The question of reliability in interpretive psychological research
The "Achilles heel" of interpretive research is that two or more researchers, confronted with the same data, posing the same question(s), will invariably express their findings differently. Different words will be used to express the "same" (i.e., congruent) meanings; it is also likely, of course, that different meanings will be thematized as well. Thus, there arises the question of the reliability of interpretive research methods. In what sense can a method (when applied by multiple researchers to the same set of data) yield results that are not "consistent" yet can still be considered "reliable"? Should we alter our definition of reliability to accommodate qualitative methods, or should we adhere to definitions of reliability that have been developed with respect to quantitative methods? Should we abandon interpretive research altogether if it cannot stand up to existing scientific conventions? These questions are merely rhetorical, for it is clear to us that definitions of methodological criteria are themselves a function of human understanding and interpretation; thus, they cannot be carved in stone and are always open to reconsideration.
Churchill, S. , Lowery, J. , McNally, O. , Rao, A. (1998)., The question of reliability in interpretive psychological research, in R. Valle (ed.), Phenomenological inquiry in psychology, Dordrecht, Springer, pp. 63-85.
This document is unfortunately not available for download at the moment.