-seeking and discovery. His
debunking of the idea that famous scientists had bullet-proof experiments and
his assertion that every partially-invalid study contains pieces of information
that are useful and valid for the progression of scientific understanding supports
this point. I like this perspective because I feel like, while increasing the
level of accuracy and validity in science is crucial, pure objectivity is not a
possible or desirable goal. Humanity would be better served through an
acknowledgement of inherent flaws and bias and a goal of identifying the helpful
and misleading aspects of a study through an ongoing process (as described in this article).
Showing posts with label #pubpeer. Show all posts
Showing posts with label #pubpeer. Show all posts
Thursday, January 18, 2018
Monday, January 16, 2017
Finding the Beauty in Negative Results
It has long been known that trying to publish a paper merely
to present negative results will not land you in a top-ranked scientific
journal, and likely not in any at all. This problem alone can very much be seen
as one of the bottlenecks of research in the scientific community.
Imagine for a moment you’re a scientist trying to discover
the mechanism behind cancer proliferation. You search the wondrous world of
PubMed, absorbing the works of brilliant scientists, and getting an
understanding as to the current state of cancer research. What you find are
papers on what appears to ‘work’; advances in the field. What you don’t find
are all the experiments and hypothesizes that didn’t work. Okay, now what? Well,
that may not be so bad, right? You can focus on an aspect/gene of interest that
has appeared to show exciting and promising results. But what if your focus was
on STAP stem cells?
In 2014, a group of prominent scientists had a famous paper
published in Nature describing the successful creation of STAP (Stimulus-Triggered
Acquisition of Pluripotency) cells, which was thought to be one of the biggest
break-throughs in stem cell research at the time. Although, luckily for the
scientific community, a peer-review website called PubPeer was also established. On this
website, anonymous members of the scientific community posted a multitude of
accusations into the legitimacy of the paper and its results. Fellow scientists
even repeated the same experiments with vastly different results than the STAP cell paper claimed, thus further
proving the paper was a fraud. (It was eventually retracted).
My point? If there wasn’t an established website, such as
PubPeer, the fallacies in this paper likely would have taken much longer to be
uncovered. This could have led to fellow scientists taking the paper as face
value and dedicating countless hours and expense to fall into this ‘rabbit hole’.
Publishing negative results could only help the scientific community further
have a system of checks and balances, to continue to push science forward. It
should be more widely accepted by all tiers of scientific publishing as negative
data can still be good data.
Subscribe to:
Posts (Atom)