Critical Realism – Discuss!

This week at school we are still feeling very philosophical about ontology, epistemology and methodology. We are looking at Critical realism as a post positive philosophical approach to reality and knowledge. I’m not sure why it is called post-positivist as it also looks post constructionist too, I think.

I am requested to:

Read through the paper by Clark (2007). (

Spend no more than 30 minutes making some notes about the following:

  • How does critical realism help the authors approach this problem?

The authors are able to recognise the complexities of the health intervention and take it beyond a positivist view of attendance and effectiveness. Utilising mixed methods they can obtain a deeper understanding of the mechanisms at play that affect attendance and effectiveness of interventions. This then a useful research project in that they can make recommendations for future interventions in given populations and situations.

  • What problems or weaknesses can you see with this approach?

I see that this was a very large and complex research project, not in the grasp of the sole researcher.

Also, I think there were assumptions on the part of the researchers dictating which mechanisms they chose to observe and which questions they chose to ask. This would be due to the researcher’s generative mechanisms at play in their contexts and  situations. The subject was so complex, I suppose they had to start somewhere and in light of new knowledge, since knowledge is fallible, then they could possibly ask new questions.

Just a thought
If reality is dependent on concepts we might not be able to see but have effect on us, then is this not relativism? The affects being relative to the mechanisms at play.

  • Looking back to the earlier slide on ‘Applying critical realism’, does critical realism fit into the area of research you’d like to develop further? Why/why not?

CR appears a useful and realistic (!) approach to large scale research where there is adequate funding and resources. For a small scale sole researcher it appears impractical unless there is already the empirical quantitative research there which can be expanded on by looking at the contextual and situational generative mechanisms that may have influenced the empirical results.

My research will not have large numbers of participants as hypnoacupuncture is not widely practiced. There are no RCT’s to suggest efficacy. I suppose, from a CR position, the RCT measuring tool is not an appropriate fit for the nature of the therapy anyway, ‘closed system’ questions being asked in an ‘open’ system.

So if I haven’t measured efficacy how can I go on to see how generative mechanisms are at play in my client’s decisions to attend and their perceived outcomes? Maybe I can use MYMOP as measuring tool but the numbers will be still small and I will inevitably affect the outcome as they are my own clients. Could the way I practice and consult then be a generative mechanism or are they larger more universal concepts? CR would at east give some ‘outcome’ results rather just clients perceptions, making it a more useful , explanatory and pragmatic undertaking.

I like the CR view, it does make sense in that not only does it explain outcomes it then goes on to make practical recommendations. It takes account of both agency and structure, midway between positivism and constructionist. This has implications for future funding and commissioning of services. The approach is useful in large institutions with lots of resources, at the moment I can see that the sole researcher could plan their research with a CR philosophy and aim to maybe capture a small part of the bigger picture, with a view to keep chipping away!

Tutor’s reply to above: –

Hi Carolyn,

These are great first ramblings, especially given that these concepts are so new. You sound like you have a good grasp of the key concepts so well done!

Good question re difference between relativism and CR. Relativists hold that there is no external reality that can be studied because in their view, all knowledge is subjective due to cultural beliefs and so forth. Critical realism accepts that there are layers of reality – the external reality operating within the closed systems at one end of the scale, and the subjective reality operating within the open systems at the other end (and with many other realities in between!). Just going back to your comment about the RCT not necessarily being an appropriate tool for an open system – whilst this is accepted in some CR circles, some critical realists do see RCTs as being valid as part of their CR evaluations. This is because they accept that there is value in obtaining a measurable ‘hard’ outcome within an RCT – the difference between the positivist and the critical realist approach to such an RCT however would be in how they use or interpret the information within the RCT. So a positivist take on the results could be ‘this RCT showed that this intervention produced this specific effect on individuals with this condition’ (power ascribed to the intervention as described in Table 2 in the Clark paper p520). A critical realist analysis of the same RCT would take into account various factors such as the RCT setting, the care given by research staff and personal health behaviours in producing the outcomes described – the power to influence health outcomes here is ascribed to a complex mix of agency and structural factors and to exploring the question ‘what works, for whom and under what circumstances’ that is central in CR.

It’s great that you’re starting to think about the generative mechanisms that could be at play in your research and also in recognising that CR tends to be outcomes-focused, rather than looking soley at perspectives as relativists do. To add to your comment on whether the way you practice and consult could be examples of generative mechanisms, I would consider these factors to be ‘context’ as opposed to ‘mechanisms’ when referring to the ‘context-mechanisms-observation’ (C-M-O) configuration that was introduced in the Connelly paper (p116). Contexts can be thought of as existing institutional or social conditions which have the power to facilitate or hinder people’s choices. Mechanisms on the other hand are the resources (or absence of resources) that allow for these choices to be made. In the Connelly paper, the example in column two (p116) is given whereby ‘laying on free and reliable transport’ (Mechanism) might be expected to enable certain single parents who are motivated and have marketable skills (Context) to find employment and move out of poverty (Outcome/Observation). In your suggestion of using MYMOP and the notion that you will still be affecting the outcome as they are your own clients – in CR there isn’t a need to ‘separate’ these effects unlike in positivist research. Instead in CR these factors are recognised for what they are and the contribution they may or may not make towards the outcome of your research. This means that the use of MYMOP (to record the outcome – O) and the recognition that you may be affecting the outcome through the consultation or other practitioner factors (context – C) remain compatible in CR. You have identified a possible ‘C’ and a possible ‘O’ now – you may now want to think of some mechanisms (M) from this example that you think could be especially relevant :)

My reply:-

Thank you Lily, this is really making sense now. So the ‘m’ part of the research would be to identify the un-observable mechanisms at play and their effect on ‘o’ and maybe ‘c’. These mechanisms are stratified or layered, in that they work at an individual, a structural and cultural level, in other words far more complicated than the positivist view. So then could I be looking at what mechanisms are at play which render the outcome favourable and compare that to an unfavourable outcome to examine the difference, this would be at the individual level. But then the individual is born into a certain structure and culture which will affect their outcome. To take it even further then, what mechanisms are at play when an individual decides to access CAM therapy? Some individuals are denied access to CAM therapy due to the mechanisms which constrain them – cultural beliefs, the dominant medical model, financial constraints for example. It’s getting wider and wider and in so doing getting more and more interesting. It’s making me think of all the people who are denied access to CAM and how we can identify these mechanisms and make it more accessible. Turn it totally on it’s head and look at the people who do not access our services and find out why by examining the mechanisms and how they affect individual decisions?

Next question

Which of these statements are true about critical realism?

A: Uses an ideological framework

    B: Prefers to adopt qualitative methods
        C: Is ontologically


    D: Can be used for exploratory, hypothesis-seeking research

A: Uses an ideological framework
I tended to think an ideological framework was about power and dominant groups in a society having the ability to indoctrinate the less dominant groups in to believing that what they postulate was the ‘truth’. Sorry, this is not very well articulated but I know what I mean! I relate ideology as being the terminology of the Marxist movement in the stance against capitalism. Cruickshank does mention this in his discussion about CR and that advocates of CR see this philosophy as emancipatory and empowering to the less dominant groups. So CR is a philosophical underpinning which determines the questions being asked and ultimately the methodology used to answer the questions. If an ideological framework is viewed as a set of beliefs which unite individuals
rather than from a structural and cultural angle of the assertions of nations or political groups and organisations, then yes, CR does use an ideological framework intended to explain how causal mechanisms affect who, why and where with a view to social change.

B: Prefers to adopt qualitative methods.
No, CR will use whatever methods and analytical tools are suitable for answering the question

C: Is ontologically pragmatic
Yes, there is a reality, it’s just fallible, the aim of CR research is to not only explain reality in a given context but to change that reality in a positive way. Seeks the information then acts on it.

D: Can be used for exploratory, hypothesis-seeking research
I think CR is described as an explanatory philosophy, but I suppose yes, the aim is to both explore the reality and the mechanisms at play and then explain and aim to change them or recommend change. The hypotheses would not be singular but complex embedded in the layers identified by CR, they would also be open to change and be fluid in the light of new evidence over time.

The video conference was very interesting this week, I felt more comfortable when speaking out. I will look forward to next week now, this is proving very interesting.

Leave a Reply

Your email address will not be published.