Generaliserbarhet och giltighet i pedagogisk forskning och teoribildning

Abstract

　

　

With the present ascendancy of qualitative approaches in Swedish pedagogic research and education science questions of generalisability have been played down, but without much discussion of either methodological consequences or possibilities of evidence-based practice. Traditionally, strict requirements for quantitatively based conclusions were adopted to ensure strong explanations and predictions as well as effective practical steering and control. Today deeper reflection on the part of the practitioner is often taken as the main road to application. However, neither statistical-technological nor qualitative-naturalistic models have been very success-ful. With the linguistic, communicative and discursive turns new approaches appear, in which validity claims as to basic relationships between theory and practice may be considered with regard to both situated conditions and wider socio-historical contexts. With reference to Habermasian discourse ethics it is argued that a model of reconstructive validity may succeed to safeguard the legitimate interests of everyone concerned in ways that highlight also moral validity claims in pedagogic theory and research.