Research questions, hypothesis and views on causality

Print Friendly, PDF & Email

(T847, Block 2, Activities 1 and 2)

In TMA01 I stated the aim of my research project as follows:

to investigate the degree to which systems thinking is an ‘absent competence’ or ‘constrained capability’ amongst those involved in leading partnership working for wellbeing and health in an English city

Block 2 asks me to consider whether I should also use research question(s) and hypothesis as a way of tightening up the scale and scope of my research project.  It also suggests that I am clear on my perspective of causality.

Research question

The first task I am inclined to do is to simply re-state my aim in the form of a question:

To what degree is systems thinking an ‘absent competence’ or ‘constrained capability’ amongst those involved in leading partnership working for wellbeing and health in an English city?

The T847 materials suggest that iterating questions to a lower more specific level helps to narrow the scope of a project.  I think I have already done that through constructing my aim.  I went from “those involved in partnership working for wellbeing and health” to “those involved in leading partnership working for wellbeing and health” to “those involved in leading partnership working for wellbeing and health in an English City“.

Although I could debate endlessly about the issue of the multiplicity of people with leadership roles, for the sake of this study I am actually thinking of those who have got formalised leadership responsibilities by nature of their membership of the shadow Health and Wellbeing Board (25 individuals) and/or their seniority in a partner organisation (approx 10 additional individuals).  The reason being is that these are the ‘types’ of person that literature seems to be focussing on when it talks about leadership competences and the need for systems thinking – and I specifically want to address the assumption that systems thinking is an absent competence.

I was interested to find this website which highlights three different types of research question:

  • Descriptive – seeking to describe what is going on or what exists
  • Relational – seeking to see if there is a relationship between two variables
  • Causal – seeking to determine whether one or more variable causes or affects other variables (or outcomes).

Looking back at my question, I can see:

– a descriptive element: is there ‘evidence’ that systems thinking is going on/exists?

– a relational element: is there ‘evidence’ that systems thinking ‘usage’ is constrained in some way by the setting?

But I am a little uncomfortable that these are yes/no questions, so let’s re-phrase them:

To what extent do those involved in leading partnership working for wellbeing and health in the participating English city:

  • exhibit systems thinking capabilities?
  • experience conditions that constrain their systems thinking capabilities?

Okay, I think this has been a useful exercise because it has helped me think about the tension between using data that will be generated through an action research intervention (see this blog for first draft interview structure) AND the need for data that is sufficient to form some sort of confident ‘answer’ to these questions.

In terms of ‘access’ to the research participants, I only have the semi-structured Appreciative Inquiry interview ‘data’ – this will not be designed to specifically elicit answers to these questions.  It is more that I am working on the assumption that if people talk about their ‘best’ experience of partnership working then they may describe instances that are consistent with systems thinking…if they don’t then it is not evidence of absent systems thinking capabilities, just an absence of them in their ‘best’ experiences.  I need to bear this in mind as I move onto data generation – could I use ‘observation’ too – do I have enough access to these individuals to use ‘observation’?  Documents are often not generated by the individuals I am concerned about, mostly officers/managers do the writing so there are few documents to turn to.

My alternative is to change the questions slightly to fit with Appreciative Inquiry approach I will be taking:

To what extent do those involved in leading partnership working for wellbeing and health in the participating English city:

  • draw on and value systems thinking capabilities in their most positive experiences of partnership working?
  • appreciate settings that enable rather than constrain systems thinking capabilities?

That feels much better – the important comment is that I have used words like ‘draw on’, ‘value’ and ‘appreciate’ – but the participants may not be explicitly making the connection with systems thinking.  That is an attribution I will be making.  (This reminds me of Ison’s discussion of systems thinking as a social dynamic (2010, p19/20) – others will be saying something or describing an experience that I will be claiming is them thinking or acting systemically.)

The questions also make me think about how I need to be ready to analyse my data.  What would I look for to make a judgement of ‘systems thinking capabilities’ or ‘conditions that enable/constrain’.  I need to firm this all up from literature reviews.

I am not really sure whether this next reflection ‘fits’ with the issue of research questions or not but I want to note it anyway.  The way in which the data I use is generated will affect what I find – if I formally interview people about Partnerships and how we ‘should’ do partnerships then they are more likely to draw on current dominant mindsets – Partnership as structure; hierarchy and so on – this is like the ‘official’ story that research keeps finding and re-iterating.  Somehow I want to get to the ‘underbelly’, the kind of unofficial type of partnership working that goes on behind this official ‘gloss’ – I believe this is how the true work of partnerships gets done and where systems thinking is more evident.  I believe that because that is what I have observed and reflected on in the course of my everyday working life – so an aspect of my research is trying to test that out using more ‘robust’ methodologies leading others to say it is academically credible.


T847 explains that some forms of research question include an assumption or suggestion about the relationship between two variables.  A research hypothesis goes beyond this and makes a specific prediction about the relationship between two or more variables.  It is derived through deductive reasoning based on existing theory.

I thought that this website explaining the differences between deductive and inductive reasoning also gave me helpful insights to understand the nature of hypotheses.

Looking at the differences between deductive and inductive reasoning, I feel as if my study is working more to inductive reasoning.  I stated above that my ‘informal’ observations lead me to think there is a pattern of systems thinking in use but constrained.  The study is about me being more robust in doing this.  I can’t deny that somewhere there is the seed of an unarticulated tentative hypothesis but I don’t feel at the stage where I could (or should) put that in writing without more observation and description.

T847 also explains that hypothesis are most often associated with positivist research paradigm and linear causality thinking.  Neither of these are suitable choices for me or for my research topic.  So I don’t think I’ll be defining a hypothesis for my research.


T847 describes the most dominant understanding of causality – known as the linear or successionist  explanation – “we did x and y happened, therefore x is responsible for y”.  This explanation is then use to apply ‘solutions’ and transfer from one context to another.

I found T551 (available on OpenLearn) useful in reminding me about causal thinking.  Causal thinking is about linking activities and events together.  Causal thinking is ‘objective’ – in that people’s opinions and perspective do not affect explanations.  Causal thinking is ‘necessary’ – the conclusion always follows from the premise (and if it doesn’t then you can seek an explanation for that).  Causal thinking can build up to chains of reasoning.

T551 says that “thinking about chains of causes and consequences or multiple causes,[…] is an important feature of systems thinking”.  This is not as linear in the same way as the conventional understanding of causality.  A systems thinkers interest is in the patterns of causes and consequences – these can help you identify where it may be possible to intervene.  Or as the world of Systems Dynamics call them – leverage points.  In TU811, there was discussion of the ‘trap of reductionism’ – a tendency to reduce complex situations into variables (that are often easier to work with) rather than consider the network of variables.  This is where I am coming from in my research – to date researchers have reduced their explanation of ‘ineffective’ partnership working for wellbeing and health down to the competences of individual leaders.  I want to open it up so we think of, and work with, a network of variables.  I started to do that in my literature review but I can see the need to understand it more and communicate it better through the use of diagramming tools – perhaps multi-causal diagram, sign graph, causal loop diagrams or even a fishbone diagram.

So I think the distinction that is important (mentioned in T551) is that traditional ‘positivist’ scientific method would use notions of causality that assume objectivity and break situations down into parts where single cause and effects are likely.  It then uses these notions of causality to accurately predict what will happen when you intervene.   Whereas systems thinking is more about interconnectedness of complex systems and here the results of interventions are not predictable – there may be unintended consequences.

I was also interested in the ‘realist’ explanation of causality mentioned by T847.  Explained as “CMO (C+M = O): context plus mechanism(s) equals outcome”.  This seems to have quite a Systems feel to it – acknowledges the setting.

I think there is a bit of a problem here with language – the word “causes” is so easily associated with the conventional ‘linear’ explanation of causality, I would be reticent to use it linked with other types of explanation.  More useful phrases are ‘influences’ or ‘the factors that interact to’ or ‘may lead to’ or ‘may contribute to’.

So what does all this ‘causality’ stuff mean for my research.  I don’t think that it changes my thoughts on data generation but it does make me think about how to ‘analyse’ the data.  In addition to excerting and codifying themes, I could use diagramming to model the causes-consequences-constraints that participants identify in their interviews.  Better still, this could be carried out collectively by the research team.


One thought on “Research questions, hypothesis and views on causality

  1. Just an extra note on causality. I’ve remembered that there is a difference in views on causality in Systems – depending on whether you are thinking systemically or systematically.

    Ison (2010, Table 8.1) gives the distinction as follows:
    systematic – systems are comprised of chains of cause-effect relationships
    systemic – systems are characterised by feedback; may be negative or positive

Share what you think...

This site uses Akismet to reduce spam. Learn how your comment data is processed.