Ten reasons why ANOVA theory and research fail to explain attribution processes: 2. Methodological problems

Abstract

In this, the second of two companion papers, five major methodological problems of questionnaire research into Kelley’s ANOVA model are identified: the formulation of instructions; the choice of stimulus events; the manipulation of information variables; the choice of attribution measures; and the application of process-probing methods. It is argued that the absence of any sophisticated process propositions in Kelley’s theory, in conjunction with the five methodological problems, has led to a poverty of acceptable evidence about how people process unfolding sequences of behaviour (episodes). In conclusion, some of the process-probing methodologies of cognitive science are considered as a remedy to the naive models and methods of attribution researchers.