Online Journal of Health and Allied Sciences

Peer Reviewed, Open Access, Free Online Journal

Published Quarterly : Mangalore, South India : ISSN 0972-5997 This work is licensed under a Creative
Volume 8, Issue 4; Oct - Dec 2009 Commons Attribution-No Derivative Works
2.5 India License

Review:
Application of Qualitative Methods in Health Research: An Overview
Amol R Dongre,
Department of Community Medicine, Sri Manakula Vinayagar Medical College, Pondicherry, India
Pradeep R Deshmukh,
Dr. Sushila Nayar School of Public Health, Mahatma Gandhi Institute of Medical Sciences, Sewagram, India
Ganapathy Kalaiselvan,
Department of Community Medicine, Sri Manakula Vinayagar Medical College, Pondicherry, India
Sanjeev Upadhyaya,
Office of Population, Health & Nutrition, USAID–India, US Embassy, Chanakyapuri, New Delhi - 110029

Address For Correspondence:

Dr. Deshmukh PR,
Professor, Dr Sushila Nayar School of Public Health,
Mahatma Gandhi Institute of Medical Sciences, Sewagram – 442102, India
E-mail: [email protected]

Citation: Dongre AR, Deshmukh PR, Kalaiselvan G, Upadhyaya S. Application of Qualitative Methods in Health Research: An
Overview. Online J Health Allied Scs. 2009;8(4):3
URL: http://www.ojhas.org/issue32/2009-4-3.htm
Open Access Archives: http://cogprints.org/view/subjects/OJHAS.html and http://openmed.nic.in/view/subjects/ojhas.html

Submitted: Oct 5, 2009; Accepted: Mar 31, 2010; Published: Apr 30, 2010

Abstract:
field of public health. Over the last decade there has been an
Qualitative research is type of formative research that includes
increase in the use of qualitative research methods in health
specialized techniques for obtaining in-depth responses about
research.2 Some of the important reasons for this revival of in-
what people think and how they feel. It is seen as the research
terest are: 1) growing realization of unsuitability of survey re-
that seeks answer to the questions in the real world. Qualitat-
search methods in the context of developing countries where
ive researchers gather what they see, hear, read from people
population is predominantly illiterate and where magnitude of
and places, from events and activities, with the purpose to
non-sampling errors is high in surveys, 2) increased interdis-
learn about the community and to generate new understanding
ciplinary team work and 3) demand of quick results from the
that can be used by the social world. Qualitative research have
ethnographic work.3
often been conducted to answer the question “why” rather than
“what”. A purpose of qualitative research is the construction The latest trend in the field of research is the combined use of
of new understanding. Here, we present an overview of applic- quantitative and qualitative research methods i.e. mixed-meth-
ation of qualitative methods in health research. We have dis- od design within a single data set. According to Morse (2005),
cussed here the different types of qualitative methods and how it is in this area that the largest abuses of qualitative data are
we and others have used them in different settings/scenarios; occurring, largely because methodological principles have not
sample size and sampling techniques; analysis of qualitative been followed. Hence, it is necessary to understand the meth-
data; validity in qualitative research; and ethical issues. ods of qualitative data collection and its analysis. 4 The present
Key Words: Participatory research techniques, PRA, In-depth article provides the brief descriptive overview on application
techniques, systematic techniques for qualitative research methods in health research.
Types of qualitative methods:
Introduction: The qualitative data collection techniques range from the
Qualitative research is type of formative research that includes highly structured systematic techniques to the highly flexible
specialized techniques for obtaining in-depth responses about people-centered participatory techniques. These can be
what people think and how they feel. It is seen as the research broadly classified under 3 major headings:3
that seeks answer to the questions in the real world. Qualitat- Participatory research (PR) techniques: In conventional re-
ive researchers gather what they see, hear, read from people search, knowledge is generated by the researchers for their
and places, from events and activities, with the purpose to academic pursuit and the study subjects have no control over
learn about the community and to generate new understanding it. Participatory methods offer collective educational process
that can be used by the social world. 1 Qualitative research which involves people as stakeholders for their empowerment
have often been conducted to answer the question “why” and assumes that the ordinary people already possess know-
rather than “what”. A purpose of qualitative research is the ledge and have an understanding of their reality which is gen-
construction (not the discovery) of new understanding. A skill- erated during their on-going struggle for survival. PR process
ful use of logically sequenced different techniques of data col- intends to change existing local problems and synthesize local
lection can maximize the validity of the data and provides in- people’s knowledge with existing scientific knowledge.
depth understanding of emotional and contextual aspect of hu-
man responses rather than objective and measurable aspect. Apart from addressing the pressing social and economic issues
like poverty, poor land irrigation, forestry, housing and water
Historically, qualitative research methods have their roots in supply etc, PR is increasingly being used for community based
several disciplines; literary criticisms, social sciences and psy- health and development project. Here, the local community
choanalytic theory. However, in recent years, there has been a members are involved in needs assessment and develop a pri-
revival of interest about qualitative research methods in the ority based health plan and act upon it. Most frequently used
participatory methods are Social Mapping, Pair Wise Ranking, Seasonal Calendar, Cobweb Diagram, Trend Analysis, Venn Diagram
1
and Transect Walk.5 Anganwadi workers and FGDs were facilitated with the group
A non-government organization, Mother and Infant Research of mothers to their opinion about supplementary food.17 A
Activities (MIRA) in rural Nepal, involved women self help combination free list and pile sort exercise was used for know-
groups through participatory research for planning perinatal ing the rural mothers’ perceptions for newborn danger signs
care services.6 In rural India, Community led Initiatives for and their household level care practices. 17 In another com-
Child Survival (CLICS) project involved community members munity based participatory research and action similar com-
through participatory research and action on key maternal and bination was used for understanding the rationale for changed
child health issues.7 In a study on health care seeking for new- health care seeking for newborn danger signs in rural Wardha.7
born danger signs in periurban Wardha, mapping exercises Sample size and sampling techniques:18
were undertaken to identify the respondents and pair wise
ranking exercise prioritized mothers’ preferences for the avail- Sample size: It is difficult to determine sample size and there
able health care services for sick newborn. 8 In rural Wardha, is no mathematical formula to calculate sample size in qualit-
an exploratory Venn diagram exercise (chapatti diagram) with ative research. It depends on the purpose of the study and
Anganwadi workers brought out their perceived job responsib- available resources. The validity, meaningfulness and insights
ilities and the seasonal calendar exercise with the mothers of generated from the qualitative data have more to do with the
severely malnourished children explored the seasonality of richness of the data obtained. The process of data collection is
childhood illness, availability of money, free time, food and continued till the saturation point i.e. where no new informa-
access to loan across twelve months in a year. 9 A participatory tion is added after the additional interviews or focus group dis-
research and action for rural adolescent girls could involve ad- cussions. Since there is no sample size estimation and there is
olescent girls to improve their practices related to menstrual use of non-probability sampling in qualitative research, the
hygiene and explored the trend of change over the period by findings are rarely used to test the hypothesis and generally it
using trend analysis techniques.10 A transect walk with directs the future course of research on relatively new or rare
primary school teachers in village Dhotra (Kasar) could ex- topic by generating research hypothesis.
plore the villagers personal hygiene practices which was used Sampling techniques: Sampling is a scientific way of select-
to devise a need based health education intervention for school ing study subjects. Since the purpose of qualitative research is
children with active involvement of school teachers and chil- to find out answer to the question why, and explore different
dren.11Noteworthy, PR ensures local people’s participation in perspectives on the research topic by generating rich textual
research and action process and strengthens their action-exper- data, Non-probability sampling is recommended. The ex-
ience-learning cycle. Hence, participatory research techniques amples of non-probability sampling are as follows. 1) Purpos-
offer wider scope and should not be seen as mere data collec- ive sampling, where sample units are selected with definite
tion tools. purpose in view, e.g. women who adopted different methods
In-depth techniques: These are qualitative in-depth flexible of contraception, victims of some events. 2) Convenient
discussions or interviews with the group or person who knows sampling, where the conveniently available respondents are
what is going in community about the topic on which we want selected, e.g. women in field, temple or common meeting
to get information. These methods are widely used for explor- place. 3) Quota sampling is a restricted type of convenient or
ing sensitive topics in medical education, demography and purposive sampling defining the quota of sample to be drawn
public health etc. Some commonly used methods are Focus from different strata and then drawing the required
Group Discussion (FGD), Key Informant Interviews (KII) and sample. 4) In Snow-ball sampling, the sample is driven by the
In-depth Interview (IDI).12 respondents. It involves asking your respondents to identify
other potential participant with specific set of characteristics
A study on newborn care practices in slums of Delhi under- and then asking the next respondent. It is used when the target
took KII and IDI of caregivers for exploring their health care population is unknown or difficult to approach, e.g. such as
seeking behavior for newborn danger signs. 13 In rural Viet- Male having Sex with Male (MSM) population and Sex work-
nam, Huy TQ et al undertook FGDs to explore socio-cultural ers etc.
and health systems factors that may impact on death reporting
by lay people to registry systems at the commune level. 14 Pro- Sequencing of the methods 3,5: The qualitative data collection
gram for Appropriate Technology in Health (PATH) have giv- should be ‘on-going’ or ‘iterative’ process. Hence, methods
en guidelines for development of health education material should be logically sequenced where one method directs the
where FGDs are advised for needs assessment for health mes- other, for example, you identify potential respondents for fo-
sage development.15 Health education materials based on such cus group discussion during social mapping exercise. This tri-
guidelines have been found effective in the field settings.16 angulation of qualitative methods ensures better validity of the
results and adds to the richness of qualitative data. In mixed
Systematic techniques: These techniques can be used with al- methods design, triangulation quantitative (survey) and qualit-
most any qualitative research methods such as focus group or ative research method is undertaken into the same research
participatory research to collect systematic and structured data design. Pre-survey qualitative research is undertaken for better
on a specific research issue. This approach is based on the pre understanding of the underlying dynamic on given re-
principle that people make sense of their words by grouping search topic in study area, for exploring local terms on re-
their observation or experiences in class known as “domain”. search topic and developing locally relevant questionnaire.
Examples are Free listing combined with Pile sorting, Delphi FGDs are undertaken as needs assessment for designing loc-
panel.12 Free list combined with pile sort can be used for ex- ally relevant behavior change strategy. Post- survey qualitative
ploring the perceptions of local people on a given research research is undertaken to bridge the gaps of information in
topic in a systematic manner. Free list exercise can be under- survey.
taken as a pre-survey qualitative research for exploring local
terms for locally relevant questionnaire on a relatively new Analysis of qualitative data:19
topic or if researchers are unknown about its underlying dy- Data analysis in qualitative research is a multi-faceted en-
namics in the study area. deavor. It requires planning, capacity for being open to views
that are different from your very own, an appreciation of pro-
In qualitative research for exploration for various reasons for visional nature of human knowledge, strong conceptual skills
malnutrition, free list and pile sort exercise were used with the and excellent scholarship. Let us understand the language and
terminology of qualitative data analysis.
Interim analysis: The qualitative data analysis is an on-going is interested in is understood (or the investigator runs out of time and
and iterative (non-linear) process in qualitative research. This resources).
is known an interim analysis. It is cyclical process of collect-
Memo: Throughout the entire process of qualitative data ana- lysis it is
ing data and analyzing it during a single research study. Inter-
good idea to engage in writing the memos. It is re- cording of the
im analysis continues until the process or topic the researcher
“reflective notes” about what you are learning from your data. The idea

2
is to write memos to you when you have idea or thought and ferent judges tend to assign exactly the same rating to each ob-
include those memos as ‘additional data’ to be analyzed. ject. It is assessed by calculating statistics such as Cohen's
Coding: It is defined as making the segments of data with kappa, Scott's pi, Holsti’s coefficient of reliability and Krip-
symbols, descriptive words or category name. A master list of pendorff's alpha. Coefficients of .90 or greater are nearly al-
codes is developed and applied to new segments of data each ways acceptable, .80 or greater is acceptable in most situ-
time an appropriate segment is encountered. A ‘priori’ codes ations, and .70 may be appropriate in some exploratory studies
are developed before examining the current data and ‘induct- for some indices. Higher criteria should be used for indices
ive’ codes are developed by the researcher by direct examin- known to be liberal (i.e., percent agreement) and lower criteria
ing the data. can be used for indices known to be more conservative.20
Content analysis: It is a widely used qualitative research tech- Step 7: Drawing conclusions from the coded data: This is a
nique for subjective interpretation of content of text data crux of qualitative data analysis. It involves reading and re-
through the systematic classification process of coding and reading of text data. The activities involve exploring proper-
identifying themes or patterns. Content analysis has three dis- ties and dimensions of categories and identify relationships
tinct approaches: conventional, directed, or summative. All between categories.
three approaches are used to interpret meaning from the con- Step 8: Reporting: While writing report it is important to
tent of text data and, hence, adhere to the naturalistic maintain the balance between description and interpretation.
paradigm. The major differences among the approaches are Here, one can use conceptual frameworks derived from the
coding schemes, origin of codes and threats to trustworthiness. data set. An interesting and readable report provides sufficient
In conventional content analysis, coding categories are derived description to allow the reader to understand the basis for an
directly from the text data. With directed approach, analysis interpretation and sufficient interpretation to allow the readers
starts with a theory or relevant counting and comparisons, usu- to understand the description. Tong et al noted that there is no
ally key words or content, followed by the interpretation of the criteria for reporting qualitative research findings and sugges-
underlying context. ted 32-item check list as Consolidated criteria for Reporting
Steps in the process of content analysis: Qualitative research (COREQ) under three main domains 1)
For qualitative research, triangulation of multiple methods and Research team & reflexivity 2) Study design 3) Analysis and
investigators is recommended for better interpretation and findings.21
validity of the findings. Use of software in analysis of qualitative data: Traditionally,
Step1: Transcription: The raw data is collected as notes or au- qualitative data were analyzed "by hand" using some form of
dio or video recordings. This needs to be transformed into filing system. For smaller data set manual content analysis is
written text format for the purpose of analysis. undertaken. Here, coding is done manually along a narrow
blank column of the text document. A computer assisted cod-
Step 2: Deciding the unit of analysis: Defining the coding
ing using software packages (that are specifically designed for
unit is one of the most fundamental and important step. The
qualitative data and analysis) has significantly reduced the
commonly used coding units are word, concept, sentence,
need for the traditional filing technique. The most popular
paragraph and theme. Changes in coding unit will affect cod-
qualitative data analysis packages are NUDIST, ATLAS-ti,
ing decisions as well as comparability of outcomes with other
and Ethnograph. ATLAS-ti 5.0 and NVivo 2.0 are among the
similar studies.
best available and potentially most useful qualitative data ana-
Step 3: From units to categories: Categories and code lysis (QDA) tools. Both are tremendously flexible programs
schemes can be derived from three sources such as 1) data it- that can be readily applied in wide range of applications. Note-
self, 2) previous related studies, 3) theories. Inductive coding worthy, computer aided content analysis should not be seen as
is done when researcher intend to develop theory rather than a shortcut to various steps of qualitative data analysis.
describing phenomenon or verifying existing theories.
Methods to ensure validity in qualitative research:
Step 4: Test coding on sample test: To ensure coding consist- Some commonly used strategies to promote trustworthiness
ency especially when multiple coders are involved, develop- and validity in the findings of qualitative research are given
ment of instruction guidelines defining rules of coding is es- below.22
sential. If there is low inter-coder agreement then revise the
Researcher as detective: The researcher has to develop the un-
rules of coding sample text and checking coding consistency.
derstanding of the data through careful consideration of poten-
Step 5: Code all text data: When sufficient consistency is tial causes and effects by systematically eliminating the rival
achieved then coding rules can be applied to code all the text explanations and hypothesis until the final cause is made bey-
data. ond a reasonable doubt.
Step 6: Assess the coding consistency: After coding all text Extended field work: For both discovery and validation, the
data, coding consistency needs to be re-checked. Human researcher should collect data in the field over the extended
coders are subject to fatigue and are likely to make mistakes as period of time.
coding proceeds. Inter-coder agreement is needed in content
Low-inference descriptors: The use of descriptions phrased
analysis because it measures only the extent to which the dif-
very close to the participant’s account or researcher’s field
notes. Verbatim i.e. direct quotations are used as low-infer-
ence descriptors.
Triangulation: Cross-checking of information and conclu-
sions through the use of multiple procedures and sources may
be undertaken for valid results. A combination of multiple
methods to study a phenomenon gives a better and in-depth
understanding of the research question. A use of multiple in-
vestigators to collect and interpret data adds to the validity of
the results.
Participant feedback: The feedback and discussion on the re-
searcher’s interpretation and conclusions with actual parti-
cipants and other members of the community helps in verifica- involved. Peer should be skeptical and play the devil’s advoc- ate,
tion and better insight into the research problem. challenging the researcher to provide solid evidence for any
interpretation or conclusion.
Peer-review: It is recommended to discuss the findings with
the disinterested peer e.g. other researcher who is not directly Ethical issues in qualitative research:

3
There is increase in publications on qualitative research meth- Childbirth 2005;5:6 [Online] [cited on May 20,
ods. Qualitative research is vulnerable to bias through the atti- 2009]; Available from URL: http://www.pubmed-
tude and qualities of the researcher, social desirability factor, central.nih.gov/articlerender.fcgi?artid=1079874
and conditions of worth.23 Except with few exceptions of nurs-
ing, sociological and anthropological literature,24 no published 7. Dongre AR, Deshmukh PR, Garg BS. A com-
articles on qualitative methods in medical research addressed munity based approach to improve health care
ethical issues. In India, Indian Council of Medical Research seeking for newborn danger signs in rural Wardha,
(ICMR) has provided ethical guidelines for biomedical re- India. Indian Journal of Pediatrics January
search on human subjects without specific reference to qualit- 2009;76:45-50.
ative research. Richards et al24 have argued need for ethical 8. Dongre AR, Deshmukh PR, Garg BS. Awareness
guidelines for qualitative research in health sciences. Unlike and health care seeking for newborn danger signs
social scientists, health professionals have poor orientation to among mothers in peri urban Wardha. Indian
philosophical aspect of research. Also, ethical committees Journal of Pediatrics. April 2009;76:691-693
have difficulty assessing ethical issues arising in relation to
qualitative studies.25 Time for research exercise should be de- 9. Dongre AR, Deshmukh PR, Garg BS. Perceived
cided in consultation with the participant. Consent should be responsibilities of Anganwadi workers and malnu-
obtained and refreshment should be arranged at the end of fo- trition in rural Wardha. Online Journal of Health
cus group discussion. Examination of self through critical re- and Allied Sciences.2008;7(1):3. Available
flection and supervision are necessary component of ethical at http://ojhas.org/issue25/2008-1-3.htm
research.23 10. Dongre AR, Deshmukh PR, Garg BS. The effect of
To summarize, the application of qualitative research methods community based health education intervention on
in health research is increasingly been seen. There are now cri- management of menstrual among rural Indian ad-
teria for reporting qualitative research findings (COREQ), a olescent girls. World Health and Population. [On-
32-item checklist, which can help researchers to report import- line] 2007[cited on May 20, 2009]; Available
ant aspects of the research team, study methods, context of the from: http://www.longwoods.com/home.php?
study, findings, analysis and interpretations. Subsequent re- cat=381
search on this check-list is required to improve the quality of 11. Dongre AR, Deshmukh PR, Garg BS. An approach
reporting. The training of the health care providers and re- to hygiene education among rural Indian school go-
searchers should be done on application of these methods in ing children. Online Journal of Health and Allied
their work settings. Also ethical guidelines should be de- Sciences.2007;6(4). Available at http://ojhas.org/is-
veloped for use of qualitative research methods in research. sue24/2007-4-2.htm
Disclaimer: 12. Dawson S, Manderson L, Tallo VL. The focus
The views expressed in this article are those of the author in group manual: Methods for social research in dis-
his professional capacity, and do not necessarily reflect those ease. Boston: International Nutrition Foundation
of the U.S. Agency for International Development or of the for Developing Countries (INFDC); 1993.
U.S. Government.
13. Awasthi S, Verma T, Agarwal M. Danger signs of
References: neonatal illnesses: perceptions of caregivers and
1. Rajesh Tandon (Ed). Participatory Research: Re- health workers in northern India. Bull World
visiting the Roots. New Delhi (India), Mosaic Health Organ2006;84:819-826.
books; 2005. 14. Huy TQ, Johansson A, Long NH. Reasons for not
2. Hoddinott P, Pill RM, A review of recently pub- reporting deaths: A qualitative study in rural Viet-
lished qualitative research in general practice. nam. World Health and Population. 2007 Feb;1-
More methodological questions than 10.
answers? Fam Pract 1997;14:313-319. 15. Zimmerman M, Newton N, Frumin L, Wittett S.
3. Sinha RK, Chattopadhyay A (ed.). Research Meth- Developing Health and Family Planning Materials
odology (Module No 10). Mumbai (India), Interna- For Low-Literate Audiences: A Guide. Washington
tional Institute for Population Sciences; 2003. DC: Program for Appropriate Technology in
Health; 1996.
4. Morse JM. Evolving trends in qualitative research:
Advances in Mixed-Method design. Qualitative 16. Dongre AR, Deshmukh PR, Garg BS. Process doc-
Health Research 2005;15(5):583-585. umentation of ‘Health Education Interventions’ for
school children and adolescent girls in rural India.
5. Training in Participation Series [PRA tips on CD- [Online] [cited on September 28, 2009]; Education
ROM]. Patna (India): Institute for Participatory for Heath. 2009;22(1). Available from
Practices; 2004. URL: http://www.educationforhealth.net/
6. Morrison J, Tamang S, Mesko N, Osrin D, 17. Dongre AR, Deshmukh PR, Garg BS. Eliminating
Shrestha B, Manandhar M, Manandar D, Standing childhood malnutrition: Discussions with mothers
H, Coatello A. Women’s health groups to improve and Anganwadi workers. Journal of Health Stud-
perinatal care in rural Nepal. BMC Pregnancy ies. May-December 2008. [online] [cited May 7,
2010]; Available from URL: http://www.esocials-
ciences.com/data/essJournalArticles/Art-
icle141120086.pdf
18. Wilmot A. Designing sampling strategies for qual-
itative social research: with particular reference to
the Office for National Statistics’ Qualitative Re-
spondent Register. [Online] [cited on May 20,
2009]; Available from URL: http://www.ons.gov-

4
 .uk/about/who-we-are/our-services/data-collection-
methodology/reports-and-publications/designing-
sampling-strategies-.pdf .
19. Zhang Y, Wildemuth BM. Qualitative analysis of
content. [online] [cited May 7, 2010]; Available
from
URL: http://www.ils.unc.edu/~yanz/Content_ana-
lysis.pdf
20. Practical resources for assessing and reporting
inter-coder reliability in content analysis research
projects. [Online] [cited on September 24, 2009];
Available from URL: http://astro.temple.edu/~lom-
bard/reliability/
21. Tong A, Sainsbury P, Craig J. Consolidated criteria
for reporting qualitative research (COREQ): a 32-
item checklist for interviews and focus groups. In-
ternational J for Quality in Health
Care 2007;19:349-357.
22. Qualitative content analysis. [Online]. [cited on
May 20, 2009]; Available from URL: www.-
southalabama.edu/coe/bset/johnson/lectures/lec17.
pdf
23. Hewitt J. Ethical components of researcher-re-
searched relationship in qualitative interview-
ing. Qualitative health research 2007;17(8):1149-
1157.
24. Borland K. “That’s not what I said”: interpretative
conflict in oral narrative research. In Perks R,
Thomson A (Ed.). The Oral History Reader. New
York: Routledge, 1998:320-332.
25. Richards HM, Schwartz LJ. Ethics of qualitative
research: are there special issues for health services
research? Fam Pract 2002;19:135-139.