Thesis Title: Subtitle

Choosing Methods: A Critical Realist Approach

The use of critical realism does not imply any particular methodological approach. It is a way of conceptualising the object of study, which then leads to certain methodological decisions. Scholars such as Danermark et al. (2002) have developed a methodological approach which accounts for a critical realist ontology in the research process. Danermark et al. (2002: 151-152) argue that a critical realist approach to ‘doing’ research means that the research must adopt a critical attitude towards the following:
“1. The claim that it is possible to understand and explain phenomena by using methods from the natural sciences, which presuppose more or less closed systems;

2. 
   methods based on purely subjectivist assumptions, i.e. that social reality is nothing but a social construction and not an interpreted objective reality; and
   
3. 
   approaches based on the objectivist assumption that subjective intentions should be excluded from a scientific study of society.”
   

In being critical of the above, Danermark et al. (2002) advocate combining methods as the best way to embody a critical realist approach in practical research work. In the context of this research the combination of methods was achieved in the construction of a case study which will be discussed in the following sections. Situating the methodological pluralism of this research within critical realism avoids some of the critiques levelled at mixed methods research. Scholars have argued that mixed methods research mixes paradigms (Creswell 1995; Tashakkori and Teddlie 1998) and are generally critical of the link between paradigm and methodology that Danermark et al. (2002) advocate. Scholars like Creswell (1995) and Howe (1988) contend that ontology and epistemological concerns should not be placed above more practical issues. This position is in opposition to a critical realist methodology that emphasises the link between ontology and methodology.
In keeping with a critical realist perspective, Danermark et al. (2002) reject some of the usual justifications for conducting mixed methods research such as explorative-directed
qualitative study in which the qualitative portion of the study informed the instruments developed for the quantitative work that follows. Additionally, they are sceptical of the use of mixed methods for validation and argue that, from a critical realist perspective, this amounts to an epistemic fallacy. Empirical connections themselves cannot identify the generative mechanisms, which are central to critical realist ontology (Danermark et al.
2002). Instead a critical realist approach holds that that interaction between mechanism and event can only be deduced through the focused study of consciously selected cases. For this research the consciously selected object of study is Facebook. This is the perspective taken for this research and is in keeping with the primary method for this type of research – the case study. Through these methods I can achieve the “intense and focused study” (Danermark et al. 2002), which can contribute new knowledge to the forces behind the patterns observed.
Case Study: Strengths and Limitations
The deep context-dependant knowledge generated by case studies is a useful way of capturing the rich ambiguity associated with lived experience (Flyvbjerg 2001). Consistent with this, the data generated in this research is context dependant (like all case studies) and constrained by my own biographical characteristics. Case studies focus on an individual unit as an object of study and Stake (2008: 119 – 120) further defines case studies as “functioning specific” or a “bounded system.” In this research Facebook represents the bounded system that constitutes the object of study. The case, or the individual unit under examination, is a network on Facebook, and how those in this network relate to the object of study (Facebook). Using case studies as a method to conduct research raises attendant concerns about the generalisability of the findings produced by this method.