Pattern matching in qualitative analysis

Sinkovics, N

doi:10.4135/9781526430236.n28

Pattern matching in qualitative analysis

Lookup NU author(s): Professor Noemi Sinkovics ORCiD

Downloads

Full text for this publication is not currently held within this repository. Alternative links are provided below where available.

Abstract

The importance of pattern matching for business and management studies -- Hodgkinson and Starkey (2011, p. 364) note that researchers in business and management studies (BMS) need to ‘re-evaluate their conceptual and methodological armoury in order to ensure the field continues to be both scholarly and relevant to a diverse array of constituents’. Although this statement is equally relevant for both quantitative and qualitative research, the latter is more prone to criticism due to its inherent messiness and complexity (cf. Sinkovics & Alfoldi, 2012). Furthermore, there are epistemological disputes about the extent to which it is possible for constituents (fellow researchers, practitioners, policy makers, etc.) to interpret the presented data in the same way and arrive at the same conclusions as the investigator(s) of the presented qualitative research. This makes the creation of guidelines/criteria for the design and evaluation of qualitative studies challenging (cf. Hammersley, 2007; Johnson et al., 2007). To this end, pattern matching can be tremendously helpful at various levels. First and foremost, it aims at externalising implicit mental models and assumptions as much as possible. This helps the readers of the qualitative piece of work to retrace the thought processes of the investigators and to better understand how and why they arrived at the presented conclusions. Second, pattern matching requires meticulous contextualisation, clear-cut theoretical formulation, as well as detailed and precise operationalisation. While it is not possible to identify and externalise every aspect of our mental models, the concious application of pattern matching will improve the way researchers go about the design, implementation and write-up of studies that satisfy the double hurdle of rigour and relevance (cf. Pettigrew, 2001). Fundamentally, pattern matching involves the comparison of a predicted theoretical pattern with an observed empirical pattern. The underlying assumption is that human beings make sense of the world by comparing what they observe externally with internal mental models (Hammond, 1966a). As a consequence, even in cases where the pattern match is not made explicit, it is still present to a certain degree. At the same time, the purposeful application of the pattern-matching logic will result in a more rigorous and structured research process and write-up as it requires systematic planning, including as much advance conceptualisation work as possible and detailed documentation throughout the entire project. While the idea of pattern matching originates from a quantitative tradition, the application of its principles does not necessarily require the use of statistics (cf. Ghauri, 2004). In this chapter, the focus is on the demonstration of how patternmatching principles can be used in the design and implementation of qualitative research. The intended contribution is twofold. First, the chapter aims at providing readers with an indepth understanding of how this idea emerged and developed over time. In order to achieve this goal, a brief but detailed overview of the historical and philosophical roots of pattern matching will be given followed by an elaboration of the general logic of pattern matching. The latter is meant to offer the reader some guidance regarding what he/she should consider when consciously applying the principles of pattern matching in a study. Second, the chapter goes beyond the existing literature on the application of pattern matching in qualitative BMS. This is done by the classification of pattern-matching applications in three categories according to the degree of the pattern match. They are labelled ‘full pattern matching’, ‘flexible pattern matching’ and ‘partial pattern matching’. For each category a detailed example from a published study will be provided.