020 7848 6782 | info@rdslondon.co.uk
 
Research Design Service London
 

 
Qualitative research
RDS London guides:

Watch 'Understanding Qualitative Research' with Jill Russell, RDS London methodology lead

Loading the player...





Writing a Qualitative Research Proposal
 
(Download as a PDF)

When to use qualitative methods

 
Qualitative methods should be used when the aim is to:
  • Investigate complex phenomena that are hard to deconstruct quantitatively, perhaps as part of a mixed methods study
  • Generate data to illustrate a problem and help others understand it
  • Gain insights into possible causality
  • Develop quantitative measurements processes or instruments
  • Study characteristics of unusual or marginalised populations
Table 1: Examples of data sources used in qualitative research and the types of research question they might answer
What the research question is about Source of data
Beliefs, feelings, perceptions, ideas about a particular topic or concept or intervention or illness Interviews and focus groups, websites and fora, media articles
Group norms and shared experiences, exploration of the socially marginalised Focus groups, websites and fora, media articles
Behaviours in natural settings, examination of situations or processes, impact of technologies and interventions Observational (e.g. ethnography, video recording of real life)
Culture (of a people, of an organisation), processes and consequences Material artefacts, real life documents, websites and fora, media articles
Development of a validated questionnaire or sets of core outcomes Focus groups and cognitive interviews
Feasibility of a process, success of a training programme, barriers and facilitators, acceptability of an intervention or study design Observation and interview
Patterns of communication, characterisations of organisations or processes Textual analysis; archival research
Development of typologies of participants or other categories Framework approach to thematic analysis

What to do before meeting your qualitative advisor
  • Check the literature. There is no point investing a lot of effort in designing a study only to find that it has been done before. Since qualitative studies tend to be less generalizable than quantitative studies, they may not be prominent in the literature but they may exist.
  • Make sure you cannot get data from a repository such as Qualidata instead. It is not ethical to do primary research if you can answer your research question using a secondary analysis of existing data for which you have sufficient contextual information about the population the data come from, the way the data were collected, and who by.
  • Develop your research question and then think of possible issues:
    • Most researchers choose focus groups / interviews, with a thematic analysis using the Framework approach, without thinking about whether this is the best match to the research question. But don’t fall into the trap of doing what feels comfortable or what is usual. The tables here may help. If a less familiar approach is needed, get a relevant expert as a coapplicant or collaborator.
    • Consider whether your field researcher should or should not match participants in terms of key variables such as gender or ethnicity. There are merits to both positions. It is now generally accepted that we all occupy multiple identities (e.g. by gender, age, social class, ethnicity) that may become temporarily relevant at any time during the fieldwork. But there are some clear situations where matching is important. For example, women should generally interview women about a personal genderspecific problem to take account of sensitivities. And what about situations where you are delivering an intervention and then wish to find out whether patients liked it or not. Should you be doing the interviews or should someone else? Either alternative is possible, so long as you provide a rationale for your choice.
    • List the inclusion and exclusion criteria for your patients and then consider whether you are including too wide or too narrow a group and the implications of changing your criteria (and research question). How will the feasibility and the usefulness of your study be affected?
    • Where will you get your participants? Will you choose a setting with a large number of relevant participants, or will you select them more randomly? Will your aim be to maximise diversity or recruitment efficiency? Will your setting drive your study design or vice versa and how can you justify this? Usually it is better in qualitative research to go for maximal diversity purposive sampling, to get rich data, or convenience sampling to be pragmatic. Given the likely small final sample size and the possibility, depending on your topic, setting and resources, that a large proportion of people will not want to participate, random sampling is unlikely to be sustained. With hard to reach groups, snowball sampling may be best, where one participant recommends another. But be clear about any biases. Where the aim is to develop theory, sampling should be theoretical, that is predictions should be made based on the existing theory and then participants sought to test the theory and predictions and fill in any gaps in the existing theory.
  • Begin to think about patient and public involvement (PPI) and make sure you do not confound this with research. A focus group with community representatives who are evaluating your intervention as part of your study and who need to sign consent forms does not constitute PPI involvement.
  • Sketch a brief outline of your study that you can work on with your advisor.
Things to watch out for when writing your application
 
Method
  • Your audience is likely to include:
    • Qualitative panel members who are likely to be experts in only one or a few methodologies and so may need convincing of the merits and quality of another approach.
    • Quantitative researchers, as the majority, with many considering qualitative research as lacking in rigor. You need to show them otherwise.
  • Give your approach a theoretical underpinning. It is not sufficient to say that theory will emerge from the study – this is different to the theory that oriented the research design. For example, you may choose to consider agency in accessing healthcare, underpinned from the start by theories of self-efficacy; your later emerging theory may show that agency is affected by barriers to access that reduce self-efficacy.
  • Tie in your methodology to your research question and goals. Qualitative research can often sound quite aimless and non-rigorous so you need to show that your work will meet the highest scientific scrutiny.
  • If using interviews state how directive or non-directive you plan to be, give some sample questions and indicate how long you expect the interviews to last.
  • Make sure your team of researchers has sufficient breadth of expertise to cope with changes to the design that the panel may request. For example, if you have suggested an interview based study but the panel consider an ethnographic study to be more appropriate, will someone on the team be able to manage this to a sufficiently high standard? Or will the panel reject your application because they cannot?
 Impact
  • Don’t exaggerate the potential impact, relevance and transferability of findings. Qualitative research is not generalizable to the general population but is relevant only to the cases considered.
Outcomes
  • Qualitative research is often undertaken when little is known about a topic. This means a qualitative research proposal cannot be as clear in the detail as a quantitative one. Qualitative research is often exploratory and develops iteratively. It may be hard to specify what your outcomes are likely to be, beforehand. This should not stop you from writing your research goals and suggesting likely outcomes. For example if you intend to develop a new training programme and the format this will take is dependent on your preliminary qualitative research, this does not prevent you from giving some broad indication of expected outcomes. It is all right to say “Findings from the qualitative study will dictate the format of the training programme and so we cannot specify this in detail beforehand. However, our preliminary scoping exercise suggests it is likely that it will involve a face to face workshop-style component and online exercises.”
Transparency in your decision making
  • Qualitative researchers often have to make pragmatic choices that would not be acceptable in quantitative studies. If you have to do so, state this clearly rather than attempting to hide it. For example, if you choose to study your own work environment because you already have access to this, write this up as strength that increases the feasibility of your study.
  • If you are holding focus groups be sure to explain what characteristics (e.g. gender, ethnicity) you will use to organise your groups and why these characteristics are the ones of most relevance to your research question.
Making comparisons
  • Try to include multiple sites if you can, that differ in ways relevant to your research question. This will give you a richer study and is likely to find more favour with quantitative researchers than a single site study.
  • Alternatively you could compare different populations within one community.
  • Mapping the various viewpoints of different stakeholders in your topic of interest can be a very powerful approach. For example, comparing the viewpoints of healthcare providers and patients by mapping similar themes onto each other can show you where service improvements are more or less likely to succeed. 
Analysis
  • Quantitative reviewers like numbers. So don’t simply say you will sample until saturation of themes, but that you will spend x hours a week for y weeks sampling, until you have reached saturation of themes, which you would expect to achieve with 20-30 participants. This enables reviewers to consider the feasibility of your plans.
  • Allow sufficient time for analysis which can take from a few days or weeks to several months depending on the type of qualitative research.
  • Describe the analytical steps in detail so the panel can see you know your stuff.
Table 2: Analytical approaches
Method What they consider
Grounded theory, thematic analysis Processes
Ethnography Behaviours of everyday life
Phenomenonology Meaning and feelings
Discourse analysis Deconstructive reading and interpretation of a problem or text showing how people construct concepts of the social world, social practices
Conversation analysis Communication, language, interaction
Narrative analysis How (life) stories are constructed and what influences them
 
Author: Dr Carol Rivas, RDS London and Queen Mary University of London

Back to top