Biomedical researchers should declare their assumptions.
Recently, the Editor-in-Chief of Science asked its reviewers and
editors to identify papers submitted to the journal that demonstrate
excellence in transparency and instill confidence in the results (McNutt
2014). This was done in response to issues with irreproducibility
(primarily) in preclinical studies. There may be other ways scientific
papers can be made more transparent and reliable. Sometimes very solid,
or at least well-intentioned, scientific reports contain assumptions
which, if false, severely weaken the work. Assumptions are facts taken
for granted, and in an argument, assumptions must be allowed as true for
that particular argument to advance. Calling assumptions into question
is a legitimate way to undermine a claim or simply deepen our
understanding of the issue; however, if assumptions are never identified
(declared), then spurious claims sometimes may be advanced as fact. One
example of this occurred in a 1978 paper in Nature that is frequently
quoted concerning cultural evidence of dog domestication. That letter
reported a human skeleton found in a grave in what is now Israel with
its hand resting on the chest of a puppy skeleton (Davis and Valla 1978)
and the authors interpreted this as evidence of affection between the
human and the dog. They said, "The puppy, unique among Natufian
burials, offers proof that an affectionate rather than gastronomic
relationship existed between it and the buried person." Offers
proof? Several assumptions are actually necessary to draw this
conclusion which were not identified in the paper: 1) that the puppy was
purposely buried with the human, 2) that the hand was intentionally
placed on the puppy, and 3) that this placement meant affection or
familiarity between the two. Any one of these assumptions could have
been false. In our opinion, readers of scientific and/or biomedical
papers would benefit from a brief textbox or section at the beginning of
each paper wherein underlying assumptions are disclosed. Mathematicians
know all too well the dangers of hidden assumptions. There is perhaps no
other discipline that can so readily prerequisite the importance of
clearly stating all relevant assumptions. The counterpart in mathematics
of hypothesis-driven experimentation and conclusion is the statement of
proposition (or theorem) with a corresponding proof using formal logic.
In this paradigm, the current setting (or global assumptions) and
assumptions related to the problem at hand are carefully crafted and
explicitly listed. From these assumptions and previously established
results, the desired result may be deduced through an application of
formal logic (proof by induction is an exception). Making a false
assumption, either explicitly or implicitly, will always lead to
erroneous results. Thus, it is a common practice in mathematics papers
to a priori list all of assumptions regarding the proposition to be
stated and proved (Lee 2010). "Perhaps biologists should do the
Davis, S. J. M., and F. R. Valla. 1978. Evidence for domestication
of the dog 12,000 years ago in the Natufian of Israel.
Nature 276: 608-610.
Lee, K. P. 2010. A guide to writing mathematics.
http://www.cs.ucdavis.edu/~amenta/w10/writingman.pdf McNutt, M. 2014.
Reproducibility. Science (Editorial) 343: 229.
Jerome Goddard (1) and Jerome Goddard II (2)
(1) Department of Biochemistry, Molecular Biology, Entomology, and
Plant Pathology, Mississippi State University, Mississippi State, MS
(2) Department of Mathematics and Science, Auburn University
Montgomery, Montgomery, AL 36117