In this section we will cover some of the reasons why your work may benefit from qualitative research.
Solutions to large global challenges tend to involve designing new systems, protocols, or technologies. To ensure that these new interventions fully meet the needs of the people they are intented to help, users’ unique perspectives and experiences need to be taken into account. Qualitative research can help you learn about people’s attitudes, behaviors and practices more in-depth, and also pays attention to the contextual factors that contribute to the persistence of challenges and different responses to them.
With challenges or contexts that are less understood, you may not always know what questions to ask or who to direct them to. Rather than assuming a certain knowledge, qualitative research helps you identify what you don’t know and offers you a way to begin to find answers. Unlike quantitative research, qualitative research allows you to be more explorative and flexible in your research approach. It can also help you identify the right stakeholders to work with and get detailed user feedback on interventions, helping to avoid spending resources on mis-diagnosed problems or inappropriate design interventions.
While quantitative data provides generalizable, definitive answers, qualitative research can offer important nuance by considering the unique cultural, behavioral, and environmental context that shape the lives of research participants. It is particularly attentive to holding multiple perspectives and more readily seeks out voices that are not usually heard, thereby making sure that those most in need are taken into account when developing interventions. Such nuanced understandings can also inform further quantitative research and thereby help generate multi-perspectival data.
Qualitative research can more organically adapt to research participants’ reactions and responses as they arise in the research process. This allows for questions that are more relevant to the person or situation to be asked. Conversely, if useful insights are not being captured, researchers can pivot more easily to introduce new questions and change the setting or any other variable to improve responses.
Because it builds research rapport and trust with research participants, qualitative research allows you to establish stronger relationships with the people for whom you’re designing an intervention. This is especially important for the sensitive topics that often arise is public health research.
“There was an insight early on around toilets… women wanted to not leave their child outside the toilet door because of the safety of their child. And the insights helped to identify that you're asking women to make choices between using a clean toilet and putting the safety of their child at risk and then helped to drive the design work that created a larger latrine so that she could comfortably bring in her baby.
We have great quantitative data, but the survey data wouldn't necessarily get at that insight.”
“Let's say you're interviewing me you'll get to a point in the questionnaire where you're going to ask me something like ‘you've said you don't want to get pregnant for the next two year period, and yet you're not using contraception, can you tell me the reasons why?’
In a typical quantitative survey, the instructions to the interviewer is to capture the first thing that the respondent says... And the rest of the stuff that I say, it's not going to show up, and it's generally not going to show up in your quantitative codebook.”
Despite the clear benefits of qualitative research, you may be hesitant to include it in your investments or feel uncomfortable managing work with qualitative components. You might feel that qualitative research is not rigorous enough or cannot produce unbiased information. By using this tool, you will be able to ensure that research is carried out in a rigorous, holistic and ethical way, often in combination with quantitative research. This next section gives practical guidance around designing and executing qualitative research projects.
Qualitative research can be stand-alone but is usually used together with quantitative data collection. Below is a brief description of the different ways in which they can be combined, with an emphasis on the unique contribution of qualitative research.
Use qualitative research early on in a project to better inform the scope and direction of the project
Use qualitative data before quantitative research to help narrow in on a more appropriate focus, for example by generating the right questions for a survey
Use qualitative research after quantitative research to dig deeper into survey results, for example by asking how and why follow-up questions
There is no one right way to combine qualitative and quantitative methodologies. It’s more an issue of the questions you are asking that drive decisions about how to sequence, or integrate, the different efforts. Some budget and timing considerations to keep in mind, include:
When you’re integrating multiple research approaches, you may need a longer planning phase.
You will want to think about how to build flexibility into your time frame when you are sequencing approaches.
Be wary of the challenges of "inviting" different partners to work together, without a prior existing relationship. You will want them to determine how to best work together but you may have an important role to play in facilitating those discussions.
Be judicious in selecting and thinking about which organizations are good at collaborating and then understand what it means to actually integrate different methodologies and approaches.
Building more contact points for partners to come together to integrate findings or clarify any issues will make for a more realistic timeline and better coordination.
For the sequencing of the approaches, plan that all the partners are present throughout the full process, with different emphases at the various project stages and despite the fact that their contribution is sequenced.
Since all qualitative research approaches involve interactions with human participants, any project that uses them should get ethical approval to ensure participants can give informed consent. Any potential risk for the participant is clearly explained and mitigated as much as possible, and that confidentiality and anonymity are protected. The latter can be difficult when visual documentation of research in photos or videos is taking place, and requires careful consideration of how to represent research participants. While ethical approval is especially important when research involves vulnerable groups and when the goal is to publish findings in academic journals, many country governments are demanding that ethical approval is sought when there is any interaction with human participants, regardless of plans to publish. You will need to consult in-country partners and government regulations to determine the standard to which your project will be held.
Qualitative sampling methods strategically select small samples of people or cases with the intention of understanding the complexity, depth, variation, or context surrounding a phenomenon (This is in contrast to quantitative sampling methods which use random sampling to generate representative data). This means that qualitative sampling approaches generate information-rich data from the most knowledgeable respondents.
Researchers begin with specific perspectives they wish to examine and then seek out research participants who cover that full range of perspectives. While purposive sampling is often used when the goal is to include participants who represent a broad range of perspectives, it may also be used when a researcher wishes to include only people who meet very narrow or specific criteria.
Researchers recruit participants who are easily accessible and convenient to them. This often includes tapping into existing social networks or finding participants in a specific location. This method is most useful in exploratory research but caution should be applied to generalizing any insights.
Researchers start with a small number of participants and then rely on referrals from the original group to find new participants. This is especially useful when studying populations which may be difficult to find or are stigmatized, making larger scale recruitment difficult. By utilizing referrals from participants, the researcher is able to gain additional trust.
Researchers intentionally select extremes and try to identify the factors that affect them. It’s most commonly used in design research, where outliers serve as inspiration for solutions that can be introduced to a wider population. This sampling method rests on the notion that products or services designed for the most difficult to reach user will also speak to the more easy to reach user.
Researchers identify categories of perspectives that are important to the study, creating a subgroup for each category, and decide how many people to include from each. One thing to note is that segments can be both qualitatively and quantitatively derived, and the level of rigor depends on the questions your research study seeks to answer as well as how results will be triangulated or validated.
There is no straightforward answer to what the ideal sample size for a qualitative research project is. This depends on the nature of the topic, study scope and design, diversity of participants, number of geographies and the richness of data collected from each participant. In general, studies asking broader questions of diverse participants need a larger sample size than studies that focus on a homogeneous participant group and more specific questions. While it is therefore difficult to put specific numbers on sample sizes, for qualitative research they generally vary between 10 and 60 participants.
The most common, although not uncontested, principle to determine sample size is data saturation, which is that point in a research process when the same findings keep surfacing and no new information will be gained from adding more research participants. Also called informational redundancy, saturation signals to researchers that data collection can be wrapped up. Other criteria sometimes used in determining sampling size are pragmatic considerations such as budgetary or time constraints or the difficulty of accessing more research participants.
Sample size and composition are important to ensure the quality, validity and generalizability of qualitative research. Because in qualitative research sample sizes are small, an additional way to ensure quality and validity is data triangulation, which uses two or more methods for the verification of research findings. This can include using different data sources (e.g., field and desk-based research) or methodological triangulation (e.g., interviews, participant observation, and surveys).
Single project in a single geography
Project: Barriers and Enablers to Ebola Treatment-seeking behavior
The team combined qualitative and quantitative research methods with 350 individuals across 11 chiefdoms in Port Lok. They conducted 20 semi-structured interviews and 10 focus groups with community members, as well as 8 structured interviews with community health volunteers, 108 questionnaires with volunteers, and observations across the board.
Medium project in multiple geographies
Project: Vx Data Insights
Kenya, Mozambique, DRC
The team used a human-centered design approach to study data underutilization in immunization programs and identify priority intervention areas. They conducted a total of 198 in-depth contextual interviews with health care workers and managers from community to national level (83 in Kenya, 64 on Mozambique, 64 in DRC)
Large project in multiple geographies and different participant categories
Kenya, India, Nigeria
2018 - ongoing
The team used a design-led mixed-methods approach combining ethnographic research, cultural analysis, behavioral and data science. So far this has involved X interviews and Y focus groups with community members.
In this tool, research approaches are defined as the overarching frameworks that guide the collection, analysis and interpretation of data. There is no complete agreement on what these approaches are, and in order for this tool to be most useful to practitioners making investment decisions, five different approaches have been selected to dive further into. These approaches may be more or less relevant to your work, depending on the scope, outputs, and type of engagement of your project.Compare all approaches
In this approach, research participants are actively involved in the research process, keep control of its results and participate in translating research findings into concrete change. A typical process follows an iterative cycle of data collection, reflection, discussion and action, with participants collaborating at every step.
Strength in understanding how to address a problem
Biased towards action
Aims to co-create
A highly immersive approach that relies on researchers interacting with participants in their natural environment. It is particularly useful when sensitive topics and/or marginalized populations are involved, where getting participants to speak openly in formal settings is more difficult. The building of rapport and trust between researcher and participants is essential.
Strength in exploring little known problems
Biased towards rich knowledge generation
Aims to understand deeply
This research approach supports the development of products, services, and programs. It is iterative, rapid and adaptive, giving researchers the flexibility to follow hunches and to change their questions mid-stream. An important aspect of design research is immersion, in order to develop empathy with end-users and to elicit useful feedback on initial designs and successive iterations.
Strength in generating actionable insights
Biased towards action
Aims to empathize
This approach focuses on the study of individual and social behavior through replicable research following strict protocols. It is used to generate theories that could be tied to design or intervention goals, to build a large evidence base and to contribute to new scientific knowledge.
Strength in understanding a specific problem
Biased towards scientific knowledge generation
Aims to remain objective
This is a systemic approach to understand what, why, and how interventions work in real world settings and to test approaches to improve them. It focuses on generating evidence to understand the factors that support or inhibit high quality implementation, in order to achieve the intended effects of interventions.
Strength in understanding a specific intervention
Emphasis on evidence generation
Aims to remain objective
For the purposes of this tool, four high level objectives have been identified that your project may be working towards: (1) explore problems and identify user needs, (2) convert these needs into interventions, (3) test and evaluate interventions, and (4) adopt and scale interventions. An investment can focus on a single objective or span across all four. Some qualitative research approaches are better suited to address an objective than others, which we explore in the chart below.Compare all approaches
Qualitative research can contribute to understanding problems' contexts and root causes and identifying user needs.
Design, implementation and evaluation
Qualitative research can provide in-depth actionable insights that inform the design and implementation of interventions.
Qualitative research can provide user insights and additional contextual information into an intervention’s performance.
Adoption and scale
Insights about specific user groups and contexts can inform how an intervention can be scaled to maximize impact.
Get inspired by exploring the library of case studies and research methods so you can take your investment further.
A list of 30+ research methods where you can learn more about them and how they are linked with the different approaches.Explore all methods
Read a selection of case studies where you can learn how approaches and methods were used throughout the investment.Explore projects
Consult all the qualitative expert vendors to consider for your next investment.View vendor list