Modelling Shared Reality:
avoiding unintended influence in qualitative research
Annemiek van Helsdingen and James Lawley
This article introduces Modelling Shared Reality, a new qualitative research methodology which is rooted in Clean Language and Symbolic Modelling. Using the protocol explained in this article, undesired influence of the researcher is minimized during all stages of the research: design, interviews, analysis and reporting. The methodology is action oriented: both the process and the results function as a catalyst for action or behavioural change (Wadsworth, 1998). It is used in the context of organisational change and development, as well as policy-making, project evaluation and 'civil participation' where the government consults with the public.Unintended influence
Research by Elizabeth Loftus and her colleagues found that the way in which questions were worded altered subjects’ memories of events they had witnessed. For example, Loftus and Palmer (1974) showed that changing a single word in a question could make a difference. After watching a film of a car accident, being asked "Did you see the broken headlight?" instead of "Did you see a broken headlight?" doubled the chances of a subject saying they had seen a broken headlight when there were none in the film. In another experiment, Loftus and Zanni (1975) asked subjects, "About how fast were the cars going when they X into each other?". Significantly higher estimates of speed were given when 'X' was smashed, compared to collided, bumped or hit.Clean Language and Modelling Shared Reality
More recently, after conducting an experiment to test the effect of changing a single metaphor in the beginning of a report on crime (crime was described either as a 'virus' or a 'wild animal') Thibodeau and Boroditsky (2011) concluded that "even the subtlest instantiation of a metaphor (via a single word) can have a powerful influence over how people attempt to solve social problems like crime and how they gather information to make ‘well-informed’ decisions. Interestingly, we find that the influence of the metaphorical framing effect is covert: people do not recognize metaphors as influential in their decisions."
These studies, and many others, have demonstrated that leading questions and statements loaded with presupposition can have a major effect on how people describe their experience. Crucially, an interviewee is highly unlikely to be aware that how the questions are asked is influencing their answers. Likewise, if a researcher is not aware of the metaphors, presuppositions and assumptions in their questions there is little they can do to avoid unwittingly biasing the answers.
Does this actually occur in academic interviews? A review by Tosey (2011) of interview questions asked in a qualitative study published in a respected academic journal found that fifty percent of the pre-designed questions contained metaphors that were likely to lead the interviewee’s answers, e.g. "What is the image you carry around that drives your actions today?"
The discipline of using Clean Language in qualitative research directly addresses the issues of priming, leading and loaded questions, thereby reducing the potential for unconscious interviewer bias.
Clean Language, a technique developed by Grove (Grove & Panzer, 1989) and modelled by Lawley and Tompkins (2000), uses specific formats designed to allow interviewees to express their experience using their own lexicon, and to minimize unintended influence or interpretation by the interviewer.
Although Owen (1996) was the first to recognise the value of Clean Language as a "linguistic-experiential phenomenology", it was only recently that Tosey (2011) published the first study on the use of Clean Language and Symbolic Modelling as a qualitative research methodology.
Clean Language uses clarifying questions exclusively. These questions have a strict syntax and are designed to be content and judgement free. To explore an element of the interviewee’s experience, the interviewer uses the interviewee’s precise words. Symbolic Modelling is a methodology by which the interviewer uses the interviewee’s descriptions to model the organization and logic of his or her experience.
Examples of Clean Language questions:
And want kind of [result] is that [result]?
And is there a relationship between [result] and [working together]?
Note: [...] indicates the precise words used by the interviewee.
Modelling Shared Reality was developed by Stefan Ouboter, and extended by James Lawley, Annemiek van Helsdingen, Wendy Nieuwland and Maaike Nooitgedagt. It is a qualitative research method that uses the techniques and underlying principles of Clean Language and Symbolic Modelling to provide insight into the shared experience of a population which is then used as a springboard for decision making and action. It is based on the premise that how people experience ‘reality’ is a strong predictor of how they are likely to act (Homan, 2005).
Since 2006 Modelling Shared Reality has been used in 15 projects involving groups as small as a dozen people to multi-organisation projects. It has been applied both within and across organizations such as government programmes, urban development projects and to increase intra-company collaboration. To illustrate how the methodology works, this article uses the mid-term review of a six-year research project “Leven met Water
” (Living with water). Stakeholders were distributed over 120 governmental and commercial organisations that were involved in around 60 different multidisciplinary sub-projects.
There are six phases to the Modelling Shared Reality methodology. These are congruent with Kvale’s (1996) protocol for interview-based qualitative research. 1. Defining the research
The research project is defined in conjunction with the people who commission the research and other key stakeholders. In this phase Clean Language is used to create a model of the questions, thoughts, and interests of the research sponsors. Since Modelling Shared Reality is a form of action-oriented research, special attention is paid to desired outcomes. A Clean Language question asked at this stage is, “And what would you like to have happen as a result of the research?”.2. Interviews
The interview protocol is semi-structured and is based on a main research question plus a small number of criteria – on average six. The criteria are topics that, based on the analysis at the definition phase, are expected to relate to the main research question. A small number of people are chosen for their strategic placement within the network of people involved in the programme. Each interview begins with a pre-defined question after which the course of the conversation is mostly determined by the answers given by the interviewee.
Thirty interviewees were selected for the mid-term review of the Living with Water project. These included group leaders from projects spread across different subjects, project sizes, organisations, and functions.
The initial interview question had two parts: What is the potential impact of the knowledge gained by your project; and what is needed to successfully complete the Living With Water programme?
Although the interviewer only asks Clean Language questions, this does not prevent them from guiding the conversation. The format of the questions enables the interviewer to keep the interview within the frame of the research question while simultaneously discovering elements of the interviewee’s experience that is most relevant for him or her.
When an interviewee does not mention one of the criteria themselves, the interviewer introduces the missing criteria in the second phase of the interview (using as few metaphors and presuppositions as possible). Thereafter the interviewer reverts to using Clean Language and the precise words used by the interviewee.
Equally the reverse can occur when one or more interviewees introduce a topic that was not part of the original interview protocol. If a number of interviewees mention the same topic, that topic is added to the protocol as a new criteria and included in subsequent interviews.
A common side-effect of Clean Language interviews is that interviewees are surprised by the depth of insights they gain into their own thoughts and feelings about the project. Many interviewees have said they not only enjoyed being interviewed in this way but that it also provided them with a renewed sense of purpose and enthusiasm. 3. Modelling the information
Each interview is recorded and documented in the form of a mind-map with quotes from the interviewee grouped by the criteria. The interviewees’ word-for-word quotes form the basis of all subsequent phases.
The modelling starts with an overview of all the mind-maps and recordings. Researchers look for patterns across the interviewee’s descriptions with the aim of identifying an underlying structure that captures the logic of the whole system. These can be patterns of similarities as well as differences. The researchers consider: What needs to be ‘true’ to connect the individual realities into a shared reality? They keep asking themselves: Did they really say that or do we just think they did? This is part of continually checking all assumptions and conclusions against the precise description given by the interviewees.
The researchers formulate six to ten ‘red threads’ which exemplify key elements that emerged as important to the group of interviewees. Together the red threads provide a coherent description of the shared reality that acts as a context for all of the individual experiences.
4. Draft research report
An example of a red thread that came out of the Living with Water mid-term review was, “There is a paradigm shift taking place in how we think about motivating people.” The interviewees maintained for instance that instead of “first finding an answer and then convincing other people” there is an increasing emphasis on “all stakeholders collectively finding a solution as part of an effectively managed process”.
An explanation of why specific red threads were chosen and the evidence of their veracity is documented. Each red thread needs to be supported with eight or so word-for-word quotes from the interviews. These quotes make up around half of the draft report. In this way a large amount of the source data is available for everyone. Keeping the source of each piece of information clear – sponsor, interviewee, interviewer or modeller – is central to a ‘clean’ approach.5. Workshop(s)
Permission is always obtained to use quotes, and once given the draft report is distributed to the interviewees, commissioning agents and other key stakeholders. The report can be in multimedia form so that it is possible to listen to the participants’ actual quotes.
An integral part of the Modelling Shared Reality methodology are workshops which bring the interviewees together. As a group they are asked to evaluate the red threads for accuracy and completeness. Then they consider the red threads as a whole and draw conclusions from them. In this way those who participate in the workshops take ownership of the results of the research. In the second half, the workshop participants use the red threads and their conclusions to formulate action directly related to the research.6. Final research report
In the final report conclusions, action plans and recommendations from the workshops are documented. Transparency in the process used to reach the collective conclusions and subsequent actions is achieved by preserving the source of quotations and highlighting changes that resulted from the workshops. Results of Modelling Shared Reality
Those who commission the research gain a clear overview of the reality shared by those directly involved in the programme. This provides a platform grounded in personal and collective experience with which to modify existing policy and create new plans to achieve the objective of the programme. Through the research process, interviewees and stakeholders widen their view of the entirety of the programme, whereas previously many had been focussed only on limited parts of the whole. In addition, everyone involved is given an opportunity to consider his or her desired outcomes based on the shared reality of all. Frequently individuals report that they initiate action based on the insights which occurred during the interviews or workshops.
Other qualitative approaches
Corné Nijburg, the programme manager of the Living with Water initiative, said:
My most sceptical and no-nonsense project leader said to me after their interview: “I’ve worked on a lot of evaluations, but never before has one interview given me so much insight in to my own project”.
The mid-term review gave us a new focus for the second half of the programme. We allocated more time and money to creating synergy between projects and learning from each other. We would not have gained such clear insights with a standard method of evaluation.
Modelling Shared Reality (MSR) is a new qualitative research methodology. It can be situated it in relation to more established approaches such as Grounded theory, Ethnography and Phenomenology.
The Grounded Theory method developed by Glaser and Strauss (1967) is primarily an inductive approach. The researcher begins without assumptions about what he or she is going to find. Data from the interviews is used as an empirical source to create concepts and categories as the basis for a new theory. MSR is similarly inductive but researchers model the data in a different way. Instead of creating a theory they select a small number of representative red threads which exemplify the group view and stay close to the interviewees' descriptions.
Ethnography explores cultural phenomena that reflect the knowledge and practices that guide the life of a group (Brewer, 2000). MSR overlaps with Ethnography in that its focus is on (large) groups and draws on the ethnographic method of selecting knowledgeable informants who know the activities of the community well. However, MSR is less concerned with a detailed, in-depth description of everyday life and practice, and more interested in a snapshot of how the group perceives the 'current reality' of their situation.
Phenomenology seeks to gain an in-depth understanding of the nature and meaning of everyday experience (Giorgi, 1985) including subjective phenomena at the individual level. MSR has much in common with the phenomenological approach. In both the researcher is required to bracket his or her assumptions and presuppositions and to collect data from the interviewee’s perspective. While Phenomenology has no explicit method for achieving this aim, MSR employs Clean Language.Critical success factors
While the goal of all research methods is to collect data in a way that researchers do not impose their biases on the information, only MSR has the strict interview protocol of Clean Language to minimise the unwitting biasing by the interviewer.
Once the data is collected Grounded theory, Ethnography, Phenomenology and Modelling Shared Reality differ widely in how the data is analysed and used. MSR seeks to exemplify the current reality of those involved and to use that as a springboard for the interviewees and the sponsors to decide upon actions required to achieve the goals of the programme.
The success of research using the Modelling Shared Reality methodology depends on several factors:
- Those who commission the research need to commit to an emergent process with unpredictable results
- An overall aim for action, based on the results of the research
- The research sponsors want to know the experience of people directly involved in the programme even if there is a potential for divergent views
- The researchers are familiar with Clean Language and are able to reflect on the way their own assumptions may be influencing every stage of the research process
- The workshops are facilitated based on clean principles in order to guarantee a congruent research process from start to finish.
If one of more of the above criteria cannot be met, then consideration needs to be given to the appropriateness of MSR as a suitable research methodology.Conclusion
Modelling Shared Reality is a new tool for qualitative researchers who are interested in taking a ‘snapshot’ of the current reality of a relatively large group. The methodology can generate high-quality in-depth information fairly quickly with a relatively small number of interviews. Because of the way respondents are involved, the research process naturally translates into action initiated by the participants themselves. Currently 15 projects have been conducted using Modelling Shared Reality. The results of those projects have produced a high-level of satisfaction for the research sponsor, and provided a solid basis for policy-making, decision-taking and improving the outcomes of their programmes.References
Brewer, J. D. (2000). Ethnography. Philadelphia: Open University Press.
Giorgi, A. (1985). Phenomenology and Psychological Research. Pittsburgh PA: Duquesne University Press.
Glaser B.G, & Strauss A. (1967). Discovery of Grounded Theory: Strategies for Qualitative Research. Chicago: Aldine Publishing Company
Grove, D. J., & Panzer B.I. (1989), Resolving Traumatic Memories: Metaphors and Symbols in Psychotherapy. New York: Irvington.
Homan, T (2005), Organisatiedynamica - De complexe anatomie van de beweging. Amsterdam: Academic Service.
Kvåle, S. (1996). Interviews: An Introduction to Qualitative Research Interviewing. Thousand Oaks: Sage Publications.
Lawley J. & Tompkins P. (2000). Metaphors in Mind: Transformation through Symbolic Modelling. London: Developing Company Press.
Loftus, E.F. & Palmer, J.C. (1974). Reconstruction of automobile destruction: An example of the interaction between language and memory. Journal of Verbal Learning and Verbal Behavior, 13, 585-589.
Loftus, E. F. & Zanni, G. (1975). Eyewitness testimony: The influence of the wording of a question. Bulletin of the Psychonomic Society, 5, 86-88.
Owen, I. R. (1996). Clean Language: A linguistic-experiential phenomenology in A.-T. Tymieniecka (Ed.) Analecta Husserliana, Vol. 48. pp. 271-297, Dordrecht: Kluwer Academic.
Thibodeau, P.H., Boroditsky, L. (2011). Metaphors We Think With: The Role of Metaphor in Reasoning. PloSOne 6,2.
Tosey, P. (2011). Symbolic Modelling as an innovative phenomenological method in HRD research: the work-life balance project, paper presented at the 12th International HRD Conference, University of Gloucestershire, 25th – 27th May 2011. Full text available at: epubs.surrey.ac.uk/7135/
Wadsworth, Y. (1998). What is Participatory Action Research? Action Research International, 2.
Note 1 Thanks to Roland Hill and Wendy Nieuwland for help with the translation.