A standardized approach to qualitative content analysis of focus group discussions from different countries
Introduction
The advantage of qualitative research is the richness of the collected data. However, for research purposes, these data need to be interpreted and coded in a valid and reliable way, for instance by qualitative content analysis. Qualitative content analysis techniques seek to classify the discussion material into an effective number of categories that represent similar meanings. According to Hsieh and Shannon [1], qualitative content analysis can be defined as a “research method for the subjective interpretation of the content of text data through the systematic classification process of coding and identifying themes or patterns”. The validity of the inference is ensured by complying with a systematic coding process. In other words, content analysis allows researchers to interpret subjective data in a scientific manner.
When analysing focus group discussions, researchers have the choice between two main options, the inductive content analysis (or conventional content analysis) or the deductive content analysis (or directed content analysis) [1], [2], [3]. The specific type of content analysis approach chosen varies according to the purpose of the research and the problem being studied. This distinction is primarily based on the different way (inductive or deductive) the categories are derived from the text.
In inductive content analysis coding categories are derived directly and inductively from the raw data. Researchers avoid using preconceived categories, allowing the categories and names for categories to ‘flow from the data’ instead. They immerse themselves in the data to allow new insights to emerge. Similar to a grounded theory approach, the main purpose is to develop theories. The advantage of the conventional approach of content analysis is that direct information is gained from the study participants without preconceived theoretical perspectives having been imposed [1], [2], [3].
Deductive content analysis is guided by a more structured process than in an inductive approach. The deductive approach is based on previously formulated, theoretically derived categories and the initial coding starts with a theory or relevant research findings. Using existing theory or prior research, researchers begin by identifying key concepts or variables as initial coding categories of analysis, bringing them in connection with the text [1], [2], [3].
Hsieh and Shannon [1] indicate as appropriate the use of a conventional content analysis when “existing theory or research literature on a phenomenon is limited”. On the other hand a deductive approach should be preferred when the purpose is “to validate or extend conceptually a theoretical framework or theory”.
We decided to use the inductive approach for analyzing focus group discussions on the quality of physician communication from a patient perspective which were held in four western European countries. The analysis of these focus group discussions should offer new insights into lay people's perspective on doctor–patient communication in different countries and contribute to the development of patients’ theory on physicians’ communicative performance. Citizens of four different countries were involved in this large international multi-centre focus group study (GULiVer). The study draws its name GULiVER from the four participating centres: Ghent University (Belgium), Utrecht University/NIVEL (the Netherlands), Liverpool University (United Kingdom) and the University of Verona (Italy). A detailed description of the research protocol is provided elsewhere [4].
Cross country comparisons of focus group discussions as in GULiVer present linguistic, procedural, methodological challenges which have to be resolved in order to obtain a reliable, consensus based coding system with which to analyze the huge amount of qualitative data.
The present paper, which is the second of a series, aims to give an example of a standardized approach to analyze qualitative data from a multi-centre study and describes the step-by-step procedures leading to the final coding system, to be applied, eventually, to the entire set of focus groups. The procedures described are based on the data collected in Utrecht, Liverpool and Verona. Ghent joined the project later and did not contribute to the development of the coding system.
Section snippets
Study sample and focus group task
Recruitment of the lay people took place in and around Liverpool (UK), Verona (Italy) and Utrecht (Netherlands) in the early summer of 2008; in 2009/2010, Ghent (Belgium) joined the project. People were approached in public areas, such as shopping centres, and via calls in free local papers. The same recruitment procedure was used in all countries. Inclusion criteria were age over 18 years; at least one GP-visit over the last 12 months; speaking the country's language; not being involved in a
Inter rater reliability phase 2
As shown in Table 5, the revised rules of turn segmentations and definitions of critical categories improved inter-rater reliability (alpha = 0.48). This value corresponds to a moderate agreement, but given the great number of categories, the use of more than two raters, and the more conservative index of Krippendorff's alpha, such a finding may be considered acceptable [18]. However, in order to further improve the reliability of the subsequent coding process of all focus groups it was decided
Discussion
Qualitative research is particularly appropriate in relatively new areas when the main goal of the study is to learn more about people's perspectives and experiences. Accordingly, qualitative approaches to doctor–patient communication can provide precious information about the quality of communication from a patient perspective and we can observe a growing use of focus groups for such a purpose [20], [21], [22], [23]. This methodology “empowers” participants to become active partners in the
Funding
This study was made possible through a grant of The Dutch Ministry of Health, Welfare and Sports (National Fund for Patient-Oriented Research).
Disclosure
We confirm all patient/personal identifiers have been removed or disguised so the patient/person(s) described are not identifiable and cannot be identified through the details of the story.
Conflict of interest
The authors have no conflict of interest that could have influenced the paper.
Acknowledgements
The Clinical Skills Team at The Medical School in University of Liverpool for supporting the study and assisting the recruitment and videoing of the summative examinations.
The lay panels in Utrecht, Liverpool and Verona for their committed participation in the study The Dutch Ministry of Health, Welfare and Sports (National Fund for Patient-Oriented Research) for their financial support of the study.
References (30)
- et al.
Qualitative content analysis in nursing research: concepts, procedures and measures to achieve trustworthiness
Nurse Educ Today
(2004) - et al.
The Roter interaction analysis system (RIAS): utility and flexibility for analysis of medical interactions
Patient Educ Couns
(2002) - et al.
Development of a questionnaire to assess communication preferences of patients with chronic illness
Patient Educ Couns
(2011) - et al.
How patients perceive a doctor's caring attitude
Patient Educ Couns
(2008) - et al.
Three approaches to qualitative content analysis
Qual Health Res
(2005) - Mayring P. Qualitative content analysis. Forum Qual Soc Res 2000;1....
- et al.
The qualitative content analysis process
J Adv Nurs
(2008) - Moretti F. GULiVER—travelling into the heart of good doctor–patient communication from a patient perspective. An...
On the pragmatics of qualitative assessments. Designing the process for content analysis
Eur J Psychol Assess
(2006)Basic content analysis
(1990)
Qualitative data analysis for health services research: developing taxonomy, themes, and theory
Health Serv Res
Qualitative analysis of content
Principles of data analysis
Naturalistic inquiry
Cited by (264)
Father engagement in Syrian Asylum-Seeker families with children on the autistic spectrum
2023, Research in Autism Spectrum DisordersTowards the adoption of most suitable green walls within sustainable buildings using interval type-2 fuzzy best-worst method and TOPSIS technique
2024, Engineering, Construction and Architectural ManagementOrganizational learning and adaptation in participatory Islamic finance
2024, Qualitative Research in Financial Markets