Ludicrous Pretense in Scientific Papers: Medawar's Three Laws

Medawar, P. B. 1964. Is the scientific paper fraudulent? Saturday Review, 49:42-43.

1. The Law of Conservation of Information: Generalizations cannot contain more information than the sum of their foundations.
2. The Law of Entropy of Information: Orderly general statements do not somehow emerge from a disorderly array of facts.
3. The Law of Bias in Information: There is no such thing as unprejudiced observation.

While this is a sort of "tongue in cheek" summary of Medawar's paper (and one I think Medawar would appreciate, based on his writing) the concepts merit serious consideration.

Like Hoffmann, Medawar challenges the structure of the scientific paper as a misrepresentation of the nature of scientific thought.  He even goes so far as to label the standard form fraudulent, a totally mistaken conception, and a travesty.  According to Medawar, the "orthodox" IMRAD (Introduction, Methods, Results, And Discussion) format follows an inductive structure.

Figure 1. This is how I illustrate the inductive approach to my students.  After making a multitude of observations in a number of areas, Charles Darwin posited Natural Selection as a mechanism of evolution.

So, what's the problem with the inductive approach?  Well, using induction, one cannot both discover and prove at the same time.  Doing both requires the hypothetico-deductive interpretation.  In other words, scientists generally have the end in mind when they begin an experiment.  Hypotheses allow for focus in experimental design.  Deduction gives us a framework or a lens through which to view data.

Figure 2. This is how I illustrate the deductive approach to my students.  Using theory, in this case Natural Selection, we can create hypotheses and posit an explanation for results.  Here, antibiotic resistance is explained.  Bacteria able to survive exposure to an antibiotic because of a mutation reproduce, whereas bacteria without the mutation die out.

Hoffman touched on this very same idea... that scientists do view studies through lenses; they have particular motivations.  As Hoffman says, "But of course scientists are human, no matter how much they might pretend in their articles that they are not."

This is a HUGE difference between the physical sciences and the social sciences.  From what I've learned so far, social scientists are all about identifying bias.  They know its there, acknowledge it, and move on.  I'm beginning to think we scientists are a repressed lot...

Anyway, Medawar's big take-home message here is this:  Turn the scientific article upside down.  Put your conclusions first and then set about showing support with the data. 

I find this suggestion rather intriguing, especially recalling Chris Mooney's post, "Good communication is good scientific practice."  He discusses the most important facets of communicating science to the public: put the bottom line up front.

In a response to last week's post, I talked about why I thought Hoffman's suggestions for change to the scientific paper were nigh impossible.  Medawar's may be closer to our grasp.  Indeed, some elements are already sporadically present.  While the discussion remains at the end of the paper, it is occasionally combined with the results, and introductions are often direct in the intent of the researcher to apply biological principles to some ecological "problem" or at least identifies the theory through which the study finds its meaning.

I have to run off to do a class visit, ironically on how to write a scientific paper.  I'll share my thoughts on this later.