Methods: Interpretative phenomenological analysis

If you study, research or teach within the applied psychologies, you will no doubt have heard of the fashionable qualitative method interpretative phenomenological analysis (IPA: Smith et al., 1999; Smith & Osborn, 2003). Indeed, between the years 1996 and 2008, there were 294 empirical papers published, indicating the increasing popularity of this methodology (Smith, 2010).

Conceptualised in the mid-1990s (Smith, 1996), this methodology could be argued to have already assumed a dominant position in qualitative research (Smith, 2010; Willig, 2008). Ultimately, this rise in popularity will have some effect on the quality of published IPA studies and student work alike. There is a clearly visible community associated with IPA that includes a website and discussion forum, regional IPA groups, an annual conference and a range of training workshops from basics to advanced to training for trainers.

With this in mind, and drawing on our varied experience conducting and supervising IPA research as well as facilitating regional IPA research groups in both Scotland and London, this article will tackle a number of pertinent issues that have developed due to this rise in popularity. These issues will be relevant for researchers, supervisors and examiners of IPA.

We are addressing a perhaps surprisingly wide audience here. Despite the early emphasis on health psychology, we have noticed an increase in the diversity of the student population attending the London Regional Group that has now branched out into other applied psychologies, primarily clinical, counselling, educational and occupational (Smith et al., 2009). This is reflected in recent publications such as Todd et al. (2010), Rizq & Target (2008) and Millward (2006). Interestingly, there is a conspicuous absence of peer-reviewed educational psychology articles.

IPA also appears to have become the 'default' option for many students at many levels. This tends to result in poorly constructed, primarily descriptive projects that do not reflect good-quality IPA. Why are so many students 'defaulting' to IPA? We suspect that for postgraduate students in the applied fields of health, clinical and counselling, IPA's focus on subjective lived experience is something that intuitively appeals. For undergraduates, we would suggest that an apparent lack of training in a broad variety of qualitative research methods creates a situation in which IPA becomes the default option. In addition, there appears to be a general misapprehension that IPA is simply a form of thematic analysis with little emphasis on interpretation and is therefore, dare we say it, the easy option (see Braun & Clarke, 2006, for an excellent overview of thematic analysis).

Amongst students, we have seen distinct and persistent issues emerge over the years that relate to the planning and conducting of IPA research. Students tend to choose the methodology in advance of developing the research aims and questions. The project is then shoehorned into the methodology, with students failing to understand that they have effectively engaged in the research process in reverse. Students are then also unable to comprehend why or how this might generate problems in executing the research project (Barker et al., 2002; Punch, 2006).

There is also an apparent lack of understanding in both students and supervisors that IPA is primarily an interpretive approach. This misconception, accompanied by a lack of confidence in raising the level of interpretation in analyses, results in broadly descriptive IPA that lacks depth and therefore demonstrates little difference to a standard thematic analysis. This does not represent good IPA (Smith, 2010).

Given these issues, good-quality supervisory input is clearly paramount. Unfortunately, attendance patterns and numbers in the London Regional Group, along with frequent private requests from students for supplementary research supervision, suggest that it is not always forthcoming. We now turn to common misconceptions in the 'supervisory space' between student and supervisor, which in our experience appear to create multiple difficulties in producing good IPA.

More is not always more

Students consistently appear to experience pressure to include too many participants, seemingly in order to placate research boards and supervisors in line with the quantitative monopoly within academic research. This necessarily de-emphasises IPA's commitment to idiography. As first suggested by Reid et al. (2005) less is more in IPA: fewer participants examined at a greater depth is always preferable to a broader, shallow and simply descriptive analysis of many individuals, as commonly seen in thematic analysis, grounded theory or poor IPA. Smith et al. (2009) highlight the fact that sample size is contextual and must be considered on a study-by-study basis. However, as a rough guide they suggest between three and six participants for an undergraduate or master's level

IPA study and from four to ten data points for professional doctorates, with recommendations for PhD becoming less easy to explicate. Following on from these guidelines, and our experience of student difficulties in this area, there appears to be a clear need for supervisors or academic boards to embrace IPA's idiographic commitment in terms of smaller sample sizes, perhaps focusing more on the adage that 'less is more' in IPA.

'Where's your control group?'

It is a common misconception that it is desirable for a novice to compare groups using IPA, and that this would be reflective of good practice. Making a comparison within an IPA study is actually quite difficult to achieve within the confines of most student IPA as it requires greater numbers of participants and therefore tends to result in studies that are primarily descriptive and lack depth. Smith et al. (2009) encourage a less ambitious project for beginners, focusing on perhaps comparing one dimension in a single group, thus maintaining a deeper level of analysis with a more interpretative focus. IPA's focus on convergence and divergence within a participant group's experience of a phenomenon naturally requires comparisons at the individual level (comparing case to case). Participants are purposively recruited to present a homogeneous sample in terms of the particular research topic.

Too many questions

A problem which students continually bring to the Regional Group is difficulty constructing a good quality semi-structured interview schedule. Students tend to produce schedules that are too long, overly extensive and detailed, and therefore constraining. This appears to arise from an erroneous belief and fear that topic areas of interest to the researcher will not be covered. In our experience however, on the whole, these fears are unfounded as these areas can be addressed later on in the interview process once the participant has expressed their interpretation of their lived experience. Gathering good-quality data for IPA asks for a more open-ended interview maintaining a careful balance between guiding and being led. Thus, interview schedules should be short, starting with broad, general questions that allow the participant to set the parameters of the topic, not the other way around. This is so that the researcher does not impose their understanding of the phenomenon on the participant's narrative (Smith et al., 2009).

Simply describing is not enough

As mentioned above, it is evident that students and supervisors tend to fail to develop the analysis to a sufficient interpretative level. Alongside this there is an inclination for students to present too large a number of descriptive superordinate and subordinate themes with insufficient data extracts presented to support each theme. At the other end of the scale, students may present a large number of short descriptive quotes in an attempt to demonstrate good frequency for a theme. However, once again, this tends to result in a lack of depth that does not represent good IPA.

A tip here is that there is no hard-and-fast rule for how many themes are appropriate for what size of project, but that a smaller number of themes tends to represent a more thorough and synthesised analysis. There are a number of theoretical papers that demonstrate how higher levels of interpretation might be achieved (e.g. Smith, 2004), along with a number of examples of good-quality published empirical IPA that demonstrate a depth of analysis with a good level of interpretation (e.g. Chapman et al., 2007; Dickson et al., 2008; Eatough & Smith, 2006; Smith & Osborn, 2007).

Deeper reflection on quality in IPA

Evaluating quality in qualitative research is a bone of contention within academia in general. This is clearly reflected in an apparent lack of understanding of how to demonstrate validity in an IPA research project. We would suggest that Yardley (2000, 2008) and Elliott et al. (1999) represent useful frameworks within which to evaluate and demonstrate the validity of IPA research. We recommend that both students and supervisors familiarise themselves with these guidelines and use them to explicate the validity of their IPA (Smith, 2010). Quality in published IPA is addressed specifically in Jonathan Smith's 2010 keynote at the British Psychological Society's annual Health Psychology Conference at Aston University and recent article in Health Psychology Review (Smith, 2011). This issue is also comprehensively addressed in Smith et al. (2009), drawing on Lucy Yardley's framework for validity in qualitative research (Yardley, 2000, 2008). This involves attention to four broad principles: sensitivity to context; commitment and rigour; transparency and coherence; and impact and importance.

We also recommend reflecting on two of the publications we mentioned above for guidance and clarity regarding what constitutes good-quality presentation of IPA research. In their study of identity crisis, loss and adjustment amongst people living with chronic fatigue, Dickson and colleagues (2008) present a manageable number of master themes (three) demonstrating reduction and engagement with the data. We learn the nuances of Anne, Bartholomew and the others via the author's inclusion of multiple illustrations from the same participant (e.g p.464 'And later on…'). Furthermore, no quote is left to itself – it is embedded in the persons account and discussed with interpretation by the researcher. The researcher also attempts to demonstrate both the individual's and the group's experience intertwining each person's account with the others (e.g. p.468 'Or for Angie…'). Smith and Osborn (2007) offer a similar in-depth window into the lives of six patients with chronic benign lower-back pain. Using Helen and her experience as an example of the common thread between the group and the phenomenon (p.522), the authors carefully weave in the other five participants' quotes (e.g. Kevin, p.523) thereby giving the patients a voice, whilst still making their interpretative voice clear and distinct (e.g. 'Notice how…' p.526).

The desire to generalise

In our experience, there is an apparent expectation from both students and supervisors, in line with traditional scientific psychology, that generalisability is the ultimate goal of any research. Within IPA, and the qualitative paradigm in general, there is more of a focus on the possible transferability of findings from group to group rather than generalisation. Smith et al. (2009) also argue for 'theoretical generalisability', where the reader may be able to 'assess the evidence in relation to their existing professional and experiential knowledge' (p.4). Thus it can be argued that idiographic qualitative research such as IPA has much to contribute to our understanding of phenomena, as it can complement actuarial claims derived from quantitative studies through a focus on the particular which can help illuminate the universal (Warnock, 1987).

In conclusion, the overarching theme that appears to emerge from our interpretation of teaching, supervising and undertaking IPA, is again that 'less is more'. Our advice to students and supervisors is to include fewer participants in the sample; fewer questions in the interview schedule, and fewer superordinate and subordinate themes in the analysis. In observing these simple guidelines researchers should achieve more depth in their IPA.

A model of a Regional Group

The London Regional Group was founded at Birkbeck, University of London by Virginia Eatough in 2005, and it has gone from strength to strength. We assumed facilitation of the group in March 2009. In its present format we meet every other month to cover a range of activities including troubleshooting various problems and queries relating to research design, interview schedules, the analytical process, reflexivity, epistemological issues and debates, recent research, and validity checks on emergent themes and analyses. Occasionally, we also pre-read and discuss theoretical papers relating to IPA, phenomenology and qualitative inquiry in general. On the whole, we find attendees range from undergraduate and postgraduate through to academics. In addition, the group has attracted research- active professionals from disciplines as diverse as architecture and commercial market research.

In line with Hanley and Lennie (2008), we believe that group members value practical tips and the support provided by those in a similar situation to themselves. The group demystifies the process of conducting IPA research and can help consolidate current understanding of the theoretical and practical issues involved. This group can provide an excellent model for the provision of an additional support network during what can be a very isolating time when conducting and writing up research. In addition to this, IPA groups can also provide the opportunity for theoretical and philosophical advancement. We encourage all to attend: researchers and supervisors, students and academics alike. Use the IPA community to develop as an IPA researcher, by:

- using the resources and information

- attending your Regional IPA Group or setting one up in your area if it does not already exist (check the website for details of your local group)

- engaging with the internet discussion forum (details of how to join can be found on the IPA website)

- attending training sessions at various universities across the UK and Ireland, ranging from introductory to advanced levels, and more recently offering training for trainers/supervisors

- attending the annual IPA conference. (keep an eye on the forum and website for details)

- presenting your work as much as possible to Regional Groups and conferences.

Other ideas for the future facilitation of teaching and learning in IPA might include more one-off teaching events at specific key moments in the research process. For example, this could include preparatory workshops prior to data gathering, 'analysis clinics' where students work on their own data in small groups that are facilitated by tutors, and writing up seminars during the final stages. Currently, there is a great demand for online courses from international and home students alike. Courses such as these might offer students increased access to good-quality teaching and IPA expertise as and when is convenient for them in their studies. As yet, this type of online option has not materialised; however, we believe that it would be very useful in facilitating self-directed learning from any location around the world.

Misunderstood and misapplied

In our experience, undergraduate teaching of research methods tends to emphasise the quantitative over and above the qualitative paradigm, leaving students poorly versed in qualitative methods such as IPA. While this imbalance appears to be changing, there is no doubt that undergraduates might find useful an archive of previous qualitative projects conducted at this level that they can draw on to guide their research.

Ultimately, and in conclusion, we argue that IPA is still a misunderstood and misapplied methodology. We believe that a better understanding from both sides of the research coin (i.e. supervisor and supervisee) of the common misconceptions as described above can facilitate the production of high-quality IPA. As demonstrated, there are numerous support systems in place for researchers using IPA; nonetheless, it is still apparent that there is great demand for further innovative and easily accessible IPA teaching and learning opportunities for students across all levels.

Kate Hefferon is a Chartered Psychologist and Senior Lecturer and Acting Programme Leader for the MSc in Applied Positive Psychology at the University of East London.

Elena Gil-Rodriguez is a Chartered Psychologist and a Lecturer in Counselling Psychology at the University of Roehampton.

References

Barker, C., Pistrang, N. & Elliott, R. (2002). Research methods in clinical and counselling psychology (2nd edn). New York: Wiley.

Braun, V. & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3, 77–101.

Chapman, E. et al. (2007). Psychosocial issues for patients with ventricular assist devices. American Journal of Critical Care, 16, 72–81.

Dickson, A., Knussen, C. & Flowers, P. (2008). 'That was my old life; it's almost like a past-life now': Identity crisis, loss and adjustment amongst people living with chronic fatigue syndrome. Psychology and Health, 23, 459–476.

Eatough, V. & Smith, J.A. (2006). I feel like a scrambled egg in my head: An idiographic case study of meaning making and anger using interpretative phenomenological analysis. Psychology and Psychotherapy: Theory, Research and Practice, 79, 115–135.

Elliott, R., Fischer, C. & Rennie, D. (1999). Evolving guidelines for publication of qualitative research studies in psychology and related fields. British Journal of Clinical Psychology, 38, 215–229.

Hanley, T. & Lennie, C. (2008). Setting up a student research group. The Psychologist, 21, 129.

Millward, L.J. (2006). The transition to motherhood in an organizational context. Journal of Occupational & Organizational Psychology, 79, 315–333.

Punch, K.F. (2006). Developing effective research proposals (2nd edn). London: Sage.

Reid, K., Flowers, P. & Larkin, M. (2005). Exploring the lived experience. The Psychologist, 18, 20–23.

Rizq, R. & Target, M. (2008). The power of being seen. British Journal of Guidance & Counselling, 36, 131–153.

Smith, J.A. (1996). Beyond the divide between cognition and discourse. Psychology & Health, 11, 261–271.

Smith, J.A. (2004). Reflecting on the development of interpretative phenomenological analysis and its contribution to qualitative psychology. Qualitative Research in Psychology, 1, 39–54.

Smith, J.A. (2010, September). Evaluating the contribution of interpretative phenomenological analysis to health psychology. Keynote address presented at BPS Health Psychology Annual Conference, Aston, UK.

Smith, J.A. (2011). Evaluating the contribution of interpretative phenomenological analysis. Health Psychology Review, 5(1), 9–27.

Smith, J.A., Flowers, P. & Larkin, M. (2009). Interpretative phenomenological analysis: Theory, method and research. London: Sage.

Smith, J., Jarman, M. & Osborn, M. (1999). Doing interpretive phenomenological analysis. In M. Murray & K. Chamberlain (Eds.) Qualitative health psychology. London: Sage.

Smith, J. & Osborn, M. (2003). Interpretive phenomenological analysis. In J.A. Smith (Ed.) Qualitative psychology: A practical guide to research methods (pp.51-80). London: Sage.

Smith, J.A. & Osborn, M. (2007). Pain as an assault on the self: An interpretative phenomenological analysis of the psychological impact of chronic back pain. Psychology and Health, 22, 517–534.

Todd, D., Simpson, J. & Murray, C. (2010). An interpretative phenomenological analysis of delusions in people with Parkinson's disease. Disability Rehabilitation, 32, 1291–1299

Warnock, M. (1987). Memory. London: Faber.

Willig, C. (2008). Interpretive phenomenology. In C. Willig (Ed.) Introducing qualitative research in psychology (2nd edn, pp.50–69). Philadelphia: Open University Press.

Yardley, L. (2000). Dilemmas in qualitative health research. Psychology and Health, 15, 215–228.

Yardley, L. (2008). Demonstrating validity in qualitative psychology. In J.A. Smith (Ed.) Qualitative psychology: A practical guide to methods (2nd edn, pp.235–251). London: Sage