Revista Colombiana de Psiquiatría ISSN: 0034-7450 [email protected] Asociación Colombiana de Psiquiatría Colombia
Syed, Rebecca J. A Guide to Planning Qualitative Research Revista Colombiana de Psiquiatría, vol. XXXV, núm. 3, 2006, pp. 419-427 Asociación Colombiana de Psiquiatría Bogotá, D.C., Colombia
Available in: http://www.redalyc.org/articulo.oa?id=80635310
How to cite Complete issue Scientific Information System More information about this article Network of Scientific Journals from Latin America, the Caribbean, Spain and Portugal Journal's homepage in redalyc.org Non-profit academic project, developed under the open access initiative
A Guide to Planning Qualitative Research
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○
A Guide to Planning Qualitative Research
Rebecca J. Syed1
Abstract
Introduction: Research in psychiatry overwhelmingly relies on quantitative methodology. However, qualitative research may be an invaluable tool in psychiatry. Its potential under utilisation is multi factorial: it is still regarded by some as a “soft option”, relatively few qualita- tive papers are accepted for publication in medical journals, some assert that health researchers would benefit from training in or collaboration with appropriately trained social scientists and numerous social scientists have produced various checklists, guidelines or manuals for qualitative research, which can be confusing and even contradictory. Aims: This article seeks to describe how to plan a qualitative study, with the aim of reducing apprehension and encouraging further qualitative research in the field of psychiatry. Methods: The main considerations when planning qualitative research will be outlined particularly with regard to: the question and background of the research area, the underlying philosophy, bias, ethical consideration followed by a nuts and bolts description of the methods. Discussion: There are many potential pitfalls in conducting qualitative research. These may involve poor planning, attempting unselectively to fulfill all of the quality criterion suggested in all the various checklists, or under estimating the resources necessary to meaningfully analyze qualitative data. Conclusion: With adequate planning, training and supervision, qualitative methodology is a valid, useful and publishable tool for those researching within the field of psychiatry.
Key words: Qualitative research, depression, adolescents.
Título: Guía para planear una investigación cualitativa.
Resumen
Introducción: Aunque la investigación en psiquiatría se apoya predominantemente en metodologías cuantitativas, la investigación cualitativa también constituye una herramien- ta invaluable. Sin embargo es subutilizada por diferentes causas: algunos la consideran una opción “blanda”, se publican pocos trabajos cualitativos en revistas médicas, algunos afir- man que los investigadores en salud se beneficiarían al trabajar en colaboración con los de las ciencias sociales, y las ciencias sociales han producido encuestas, guías y manuales de investigación cualitativa que pueden ser confusos o incluso contradictorios. Objetivos: Des- cribir cómo planear un estudio cualitativo para reducir la aprensión y estimular la investi- gación cualitativa en psiquiatría. Métodos: Las consideraciones principales al planear una investigación cualitativa se definirán particularmente por la pregunta y el trasfondo del área de investigación, la filosofía subyacente, los sesgos y las consideraciones éticas, seguidas de
una descripción básica de los métodos. Discusión: Existen riesgos potenciales al llevar a
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○ 1 B.A., M.A., MRCPsych, Visiting Professor of Department of Clinical Epidemiology and Biosta- tistics, Javeriana University.
Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006 419
Syed R.
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○ cabo una investigación cualitativa, como that relatively few qualitative papers mala planeación, intentar de manera poco are accepted for publication in me- selectiva completar los criterios cualitativos dical journals (3) may reduce its pro- de las encuestas o subestimar los recursos necesarios para analizar los datos cualita- duction. Also, for health researchers, tivos de manera significativa. Conclusión: discouragement may result from the Con la planeación adecuada, entrenamiento assertion that they would benefit y supervisión, la metodología cualitativa es from training in or collaboration with una herramienta válida, útil y publicable appropriately trained social scien- para las investigaciones en el campo de la tists (4). Numerous social scientists psiquiatría. have produced various checklists, Palabras clave: investigación cualitativa, guidelines or manuals for qualita- depresión, adolescentes. tive research, which can be confu- sing and even contradictory (5). Tho- Introduction se working in psychiatry may the- refore come to the conclusion that Qualitative research is an um- qualitative research is not worth brella term describing several spe- even attempting, and should be left cific research methods that rely on to social scientists. the collection, analysis and interpre- tation of non-mathematical data. Aims Research in psychiatry overwhel- mingly relies on quantitative metho- This article attempts to provide dology (1). However, qualitative re- an outline of how those in the field search may be an invaluable tool in of psychiatry can go about planning psychiatry not only to assist in the high quality qualitative research resolution of current policy impera- whilst avoiding potential pitfalls, tives, such as calls for patient invol- using examples from the protocol “A vement and more thorough evalua- qualitative study investigating de- tion of services but also providing pressive prodrome in adolescents” insights into the phenomenology, which is published in this edition. diagnosis, treatment, management, prevention, and general understan- Methods ding of psychiatric disorders. Many factors may contribute to The main considerations when this potential under utilisation of planning qualitative research will be qualitative methods in psychiatry. outlined particularly with regard to It is still regarded by some as a “soft the question and background of the option”, lacking scientific rigour, research area, the underlying philo- open to possible bias or even fraud sophy, bias, ethical consideration (2) and a form of research unsuited followed by a nuts and bolts descrip- to evidence based practice. The fact tion of the methods.
420 Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006
A Guide to Planning Qualitative Research
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○
The study background and question qualitative study investigating de- Is the area being considered pressive prodrome in adolescents”. important and relevant? Qualitative methods may be more suited than quantitative me- Considerable time and expertise thods for describing many kinds of can be saved if these essential ques- complex settings and interactions; tions are thought about at the pla- including interactions amongst pa- nning stage. Research can be relevant tients, families, clinicians, within when it either adds to knowledge or and between professional groups increases the confidence with which and organizations, in communities existing knowledge is regarded. and within markets (7). This may be Another important dimension to rele- useful in psychiatry in exploring vance is the extent to which the areas such as attitudes, stigma (8), findings can be generalized beyond relationships and social interactions. the setting in which they were gene- rated, this also needs to be considered Underlying Philosophy in planning the sampling strategy. As stated earlier there are con- trasting ideas on quality assurance How will conducting a qualitative and assessment in qualitative re- study contribute to the field of study? search, as evidenced by the contras- ting and sometimes even contradic- Often qualitative research gene- tory guidelines, checklists and ma- rates hypotheses. If we were to con- nuals. These spring from different ceptualize research as reducing views on the philosophical founda- uncertainty, when a field is evolving, tion of qualitative methods held by not only is there uncertainty about different academics. Some believe answers but about what the right that qualitative research has entire- questions might be and how to fra- ly different theoretical under- me them in order to get answers and pinnings to quantitative research about where and to whom these and therefore cannot be assessed in questions should be addressed. the same way. Others believe that Therefore qualitative studies are although there are some differences invaluable in new fields of study and in the underlying theory, qualitative are often used when developing research can be assessed using si- theories or conceptual frameworks. milar ideas as quantitative research. For example, they have been useful On one end of the spectrum are in the developing field of early those that say that each research intervention research in psychotic perspective is unique and equally va- disorders (6) and have potential for lid in its own terms. This would mean similar use in depression, hence the that research could provide no development of the protocol for “A unequivocal insights relevant to
Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006 421
Syed R.
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○ action, and would therefore be of quently by Barbour (11), “a case of little use in applied health research the tail waging the dog”. (9). With a more realistic approach Instead, when planning qualita- in relation to health research, the an- tive research, it may be a more ti realists reject the belief that there useful strategy to consider the pur- is a single, unequivocal truth that is pose and area of research and how independent of the research process adopting a particular philosophical and as such cannot be judged by perspective may aid the research. conventional measures (such as We can then implement methodo- those used in quantitative research). logies, analytical strategies and They have an inductive approach to tools aimed at improving quality qualitative research; theory is made with sense and judgement from and grounded in the data. This For example, when planning “A is known as grounded theory. qualitative study investigating de- In subtle realism, a perspective pressive prodrome in adolescents” we advocated by Mays and Pope (10), it hypothesised that the period leading up to the first episode of depression is agreed that all research involves a could be retrospectively reconstruc- subjective position but unlike the ted. We were attempting to gain in- anti-realists, argue that there is so- sight into an underlying “reality”. We me underlying reality that can be felt that a qualitative study was nece- studied. They, therefore advocate the ssary; we could not assume that this broad concepts of relevance and vali- period is merely a milder form of de- dity to evaluate qualitative research. pression. We wanted to explore the The underlying philosophical evolution of these symptoms, possible perspective is pivotal in unders- patterns and associations, and see if tanding both how qualitative re- it was possible to “characterise” this search is conducted and evaluated. period. Our purpose was not to ex- The checklists and manuals produ- plain why it happens, but rather what ced by different social scientists re- happens and how it feels. In other flect their particular philosophical words we needed a qualitative study viewpoints. When conducting quali- to work out what questions need to tative research there is a danger in be asked in order to potentially iden- attempting to improve rigour by ti- tify individuals during this period. cking all the boxes of all of the diffe- rent checklists; in the attempt to Bias fulfil all suggested technical proce- dures, the overall, broad understan- Qualitative research, like all ding of the research may be lost. In research, is prone to bias. However other words the use of “quick fix” in qualitative research bias may be procedures and blind adherence to more difficult to identify, quantify check lists becomes, as stated elo- and reduce.
422 Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006
A Guide to Planning Qualitative Research
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○
Reducing potential sources of must undergo ethical review and bias is essential at the planning sta- approval. This is particularly impor- ge of a qualitative study. Consi- tant in when working with vulnera- deration of sources, strength and di- ble populations such as those with rection of potential biases when mental health problems. Also, spe- planning a study influence both its cial ethical dilemmas are encoun- the conduction and analysis. A sys- tered in qualitative research. Thus tematic approach to reducing bias when planning qualitative studies and adequate explanation of poten- involving mentally ill subjects this tial bias then relates to the quality must be given priority. The depth of the final piece of research. Refle- and extent of personal information xivity means the sensitivity to the collected is much greater than in ways in which the researcher and quantitative research, therefore in- the research process have shaped formed consent, handling and secu- the collected data. According to rity of research materials, de-iden- Mays and Pope (10) personal and tification of data and other issues of intellectual biases need to be made confidentiality need to be considered plain at the outset of any research and planned in detail. Special ethical report to enhance the credibility of dilemmas should be evaluated in the the findings. Also Hoddinott and Pill ethics and peer review of the study (12) emphasise the importance des- protocol before the commencement cribing explicitly who conducted the of the research. This should also research, who and how the subjects involve the development of a plain were recruited and how the research language statement, which standar- was explained to participants. dises the written explanation to par- Transparency is however diffe- ticipants of the research project. For rent to reproducibility. Hoddinott example in “A qualitative study in- and Pill suggest that adequate vestigating depressive prodrome in methodological data be included in adolescents”, ethical considerations every paper so that a reader would included the informed consent of then be able to “replicate the study adolescents, and consent for corro- and confirm the finding of required”. borative interviews. This important However others argue that due to and complex subject cannot be the nature of and subject of quali- covered in detail here, however these tative research this may be neither references may be of some use to the appropriate nor desirable (13). interested reader (14-15).
Ethical considerations Nuts and Bolts
As with any research, qualitative I set out to follow step by step research involving human subjects the development of a protocol and
Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006 423
Syed R.
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○ draw attention to some of the more Another consideration is that of important and controversial issues representativeness of cases (18-19). involved in producing high quality How widely and with what confiden- research. ce can these results be applied out- side this specific situation? At the Who or what will be studied? planning stage consider the potential usefulness of combining qualitative Sampling is an important consi- and quantitative methods to support deration when conducting any stu- generalisations (20). The external dy. It is necessary to both ensure validity of the research has many relevance of the sample to the re- implications for its usefulness. Any search question (16) and document assertions made in the final do- your reasoning comprehensively. It cument should describe this process. is important to remember that the It is important to consider the “units” of research are often not as context of the research as well as easily conceptualised as in quanti- practical aspects. For example as tative research. It may be useful to explained in “A qualitative study think about a choosing a sample investigating depressive prodrome that is theoretically likely to produce in adolescents”, a more convenience of the type of knowledge necessary method of sampling was more to understand the structures and appropriate to gain a relevant sam- processes within which individuals ple without loss of qualitative repre- or situations are located (17). sentability. Sampling in a qualitative study It may be useful to access our can involve different sampling me- subjects again after data collection for thods to that used in a quantitative a process known as member study. Quantitative studies often checking. This includes techniques in use random or consecutive sam- which the investigators account is pling in order to gain a representa- compared with those of the research tive sample. Sampling in qualitative subjects to establish a level of studies aims to achieve a qualitati- correspondence between the two sets. vely representative sample by co- These reactions to the analysis are vering a range of potentially relevant then incorporated into the study social phenomena and perspectives findings. Some researchers consider from an appropriate array of data this the strongest available check on sources. Purposive sampling is often credibility of the research project (21). used in qualitative research. This effectively means that researchers How will data be collected? deliberately seek to include outliers. These would conventionally be dis- There are several ways of co- counted in quantitative approaches. llecting data in qualitative research.
424 Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006
A Guide to Planning Qualitative Research
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○
The most commonly used methods data sources are compared. The re- are field observations, interviews or searcher looks for patterns of con- document analysis, separately or in vergence to develop or corroborate an combination. These are described in overall interpretation. In “A qualitative more detail in other texts (22). study investigating depressive prodrome in adolescents” corrobora- What is the data collection process? tive interviews were planned.
The important issue in planning How will the data be recorded? data collection is not only the type of data collection but also the data When choosing the method of collection process. This needs to allow recording data, aspects such as de- rich and robust descriptions of the tail, accuracy and practicality must observed event and be comprehensive be considered. For example in the ca- enough in both breadth and depth to se of interviews, audio-taping may be generate and support interpretations. the first stage of recording data, which For example in “A qualitative study can then be transcribed as planned investigating depressive prodrome in in “A qualitative study investigating adolescents” in depth interviews were depressive prodrome in adolescents”. 1.5-2 hours long. In qualitative stu- However, the process of transcription dies the collection of data and sam- for qualitative research purposes is pling can take place at the same time- different from that used for medical an iterative approach, and rely on dictation. Details such as pauses, each other. Especially in an inductive background noise etc may provide study, with little theorised previously, invaluable data for the elaboration of sampling and analysis may occur in the meaning of the spoken word. a constant feed back loop in which the qualitative researcher at first How will the data be analysed? makes theories and then sets out to find cases that disprove the theory. When planning a qualitative Other data collection may also be study the volume of data generated needed in order to corroborate the fin- and therefore the analysis of this dings. This needs to be planned in data is labour intensive. Far more terms of what information is going to time must be set aside for this time be useful and practically feasible? consuming part of the study than This is known as triangulation in qua- in quantitative research. litative research and addresses the The analysis itself will very internal validity of the study. In trian- much depend on the specifics of the gulation the results from either two study and the extent to which it sets or more different methods of data co- out to either generate or test theory. llection or more simply, two or more Analytical categories are used to
Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006 425
Syed R.
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○ describe and explain social pheno- Planning the write up mena. These categories can be de- rived inductively, that is obtained It can be helpful to plan how gradually from the data or deducti- this data will be explained and re- vely either at the beginning or part presented when the research is way through the analysis as a way completed. The study needs to be of approaching the data. recorded in sufficient detail so that Software packages can be used at its conclusion a clear account can to aid analysis programmes such as be provided of how early simpler NUDIST (22) or N-VIVO. These are systems of classification evolved into effectively data handling kits; use of more sophisticated coding structu- them does not ensure that the ana- res and thence into clearly defined lysis is adequate, neither is the use concepts and explanations. This is of them necessary (13) although they known as the audit trail. may save considerable time. If they are intended for use in a study, prior Discussion experience or training in their use at the planning stage is invaluable. There are many potential pitfalls Some have found that the con- in conducting qualitative research. sistency and reliability of the analysis These may involve poor planning, can be improved by the involvement unselectively attempting to fulfill all of more than one analyst (23). Ho- of the quality criterion suggested in wever this is a contentious issue as all the various checklists, or under others argue that the person who estimating the resources necessary collected interview data is more likely to meaningfully analyze qualitative to have insight and understand the data. context in which the interview took place. Morse (24) stated “the quan- Conclusion titative view of ensuring reliability and validity by using external raters is not With adequate planning, training recommended for qualitative re- and supervision, qualitative metho- search”. The value of involvement of dology is a valid, useful and publisha- external analysts, team members and ble tool for those researching within the extent of supervision need to be the field of psychiatry. considered at the onset and throughout the analysis of a study. References During the analysis stage steps can be taken to not only to search for 1. Crawford MJ, Ghosh P, Keen R. Use of alternative explanations but also to qualitative research methods in general medicine and psychiatry: publication find and discuss elements that seem trends in medical journals 1990-2000. to contradict the emerging theory. Int J Soc Psychitary. 2003;49:308-11.
426 Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006
A Guide to Planning Qualitative Research
○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○
2. Mant D. R and D in primary care-an NHS 13. Rogers A, Popay J, Williams G, Latham priority. Br J Gen Pract. 1998;48:873. M. Inequalities in health and health 3. Boulton M, Fitzpatrick R, Swinburn C. promotion: insights from the literature. Qualitative research in healthcare 2. A London: Health Education Authority; structured review and evaluation of stu- 1997. p. 384. dies. J Eval Clin Pract. 1996;2:171-9. 14. Usher K, Holmes C. Ethical aspects of 4. Harding G, Gantley M. Qualitative me- phenomenological research with menta- thods: beyond the cookbook. Fam lly ill people. Nurs Ethics. 1997;4:49-56. Pract. 1998;15:76-9. 15. Holloway I, Wheeler S. Ethical issues in 5. Chapple A, Rogers A. Explicit guidelines qualitative nursing research. Nurs for qualitative research: a step in the right direction, a defence of the “soft Ethics. 1995;2:223-32. option”, or a form of sociological impe- 16. Giacomini MK, Cook D. Users guide to rialism? Fam Pract. 1998;15:556-61. the medical literature: XXIII. Qualitative 6. Yung AR, McGorry PD. The initial pro- research in health care A. Are the results drome in psychosis: descriptive and of the study valid? Evidence-Based qualitative aspects. Aust N Z J Psy- Medicine Working Group. JAMA. chiatry. 1996;30:587-99. 2000;284(3):357-62. 7. Sofaer S. Qualitative methods: what are 17. Popay J, Rogers A, Williams G, they and why use them? Health Serv Rationale and standards for the syste- Res. 1999;34(5Pt 2):1101-18. Review. matic review of qualitative literature in 8. Schulze B, Angermeyer MC. Subjective health services research. Qual Health experience of stigma: a focus group Res. 1998;8:341-51. study of schizophrenic patients, their 18. Seale C, Silverman D. Ensuring rigour relatives and mental health professio- in qualitative research. Eur J Pub Health. nals. Soc Sci Med. 2003;56:299-312. 1997;7:379. 9. Murphy E, Dingwall R, Greatbatch D, 19. Silverman D. Interpreting qualitative Parker S, Watson P. Qualitative research data. London: Sage; 1994. methods in health technology assess- 20. Lincoln YS, Guba EG. Naturalistic ment: a review of the literature [Mono- enquiry. Newbury Park, CA: Sage; 1985. graph]. Health Technol Assess. 1998;2. 10. Mays N, Pope C. Qualitative research 21. Green, JM, Thorogood N. Qualitative in health care: assessing quality in quali- methods for health research. London: tative research. BMJ. 2000;320:50-2. Sage; 2004. 11. Barbour RS. Checklists for improving ri- 22. Richards T, Richards L, McGalliard J, gour in qualitative research: a case of Sharrock B. NUDIST 2.3 Reference ma- the tail wagging the dog? BMJ. 2001; nual. Bundoora: Replee; 1992. 322:1115-7. 23. Mays N, Pope C. Rigour in qualitative 12. Hoddinott P, Pill R. A review of recently research. Br Med J. 1995;311:109-12. published qualitative research in gene- 24. Morse J. Designing funded qualitative ral practice. More methodological ques- research. In: Denzin N, Lincoln Y, editors. tions than answers? Fam Pract. 1997: Handbook of qualitative research. 14:313-20. London: Sage; 1994.
Recibido para evaluación: 15 de junio de 2006 Aceptado para publicación: 26 de agosto de 2006
Correspondencia Rebecca J. Syed Departamento de Psiquiatría y Salud Mental Pontificia Universidad Javeriana Carrera 7.ª # 40-62, piso 8 Bogotá, Colombia [email protected].
Revista Colombiana de Psiquiatría, vol. XXXV / No. 3 / 2006 427