Pragmatic Concepts and Tools for Data Interpretation

ABSTRACT

This paper takes as its premise that some of the schism between practice and researchin social work derives from the failure of statistics and research courses to provide straightforward guidelines and tools for data interpretation. Particular attention is directed to the paucity of tools presented for assessing strength of association. The failure to provide appropriate tools invites misinterpretation and misunderstanding; specifically,students are “invited” to attach invalid meanings to statistical significance. The paper presents a five-actor model for data interpretation. The factors in the model are: statistical significance, strength of association, causality, generalizability, and importance.The model makes explicit the importance of both statistical and nonstatistical reasoning for balanced interpretation and recognizes the importance of context. When all five factors are considered and given reasonably equal importance, data interpretation can be balanced, comprehensive, and meaningful.