The group polarization effect: To be or not to be?

It really comes as a surprise to read a discussion about the group polarization effect, a field that goes back so many years (Rodrigo & Ato, 2002, 2002a; Kiers, 2002). Usually the publication rate follows an epidemic spread curve (Nowakowska, 1973), which means we should by now be immune from this bacillus called the "group polarization effect'. The recent discussion, however, is a good example of the tendencies in our world of research: After more than 40 years of research in this area, two young colleagues are able to correctly point out the "lack of fit between the definition and the statistical analysis of the group polarization phenomenon' (2002, p. 5). As someone who started his own research in this area more than 30 years ago (Witte, 1971, 1971a) I feel obliged to join the discussion from a more fundamental and historical angle. Traditionally our approach is to detect effects and then list them in our textbooks, e.g., the Ringelmann-effect, the autokinetic effect, the Asch conformity effect, Milgram's obedience effect, Latane's bystander effect, shared view effect, Koehler effect, blocking effect, group polarization effect, and many more. We are now in a situation where one of these effects is being questioned due to a fundamental error in the evaluation methods that were employed. Perhaps this deficit is much more general than we assume, and may apply to many more effects we are familiar with.