PEER REVIEW OR PRACTICAL REASON?
AN ARGUMENTATIVE APPROACH
Richard Horton
The Lancet, 42 Bedford Sq, London WC1B 3SL, UK
To limit the risk of unacknowledged bias in peer review, serious attempts have been made to systematize appraisal procedures. Although checklists provide a crude normative framework to regulate the conduct of peer review, they do not focus on what should be the central concern of any research report-namely, the validity of the core argument made by authors. The structure of arguments and the processes of reasoning have been studied by Stephen Toulmin (The Uses of Argument. Cambridge: Cambridge University Press 1958). He described the geometry of an argument in a way that is directly applicable to the scientific paper. According to Toulmin, the structure of any argument can be represented by a simple formula:
Message
D ____________________________________ Q , I
| |
since W unless R
|
on account of B
Where D represents the primary data; I, the interpretation; W, a warrant; Q, a qualifier (a measure of the force of an argument); R, the conditions of rebuttal; and B, the information backing W. For Toulmin, the relation between D and I, the argument, is determined by the justification or the warrant. Symbolically, an argument proceeds thus; if D, then I, since W. The aim of any scientific investigation is to establish the warrant. The warrant authorizes the argument; it is the subject of the hypothesis being tested. In biomedicine, Q represents a probability function; B, the degree of random and systematic error; and R, the inclusion criteria for the study (generalizability). These 6 features are essential for evaluating the internal and external validity of the argument being made. This method of argument analysis, if applied to the scientific research paper, offers an opportunity for a systematic yet more logical approach to peer review.
Return to Session Information
