1st ASSOCIATION FOR QUALITATIVE RESEARCH
INTERNATIONAL CONFERENCE
Melbourne, Victoria, AUSTRALIA
July 8-10, 1999
‘Rigorous Sex Research:
A Phenomenological Perspective’
Author/Presenter: Russell J Davey
Curtin University of Technology
Slide 1. Abstract
Studies into human sexuality pose particular problems for the researcher who invites participant involvement.
Direct observations may be impossible, and interviewing may be made difficult due to cultural, moral, and/or ethical values. The issue of trustworthiness (rigour) therefore needs close attention if data gathered and findings obtained are to be considered ‘real’ science.
Trustworthiness in phenomenological studies should first be addressed when considering the structural form of the methodology to be employed. Sampling for example, needs must be purposive (random sampling weakens the study). Provision needs to be made for participant verification of transcripts of interviews, and serial interviews (with the same individual) the expected norm.
As phenomenological studies are not driven by hypothesis, but a desire to explicate a given phenomenon (and reveal the ‘essences’ appertaining), the researcher can expect to be deeply immersed in data which may seem obtuse. This initial obfuscation can (and should) be met with an attitude of openness and a willingness on the part of the researcher to allow the phenomenon to present itself. Again, rigorous attention to method is important here, and the phenomenological ‘epoché’ (bracketing) employed so that the revealed in uncontaminated by prior learnings and bios.
This paper offers practical solutions to the establishment and maintenance of trustworthiness in phenomenological research into human sexuality through an examination of a Doctoral study entitled ‘A Phenomenological Study of Sexual Experiences’.
Slide 2. Postgraduate Learnings
As a postgraduate student, I was required to study ‘Health Research Methods’. I quickly learnt that there was only one way to do ‘real’ science, "The Scientific Method". Though there seemed to be several ways of conducting research under this banner, all must subscribe to and satisfy the conditions labeled Validity and Reliability. All data was to be numerically represented (or at least reduced to numerical form) and statistically analyzed. The goal was to formulate and test hypotheses in order to achieve results which could be generalized across populations. ‘Subjects’ were not important in themselves, and should be chosen randomly. Experiments should be controlled and manipulated, and where subjects were involved, 50% should be given no ‘treatment’ (or a placebo). No mention was made of qualitative research methodology at all.
By good fortune, and the intervention of a wise Supervisor, I was granted permission to take a unit in Qualitative Research Methods. This was an incredible discovery for me. To find that there were systems for research that a) respected humans as important sources of data – more participant than ‘subject’ b) recognized that not all data could be (nor should be) reduced to numbers.
I thought that this strange state of affairs (where qualitative methods were ‘hidden’ beneath a barrage of information about quantitative methods – the only way to do ‘real’ science) might be peculiar to my University. It wasn’t, it isn’t. Then I thought that this may be an ‘Australianism’ ie: that other countries (notably the USA and the UK) might be different. It wasn’t, it isn’t. How can this be? It seems obvious to me (and to all the qualitative researchers that I know) that we need both qualitative and quantitative methodologies if we are to do ‘good’ science.
Sadly, this totally unfounded bias toward quantitative methods has meant that a) many ‘mainstream’ educationalists and researchers have little knowledge or understanding of qualitative methods, b) qualitative researchers are often forced to answer questions of trustworthiness in quantitative terms totally ill-suited to the method employed.
Slide 3. Rigour & Trustworthiness
The term ‘rigour’ (within the paradigm of the quantitative) has come to mean ‘strict attention to validity and reliability’. Qualitative methodologists, recognising that the terms validity and reliability are essentially concerned with statistics (Diekhoff 1992; Tesch 1990; Kaplan 1987), have sought alternative terms perhaps more suited to qualitative research (see Table 1).
| Table 1: Criteria which may be used to determine trustworthiness within qualitative and quantitative paradigms (adapted from Koch, 1994, p. 977). | ||
| Quantitative Criteria | Qualitative Criteria | |
| Truth Values | Internal Validity | Credibility |
| Applicability | External Validity | Transferability (fittingness) |
| Consistency | Reliability | Dependability |
During the course of my research, I’ve found the following a useful guideline for establishing and maintaining trustworthiness:
1. Establishing & maintaining trustworthiness
Notes to above items 1-6:
Items 1-5 are derived from Streubert & Carpenter (1995, pp. 46-47) and Koch (1994, pp. 976-978). Item 6 is derived from Marshall and Rossman (1995) who suggest the use of a research ‘partner’ or person who acts as a ‘Devil’s advocate’ by critically examining the researcher’s analyses. Items 1-4 can be seen to address credibility, while numbers 5 and 6 address dependability (determining transferability is the responsibility of the potential user of the findings) (see Koch 1994, p.997; Streubert & Carpenter 1995, p. 26).
Slide 4. Confirmability Audit Trail
Phenomenological endeavours cannot be undertaken without awareness and attention to the issue of researcher bias. Though the phenomenological epoché (‘bracketing’) is employed, it is useful to consider the Confirmability Audit Trail (see Siegle 1999a and Lincoln & Guba 1985).
[Siegle (1999a) reminds us that Confirmability is the degree to which the findings are the product of the focus of the inquiry and not of the biases of the researcher]
Siegle (1999a) wrote that "An adequate trail should be left to enable the auditor to determine if the conclusions, interpretations, and recommendations can be traced to their sources and if they are supported by the inquiry".
Halpern (Lincoln & Guba 1985, and see Siegle 1999a) identified six areas of raw data that need to be reviewed when auditing. These include:
Slide 5. Transferability
Despite sharing the foregoing – that "the terms validity and reliability are essentially concerned with statistics"; that alternative terms are better suited to qualitative research; and that Trustworthiness is an issue in phenomenological studies, Despite listing the ways I have "established and maintain trustworthiness", I have found that some of my colleagues cannot accept phenomenological research (nor any model of qualitative research) as ‘real science’ – I was tempted to say ‘valid science’! – as the data generated "is not generalizable".
The literature makes it abundantly clear that Qualitative researchers do not attempt to generalize their findings. They avoid the word, and prefer to use the term ‘Transferability’ (or "fittingness") which may be defined as "the extent to which the findings can be applied in other contexts or with other respondents" (Siegle 1999b). Indeed, the onus for "demonstrating transferability belongs to … (the reader of the study)" (Siegel 1999b).
Strategies for Increasing Transferability
Phenomenology is known for the richness of participant description evident in the report. At their best, these ‘snippets’ of data reveal deep truths which automatically ‘ring’ within us – they ‘make sense’, they touch us in ways which cause us to feel - to laugh, to feel sad, to identify in some way with the unknown speaker (transferability).
Purposive sampling maximizes the chance of obtaining accurate information about the studied phenomenon for it relies upon choosing those who have both the experience of the phenomenon and also the ability to communicate their experience of that phenomenon.
Slide 6. The 5 Stages to Successful Phenomenological Investigation
So, having looked at Rigour, Validity, Reliability and generalizability, and demonstrated how a Confirmability Audit Trail may be of service, what is my method, and in what ways have I specifically addressed Trustworthiness in my research?
As is not uncommon in qualitative research, my methodology has evolved throughout the study. The methodology settled upon is, I believe, both ‘true’ to the ‘founding fathers’’ philosophical stance, and at the same time a practical, easy to follow model for phenomenological researcher.
The five stage model I developed after extensive readings of phenomenologist literature (both theoretical, philosophical works, and phenomenological research papers), serves very well as a vehicle for promoting Trustworthiness as all aspects previously discussed today are integrated within the model itself.
The major ‘stages’ of phenomenological investigation that are identified and labelled herein Desire, Surrender, Reflect, Repeat, and Complete would stand whether the investigation is researcher only (though sub-stages 1.3, 2.2 and 2.3 would become redundant), or involving participants.
[It is perhaps worthwhile to consider that few phenomenological methodologies mention the need for participants to employ phenomenological ‘bracketing’, yet make essential that the phenomenological researcher do so. I see this as a serious error, perhaps understandably ‘missed’ by phenomenologists who are at base philosophers, or by ‘Americanist’ phenomenologists who are non-critical in data gathering (anything given by participants is to be seen as ‘truth’, to be accepted without question)].
Slide 7. Stage 1: DESIRE
1.1 determine the phenomenon of focus As precisely as possible, we decide what it is that we want to study. Van Manen suggests it is appropriate to ask ourselves ‘what is the nature of this phenomenon … as an essentially human experience?’ (van Mannen 1990, cited in Crotty 1996, p. 35). It may be somewhat less useful to think in terms of ‘what is the common lived experience of this particular group of people?’. Van Mannen enjoins us to examine the phenomenon, whereas the later question may focus our attention on the mundane.
1.2 ‘bracket’ knowledge ‘Bracketing’ or ‘epoché’ is the process whereby all previous knowledge, beliefs, and common understandings about a given phenomenon are ‘set aside’. Participants may need to be schooled to ‘bracket’ in order to meet Husserl’s conditions for ‘epistemological reduction’ else the phenomenological goal (ie: what is it?) becomes obscured rather than revealed.
1.3 carefully select participants (purposive sampling) Participants should be purposively sampled according to the following criteria:
that they have experienced the phenomenon under study,
that they can communicate their experience of the phenomenon as free as possible from embarrassment and bias, and
that they are able to ‘bracket’ their knowledge, beliefs, and common understandings about the phenomenon.
*** Random sampling is inappropriate for phenomenological investigation, and may actually hinder the study (Morse 1992, p. xi). ***
Slide 8. Stage 2: SURRENDER
2.1 ‘open’ to the phenomenon (that is: ‘surrendering’ to the ‘immediate experience’; ‘transcending … the natural attitude’ (Crotty 1996, p. 148); or lifting ourselves above the mundane) Husserl’s writings emphasised immediacy of experience (or the ‘phenomenological experience’), that moment devoid of assumption, prior to thought, desire, and judgement. Bracketing helps us move into that immediacy and beyond the ‘natural attitude’ (or ‘common understandings’). It is as though we are attempting to see with the eyes of a child – involved, fascinated, focused ‘captured’ by the phenomenon, filled with wonder. In the immediacy of the moment, we open ourselves to whatever may be revealed without concern for what we have learnt, what others have said, what may have been theorised or supposed. We ‘surrender’ to what is immediately before us [one participant in the study I’m currently undertaking said that it was very difficult to stay with the actuality of his experience, that his mind kept ‘slipping’ back into what he had been taught he was supposed to feel and think].
2.2 personally interview participants to ensure accuracy. When the roles of interviewer, transcriber, and and analyst are combined, richer and more accurate (full) data can be obtained (interviews consist of not only ‘words’, but other verbalisations and non-verbal communications which may be missed when the interviewer, transcriber, and analyst are different persons).
2.3 ensure participants describe their personal experience of the phenomenon Is the participant ‘bracketing’? Is the participant able to return to the immediacy of experience? When participant responses include statements such as ‘everyone knows’ or ‘theory states’ or ‘I learnt’, what follows is unlikely to be true to the participants personal experience of the phenomenon, but rather interpretations of the phenomenon. Statements which affirm the participant can obviate this problem.
Slide 9. Stage 3: REFLECT
3.1 personally transcribe recordings of interviews to ensure accuracy Interviews should be transcribed by the interviewer (see 2.2). The researchers Journal may be used adventitiously here to record notes about the interview.
3.3 describe what is seen: ensure phenomenological character of description. Describe what comes into view about the phenomenon. Ensure that what has appeared (and therefor what has been described) is solely derived from an immediacy of experience, not ‘coloured’ by theory, knowledge, or bias.
3.3 critically analyse transcripts, and assumptions, perceptions, cognitions, intuitions, insights and interpretations. Phenomenology is both reflective and critical (Crotty 1995). All ‘givens’, experiences and thoughts are examined and questioned in the ever deepening search for ‘truth’ (for discussion on this search for ‘truth’ see Ashworth 1997).
3.4 interpret hidden meanings in transcripts. Though some would claim this is ‘Hermeneutical Phenomenology’, interpretation is commonly found within the wider genre (for discussion on this see Crotty 1996, pp. 75-106; Waters & Crook 1996; Streubert & Carpenter 1995, p. 41; Spiegelberg 1982; 1969).
3.5 explicate those ‘essences’ which delimit the phenomenon. ‘Essences’ are those essentials without which the phenomenon could not exist. Essences are not to be confused with ‘categories’ or ‘themes’ commonly found within qualitative research, they are rather those things that are indubitably associated with the phenomenon – that delimit the phenomenon as ‘this phenomenon’ and not ‘that phenomenon’.
3.6 write a ‘summary’ and/or participant ‘story’ for the purview of the participant. . At this stage, ‘themes’ or ‘categories’ could be developed, though these may be only precursors to the explication of essences.
Slide 10. Stage 4: REPEAT
4.1 return to the participant for verification of the transcript and/or ‘summary’ While transcript verification is important as it addresses ‘credibility’, the writing of a summary or participant story for participant purview, comment, and verification not only strengthens rigour, but recognises the participant as a co-researcher whose contribution is valued and valuable.
4.2 repeat steps 2. to 3. as necessary until the researcher is satisfied that essences from the individual experience have been explicated It is expected that phenomenological investigation will require serial interviews with participants (see McCraken 1988). Though ‘saturation’ has erroneously been associated with phenomenology (the term is used in ‘Grounded Theory’), the concept of saturation – the continuance of data collection until repetition of data is achieved or until no new knowledge surfaces (Streubert & Carpenter 1995, p. 24) – may be useful to bear in mind when undertaking the search for essences in phenomenologically interviewing.
Phenomenology involves reading, re-reading, and reading again all the data gathered. Each reading is an attempt to remove the dross, to reveal the gem buried within. ‘Question everything’ would be a useful tenet to hold when undertaking phenomenological research, together with ‘refine and reduce’.
Each interview should be properly analysed prior to an analysis of an entire series of interviews from one participant. Cross comparisons with other participants should only be undertaken when individual analyses have been completed.
Slide 11. Stage 5: COMPLETE
5.1 re-read all participants’ transcriptions to uncover commonalities, uniquenesses, and phenomenological essences Be prepared to discard tentative essences previously identified in individual analysis (‘question everything’, ‘refine and reduce’). If what has been discovered up to this stage does not hold true across the entire population sample, then it is not an essence.
Once again, and for the final time, we emerse ourselves in the data gathered. Phenomenological research is not easy. It is demanding, repetitious, and critical. At this stage (if not before) it would not be surprising to have such thoughts as ‘I should have undertaken a nice, easy, piece of quantitative research’. Essences may prove illusive, interpretations seem shallow or inaccurate. Do not lose heart. Your hard work is almost complete.
5.2 write a report which includes an explanation of the process of research Now it is time to research what the literature says about the phenomenon under study. Though practical considerations may have necessitated an initial literature review (for students seeking candidacy, or for researchers seeking funding), withholding until this stage may be seen to be adventitious as information gained from literature searching can influence phenomenological investigation (thus the strictures of epoché).
A phenomenological report should be rich with description derived from participant data. Included should be relevant Journal entries together with the results of the literature reviewed. Documentation needs to be such that other researchers will be able to follow the investigative process, and be able to reach similar conclusions given the researcher’s data, perspective, and situation (see Koch 1994, p.978) (this requirement strengthens research as it addresses rigour).
Husserl said to learn phenomenology we must do phenomenology.
Slide 12. In Conclusion
Trustworthiness is an issue of concern that is being addressed by Phenomenologists undertaking research into sexuality.
There are many ways to conduct ‘real’ science.
The use of alternative terms (eg: Trustworthiness; Transferability) which obviate the strictures of those which have now become synonymous with quantitative research (eg. Reliability; Validity; Generalizability) are desirable.
Trustworthiness may be established and maintained throughout a study by close adherence to a structured methodological model.
A Confirmability Audit Trail may be used, not only by the auditor but also by the researcher, to determine if the conclusions, interpretations, and recommendations of a study can be traced back to their sources and if they are supported by the inquiry.
Strategies for increased transferability are an integral part of phenomenological endeavour. Reports should be rich with participant transcription, and sampling needs must be purposive.
References:
Siegle, D. 1999a, Confirm.html June, from WWW
Siegle, D. 1999b, Transfer.html June, from WWW
Diekhoff, G. 1992, Statistics for the Social and Behavioral Sciences: Univariate, bivariate, multivariate, Wm C. Brown Publishers, Dubuque, IA.
Kaplan, R. M. 1987, Basic Statistics for the Behavioral Sciences, Allyn & Bacon, Inc, Newton, Massachusetts.
Koch, T. 1994, ‘Establishing rigour in qualitative research: The decision trail’, Journal of Advanced Nursing, vol. 19, pp. 976-985.
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Newbury Park, CA: Sage.
Marshall, C. & Rossman, G. B. 1995, Designing Qualitative Research, 2nd edn, Sage Publications, Thousand Oaks, California.
Streubert, H.J. & Carpenter, D. 1995, Qualitative Research in Nursing: Advancing the Humanistic Imperative, J.B. Lippincott Company, Philidelphia.
Tesch, R. 1990, Qualitative Research: Analysis types and software tools, The Falmer Press, Basingstoke.