Comment or read comments on this essay

<BACK to MAIN   PAGE 1     NEXT>

 

Posted: July 2009

 

When is a conceptual framework also a theoretical contribution?

Matthias Holweg (Judge Business School, University of Cambridge, United Kingdom)
Dirk Pieter van Donk (Faculty of Economics and Business, University of Groningen, The Netherlands)

 

 

Introduction

This note started out as a very harmless conversation between us – we share an interest in supply chain management, and specifically, what factors make supply chains responsive to customer requirements. In the course of our debate we compared a range of papers that we and others had written on the subject, and soon discovered that the conceptual frameworks that each of these papers promoted differed significantly in form, variables and relationships considered. What struck us was that we could each logically ‘defend’ or ‘justify’ our respective frameworks, yet despite the considerable disagreement between them – in  terms of variables considered, for example – there seemed no obvious reason to either prove or disprove any of them. As we described how we had arrived at our respective ‘magnum opus’, it soon transpired that there is little guidance on how to build such frameworks in the first place. Yet without such guidelines, how do we assess the quality, validity, or contribution made to theory – considering that Operations Management (OM) research spans across the epistemological spectrum whereby many frameworks are never subjected to formal theory testing?

These questions intrigued us, in particular in the light of Roger Schmenner’s recent critique of the use of theory in OM, highlighting a tendency to draw up new theories without ever rejecting old ones (Schmenner, 2009). The way we tend to position conceptual frameworks in OM research marks a case in point, even more so, as we tend to encourage our postgraduate students to draw up conceptual frameworks as an important part of their doctoral work. These often then become part of what the student claims as ‘theoretical contribution’. But can this really be a valid theoretical contribution, if a conceptual framework is essentially bound by the subjectivity of the researcher proposing it for the purpose of her study? In the extreme case, would this not make a conceptual framework essentially non-refutable, apart from judging its consistency and underlying logic? What guidelines do we provide for building sound conceptual frameworks, and when does a conceptual framework also become a contribution to theory? We found the answers to be far from obvious, so set out to enquire.

 

Different forms, different purposes

Our very first observation is that conceptual frameworks actually differ considerably in their form. Miles and Huberman (1994, p. 18) distinguish between a graphical and a narrative form, adding that ‘frameworks can be rudimentary or elaborate, theory-driven or commonsensical, descriptive or causal’. We build on this notion, and identify common types in OM research: firstly, descriptive mindmaps that lay out the main factors or constructs in a certain field of interest and shows a connection between them, but without causality or directionality. Within this framework variables might be shown in a hierarchy of different levels of aggregation (see for example Koste and Malhorta, 1999, p87). Adding a further level of restrictions, relational frameworks highlight the main variables of interest (to the researcher proposing it), alongside the connections between these variables. Connections are generally unidirectional, but might be bi-directional to depict mutual influence. A special case here is the causal framework where connections direct into one way to imply causality. In the latter case the model shows dependent and independent concepts, constructs, or variables (see for example Frohlich and Westbrook, 2002, p733). Thirdly, Venn diagrams (or set diagrams) show all hypothetically possible, logical relations between a finite collection of entities. The aim of this type of framework is to show all entities at the same level, in one space, where each entity can be associated with a specific set or attributes of entities. Circles (or similar shapes) show the belonging and/or identity of each element (see for example Hamilton and Selen, 2004, p273). Finally, narrative frameworks that can be based on of any of the above forms verbally express what the above show in a graphical form. In summary, the main distinguishing characteristics of any form of a conceptual framework are (1) the type of elements considered, (2) their relationships in terms of causality and directionality, (3) the hierarchy between these elements, and, finally, (4) the representation of these elements.

 

 

 

<BACK to MAIN          NEXT to PAGE 2>

 

 

Comment or read comments on this essay