Windows into grounded theory

Anne Mottram on Use of the literature in grounded theory research

I became aware of grounded theory during my undergraduate days when we studied “Awareness of Dying”, a seminal work by Barney Glaser and Anselm Strauss published in 1965. This study was influential on starting a debate within the nursing and medical professions on the importance of rethinking the silence of professionals towards the dying patient. Of particular interest was that this was the first study to apply the principles of grounded theory: namely that theory should emerge from the data, and that fieldwork is essential to allow researchers to immerse themselves into social settings from which theory will develop.

The authors (Glaser & Strauss, 1965) stress the importance of the researcher keeping memos to aid theory development, when working in the field.  The main goal of memo writing is “to theoretically develop ideas” (Glaser, 1978, p. 83) and to raise theoretical awareness. Glaser (1978) provides practical advice, suggesting that when a researcher has a good idea they should immediately write a memo on it - it need only be a sentence or a short paragraph. Glaser (1978) also contends that if researchers do not write memos, that they are not doing grounded theory.

With the above in mind I will describe the first memo I wrote in the early stages of fieldwork of a study which examined patient experiences of day surgery. A man, 40 years old, was attending the pre-operative assessment clinic to ensure his fitness to undergo surgery and discharge on the same day. Whilst awaiting his appointment he would not sit down but paced up and down talking into his mobile phone. When he momentarily paused from his telephone conversation I asked would he agree to take part in the study. He readily agreed. With his permission I attended the pre-operative assessment with him where he hurriedly answered the assessing nurse’s questions. Following this, he left the clinic quickly and commenced another call on his mobile. I had to run down the corridor after him to ask him to sign a consent form for me to contact him following his surgery. He signed the consent form without reading it and went on his way.  On the appointed day of surgery he did not attend nor did he respond to telephone calls from the clinic to see if he wanted to re-schedule.

I felt very disappointed that I had ‘lost’ a research participant. I mistakenly felt that I had no data worth reporting. I wrote a memo which, at the time, appeared to be of no consequence. The memo recorded my perceptions of the discomfort of the patient, the importance of his mobile phone, the desire to get away from the hospital. Later I found that this memo contained an important idea, which when compared to memos relating to other patients, gave rise to an important theoretical recognition related to the concept of speed and time and the desire of patients to have control over their habits, routines and timetables. As Glaser (1978) states, memos are vital as they provide “a bank of ideas which can be revisited in order to map out the emergent theory “(p.84).   

References

Glaser, B. (1978). Theoretical sensitivity. San Francisco: University of California.

Glaser, B., & Strauss, A. (1965). Awareness of dying. Chicago, IL: Aldine Publishing.

Wendy Johnson on Memoing in a grounded theory study

I come from a background of journaling and have always enjoyed writing. I’d write about things that were interesting to me on a very basic level and often without any actual depth to it. There was no particular structure to my writing as it covered everything on a superficial level. For years this type of writing worked for me. So when I was introduced to memo-writing in research, I immediately believed that it comprised the same type of format – basic and without depth.

I started memoing very early in the research process, even before my research question was fully developed. I used the same format for writing as I had in journaling. Initially, I didn’t see the importance of focusing on one concept at a time. I just wanted to write about my research and include everything in that memo. What I found was that each memo consisted of tightly packed ideas that made it almost impossible to put them into any type of concrete category. The memos were superficial and without depth just as my journaling had been for years. Nuances that I had hoped would emerge from the memos were barely recognizable.  

I decided to go back to the beginning and try to get a better understanding of memoing and how it was different from journaling. I knew that this was an important first step as I needed to fully understand the purpose of memoing in research and the various types of memos that I could use. I also knew that if I didn’t master the art of memo-writing early on in the research process that it would only become much more difficult for me later when I tried to develop codes. I began by putting a date and a title on each memo, even before I wrote the first word. This step helped me to focus on one idea at a time and to make it easier on me later when I placed each memo into a specific category. I then grouped the categories developed from the memos together and began looking for relationships between them. With this next step I was able to see the nuances amongst the categories that I had searched for earlier. This step also provided me with the opportunity to see how my thinking had deepened and progressed as each memo was specifically targeted to one concept. As I progressed in my research study, my memo-writing became fine-tuned and I was able take the concepts that had developed and use them in more productive ways. I was able to lay out the concepts, build them up, deconstruct them and then finally reconstruct them in preparation for the higher levels of analysis I had sought for earlier in my writings. 

Helen Hall on Using data from non-participant observation 

Experience has taught me the value of seeking out a variety of data sources to gain deeper insight into my area of interest; midwifery practice.  Indeed Glaser’s dictum ‘all is data’ reflects the fundamental grounded theory position that the researcher can, and should, gather any form of data they believe will be useful.

While interviews and focus groups give voice to the participants and are commonly used, I also found that non-participant observation can play a valuable role. Non-participant observation literally places the researcher where the action is, without significantly influencing it.  Data gathered in the field can enable grounded theorists to gain a deeper insight into the complexities of human action and interaction.  Furthermore, observational data can illuminate any disparity between the participant’s words and their actions. Similar to other researchers, I have found that individuals are not always cognisant of subtle mediators that impact their behaviour. For example, in a study exploring the use of complementary therapies in the maternity setting it became evident through observation, that regardless of the midwives’ attitude and beliefs, the context of their clinical work had a significant impact on their behaviour.

 I recommend that observations be used in combination with interviews, to ensure that the researcher is not misinterpreting the participant’s experience. Follow up interviews also provide a valuable opportunity to explore any inconsistencies between participants’ words and actions.  I have found it works well if I first interview several participants, observe one or two in practice and then use theoretical sampling to guide where I should collect data from next. For example, I interviewed midwives about their attitudes and behaviour around the use of complementary medicine in practice and then I observed some of them interacting with women and colleagues in a variety of antenatal settings including clinics and child education classes. During observational periods I took detailed field notes in order to record important characteristics of the exchange and I sought clarification from the participants following.

However, I must highlight that while observation of participants in their natural environment may provide valuable insights, it is not always appropriate or feasible. Apart from constraints around time and other logistics, many areas of interest may not lend themselves to observation. For example, observation of midwives interacting in the birth suite may be inappropriate, while interviewing them retrospectively about their experiences may provide valuable data. 

When using grounded theory methodology, I aim to enter into participants’ worlds and to represent it through their eyes. I have found that combining the use of interviews/ focus groups, to interpret the participant’s words, with non-participant observations in the field, to interpret their action, has enabled me to gain insight into the phenomenon from different perspectives and I would encourage other grounded theorists to explore the possibilities. 

Patrizia Schettino on Using diagrams in grounded theory 

Strauss and Corbin (1998) define diagrams as “visual devices that depict the relationships among concepts” (p.217). They can be considered “visual memos” (p.217) and as such can be produced, from simple to complex, during the entire research process. Diagrams provide a visual representation of relationships between categories and between categories and subcategories. Diagrams show the density and complexity of the theory. Charmaz (2006) refers to diagrams as tools that can enable the researcher to see the “power, scope and directions of categories and the connections among them” (p.118).

In my research, diagrams are not a secondary way of presenting study results; they are central to the process. I use them as visual memos to define my research question and during all the stages of the analysis process. I am a designer and a researcher of design and digital heritage, so for me it is the most natural thing to do to produce diagrams to visualize my interpretation of data.

In an exploratory study I conducted prior to a pilot study my initial diagram was very simple. In it I basically visualized the change of my research question, from “How can designers take cultural diversity into account in their creative process” to “How can we describe the mediated designer-user communication process”. The diagram included representations of key players (designer, user) and the process itself, as it emerged from the exploratory study. After conducting field work, I produced a further 15 diagrams.  These were grounded in my data and were used to explore and represent relationships between categories. I often started by drawing them on paper and later making a digital copy.

I tried to use the modeling tool in NVivo but found it to be too basic compared to other more advanced software for visualizing maps, such as Visio, Power Point or Illustrator. It would be a great improvement if NVivo had a more advanced modeling tool. My research focuses on processes, actions and interaction and one of the key results of my PhD research was a model of mediated communication designer-visitor-customer service in museums in 11 stages. The first draft of this model, grounded on the first part of the data analysis, was shared both with designers and museum staff members. To share diagrams and models with study participants and stakeholders is a way of validating models: this validation strategy is called ‘participant validation’. Sharing diagrams also contributed to building relationships with stakeholders, enabling me to communicate my research using a visual tool, stimulating communication and collecting comments from participants about my visualization of the design process and my own understanding of processes inside the museum. Later, I supplemented participants’ comments in a more advanced version of my diagrams and built a more advanced version of my model. This version of the model was shared with the museum and was presented at an international conference (Schettino, 2013). As I produced around 15 diagrams in my study, a key objective was to develop a final model, which connected all of them. I wanted a model that provided an overview of the complexity of the process, without compromising legibility due to complexities, but also without oversimplifying what emerged from the data.

Reference

Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative research.

Schettino, P. (2013). Emotions, words and colors: A strategy to visualize and analyze patterns from visitors' narratives in museums. Paper presented at the Information Visualisation (IV), 2013 17th International Conference, London.

Oliver Thomson on The use of non-participant observation and video-prompted reflective interviews to facilitate theoretical sufficiency 

Through engaging with the key grounded theory literature it became clear to me, early in the research process, that a central requirement for developing a theory with explanatory power was to reach a point of theoretical saturation in data analysis. The term ‘theoretical sufficiency’, as offered by Dey (1999), fitted better with the subjectivist epistemology of my study. In this regard, data served to suggest categories rather than saturate them. Early data analysis followed an iterative process of reading and re-reading transcripts from individual interviews with purposively sampled participants (often while listening to the interview audio-recordings), coding and writing copious analytical and reflexive memos. Staying true to these procedures meant that mid-way through my study I had obtained sufficient theoretical purchase on data analysis and I had developed theoretical insights and ideas with shape and structure. By this point, I moved to theoretical sampling to collect specific data relevant to my theoretical ideas. Theoretical sampling needed to facilitate the sufficiency of categories which would ultimately add to the depth, breadth and explanatory power of my theory.

When using theoretical sampling I not only had to decide what data to collect and who from, but also how data was going to be generated to progress my developing theory. Part of my theoretical sampling strategies involved moving from individual interviews to non-participant observations and video-prompted reflective interviews with new participants. By this point in the study I had become so immersed in the data and theoretically sensitive, that observing participants enabled me to make theoretical connections between what previous participants had said during individual interviews and what I saw during observation sessions. Non-participant observation provided an opportunity to test my theoretical ideas whilst also offering new perspectives and deeper analytical insights. Using video-recording as a reflective tool during interviews served to remind participants about prior observed sessions and encouraged them to reflect more deeply about their practice and decision-making. This ensured that their answers to questions were closely tied to their actions, which took place during the clinical appointment. For example, participants talked through their specific thinking and reasoning immediately after viewing (and at times whilst viewing) aspects of their clinical consultation on the video-recorded footage, and provided specific detail about why they performed a particular clinical procedure/action in a particular way.

I decided to use video as a reflective modality to prompt participants during interview (which took place immediately after the observation session), rather than as a specific source of data to be analysed. Using the video-recording in this way helped participants to articulate the connections between their knowledge and action, which was proving difficult to access through previous interviews alone (Haw and Hadfield 2011), and generated rich, context- and situation-specific data. Observing and video-recording participants in their natural setting enabled me to ‘see’ the developing ‘theory in-action’ and highlighted similarities and differences compared with the ‘espoused theory’, which was developed from previous interviews (Argyris and Schön 1974). Comparing data with data in this way, was an essential part of my constant comparative method of analysis (Charmaz 2006). Incorporating observation and video-prompted reflective interviews helped move major categories towards theoretical sufficiency and develop the relationships between them, thereby enabling me to construct a substantive theory with explanatory power.

Argyris, C. and D. A. Schön (1974). Theory in practice: Increasing professional effectiveness, Jossey-Bass.

Charmaz, K. (2006). Constructing grounded theory. London ; Thousand Oaks, Calif., Sage Publications.

Dey, I. (1999). Grounding grounded theory : guidelines for qualitative inquiry. San Diego, Academic Press.

Haw, K. and M. Hadfield (2011). Video in Social Science Research: Functions and Forms, Routledge.

Intan Nurjannah on Translation and analysis of interview data in a grounded theory study

In my study titled “Connecting care for individuals living with a mental health issue in Indonesian: A grounded theory study”, the participants were Indonesian so the raw data was in Indonesian language. This raised an important issue during processes of analysis, as not all members of the research team understood Indonesian. A translation process was therefore developed, so that they could be involved in data analysis.

The large quantity of data, including memos, to be analysed led to a unique translation process. Moreover, using gerunds (verbs ending in ‘ing’) when coding data, which is a guideline suggested by some grounded theorists, required consideration, as this does not exist in Indonesian. Additionally, it was crucial to be cautious in the translation process, to ensure that the theory was ‘really grounded’ in the data. There were three key issues to be considered when developing the translation process: how, who and when would the translation be conducted.

In addition, it was important to understand the underlying philosophical foundations of the study. The researcher’s philosophy is that ‘the truth concept is a relative stance’, meaning that the researcher acknowledged that theory exists through a process of theoretical development. With this in mind, the researcher chose to use two party translation: the researcher herself and a professional translator. The researcher conducted ‘coding’ translation, which formed the basis of the developing theory from the raw data. The professional translator focused on the translation of participants’ quotes, and this was conducted when the study was nearing completion.

The choice of words used in the translation process was also a challenge. This was because some English words are not necessarily interpreted in the same way when translated into Indonesian. The researcher found that there were some words, which have strong associations with grounded theory. For example, the term ‘emerged theory’ in Indonesian does not have as deep a meaning as its English counterpart, which is strongly associated with the Glaserian variant of grounded theory. It is similar with the term ‘developed theory’, which in English is associated with the Straussian variant of grounded theory.

More information about how the translation process works to ensure the integrity of translation results can be accessed via: Nurjannah, I., Mills, J., Park, T. & Usher K. (2014). Conducting a Grounded Theory Study in a Language Other than English: Procedure for Ensuring the Integrity of Translation. SAGE Open: 1-10 DOI: 10.1177/2158244014528920).

Huibrie Pieters, Katrina Dornig & Tomoko Iwaki on Collaboration in grounded theory analysis

Doctoral research is an independent and sometimes lonely process by nature, even with the best mentorship.  As advanced doctoral students, the first two authors co-created a supportive dissertation analysis partnership to reflect on our thinking and propel us forward in our data analysis and writing.  We wanted to share our experience of how we created this collaborative partnership to help other doctoral students navigate the dissertation process. 

It is, of course, of utmost importance to find a partner who is a good fit.  A good fit is not necessarily someone exactly like yourself.  A good partner will have different skills and attributes that will push your analytic thinking forward.  Take the time to choose a partner carefully.  Some important considerations are your expectations, needs, and goals and what contributions each of you bring to the analytic dyad.  For example, we were at similar points in our dissertation research and thus, our analytic needs and timeframe for analysis and writing were really compatible.  We were also both using the constructivist grounded theory methodology for our research.  Consider similarities and differences between you and your potential partner and whether they would be helpful or a hindrance.  We both brought different expertise and experiences that complemented both of our research. 

Once you have chosen a partner and are deciding how to proceed with meetings and work sessions, it is critical to pay attention to structure and process.  Set up regular meeting times in a place that is convenient and conducive to work for both of you.  Have a clear agenda for your meetings and respect each other’s work by being on time, prepared, and turning off all other distractions such as cell phones.  We found that approaching analytic discussions as a space to push each other’s thinking forward rather than as a space to find the ‘right’ answer more encouraging of creative thought, and ensured that we built a trusting relationship.  Check in with each other on a regular basis to ensure that the meetings are mutually productive, and are working, both structurally and intellectually.  As you progress through your analysis, your analytic needs may change so allow for flexibility.  We did not memo on our collaboration, but reflecting back, it would have been helpful to do so to capture our thoughts on what was working and what could be changed to be more useful.  More details, as well as a useful map for creating a successful analytic dyad, can be found in our original article (Pieters & Dornig, 2011).

Our relational support provided the space for us to interact with the data in a deeper way, to contain the complexities, ambiguities and multiple perspectives of our data.  Though data analysis was the driving motivation for creating our analytic dyad, our partnership evolved to also be emotionally supportive and through this process we have since become friends.  We hope that other doctoral students, early career researchers, and mentors can use our experience to make the dissertation process a shared and supportive journey.      

Reference

Pieters, H.C., & Dornig, K.D. (2013). Collaboration in grounded theory analysis:
Reflections and practical suggestions. Qualitative Social Work, 12(2), 200-2014. doi:10.1177/1473325011424085