Opinions versus Facts: A Bio-statistical Paradigm Shift in Oenological Research

A substantial oenological literature exists on opinions of experts and neophytes as they relate to opinions about the quality of wines (Ashenfelter and Quandt, 1998; Cicchetti, 2004; Lindley, 2006). These opinions can be contrasted with factual binary questions about wine: Is it oaked? Does it contain sulfites? Is it filtered? Is the grape varietal Cabernet Sauvignon or Cabernet Franc? Syrah or Grenache? Pinot Noir or Gamay? Such factual binary issues are examined within the broader context of the various measures of factual judgment: Overall Accuracy (OA), Sensitivity (Se), Specificity (Sp), Predicted Positive Accuracy (PPA), and Predicted Negative Accuracy (PNA). The resulting biostatistical methodology derives from biobehavioral diagnostic research investigations. The purpose of this report is to apply this methodology to the discipline of oenology to compare wine judgments with wine facts. Using hypothetical examples, wine judges’ classifications of wines as oaked or unoaked were analyzed for their degree of accuracy. The results show that OA is a poor measure of the accuracy of binary judgments relative to Se, Sp, PPA, or PNA. The biostatistics of the problem could have wide-ranging applications in the design of future oenological research investigations, and in scientific research more broadly. (JEL Classifications: C1, L15, Q13)