Choosing a research method is one of the most important decisions in a study. The method affects the questions you can answer, the data you collect, the people you involve, the analysis you perform, and the strength of your conclusions.
The two broad approaches are qualitative and quantitative research. Qualitative research explores experiences, meanings, perceptions, motivations, and processes. Quantitative research measures variables using numerical data and statistical analysis. Neither method is automatically better. The right choice depends on the research question and the decision the evidence needs to support.
Qualitative research is used to understand how people interpret experiences, how processes work, and why a particular situation occurs. It produces detailed, non-numerical information, usually through interviews, focus-group discussions, observation, document review, or open-ended questions.
For example, a qualitative study might explore why community members are not using a health service, how staff experience a new organisational policy, or what prevents local partners from applying a capacity-building intervention.
Qualitative research is particularly useful when:
•the topic is not well understood;
•the study seeks to understand perceptions or motivations;
•the context is central to the research question;
•the researcher needs to explore unexpected issues; or
•participants’ experiences and voices are important to the decision.
The output may include themes, explanations, narratives, process maps, quotations, and interpretations.
Quantitative research collects numerical data to describe patterns, compare groups, test relationships, estimate prevalence, or measure change. Common methods include structured surveys, assessments, routine monitoring data, experiments, and analysis of administrative records.
For example, a quantitative study might estimate the proportion of staff who completed training, compare service satisfaction across locations, measure changes in knowledge scores, or examine whether a programme reached its target population.
Quantitative research is particularly useful when:
•the research question requires measurement;
•the study needs to compare groups or locations;
•the organisation wants to estimate the size or frequency of a problem;
•change needs to be tracked over time; or
•findings need to be summarised across a larger population.
The output may include percentages, averages, rates, comparisons, statistical tests, charts, and tables.
|
Dimension |
Qualitative research |
Quantitative research |
|
Main purpose |
Understand meaning, experience, and process |
Measure patterns, size, change, or relationships |
|
Typical data |
Words, observations, documents, narratives |
Numbers, scores, counts, ratings |
|
Common methods |
Interviews, focus groups, observation |
Surveys, assessments, datasets, experiments |
|
Sample |
Usually smaller and purposive |
Often larger and designed for comparison or estimation |
|
Analysis |
Coding, thematic analysis, interpretation |
Statistical analysis and numerical comparison |
|
Main strength |
Depth and contextual understanding |
Measurement and comparability |
|
Common limitation |
Findings may not be statistically generalisable |
Results may not explain why a pattern exists |
The first question should be: what decision will this research inform? A study designed to improve a service may need to know how many users are affected, why the problem occurs, and which solution is practical. That may require more than one type of evidence.
Questions that begin with “how many,” “how often,” “what proportion,” or “to what extent” often require quantitative methods. Questions that begin with “how,” “why,” or “what is the experience of” often require qualitative methods.
This is a useful starting point, not an absolute rule. A well-designed study should consider the full context of the decision.
Quantitative studies often require a sample that supports the intended comparison or estimate. Qualitative studies generally select participants because they can provide relevant experience or insight. In both cases, the sample should be justified rather than chosen only for convenience.
The best design must be feasible. Consider the time available, research skills, access to participants, data-protection requirements, travel, translation, analysis capacity, and budget. A small, well-designed study is better than an ambitious study that cannot maintain quality.
Mixed-methods research combines qualitative and quantitative approaches in one study. It is useful when numerical patterns need explanation or when personal experiences need to be understood alongside broader measurement.
For example, an organisation might conduct a survey to measure staff confidence in a new system and then hold interviews to understand why confidence differs between departments. A programme might analyse attendance and outcome data and then speak with participants to understand barriers to access.
Mixed methods can be structured in different ways:
•Sequential explanatory design: Collect quantitative data first, then use qualitative research to explain the results.
•Sequential exploratory design: Conduct qualitative research first, then use the findings to develop a survey or measurement tool.
•Concurrent design: Collect qualitative and quantitative data during the same period and compare the findings.
The approaches should be connected by a clear research question and analysis plan. Simply collecting two types of data does not automatically create a strong mixed-methods study.
One mistake is choosing a familiar method rather than the method that fits the question. Another is writing a questionnaire before defining the information needed for the decision. Researchers may also collect data from a convenient sample without explaining its limitations.
Other problems include poorly worded questions, inadequate training for data collectors, weak consent procedures, inconsistent coding, and presenting results without discussing uncertainty or limitations.
A credible study should document the research design, population, sample, data-collection tools, analysis process, ethical safeguards, limitations, and implications for decision-making.
|
If you need to understand… |
Consider… |
|
The size or frequency of a problem |
A quantitative survey or routine dataset |
|
Why people behave in a certain way |
Interviews, focus groups, or observation |
|
Whether outcomes changed |
Baseline and follow-up quantitative measurement |
|
How an intervention was experienced |
Qualitative interviews or case studies |
|
Differences between locations or groups |
Quantitative comparison, supported by qualitative explanation |
|
An unfamiliar issue before designing a survey |
Exploratory qualitative research |
|
Both the scale and the reasons behind a problem |
A connected mixed-methods design |
Qualitative and quantitative research answer different types of questions. Qualitative research provides depth, context, and insight into experience. Quantitative research provides measurement, comparison, and evidence about scale or change. Mixed-methods research can combine their strengths when the decision requires both measurement and explanation.
The most important principle is to begin with the research purpose. Define the decision, formulate clear questions, choose a feasible design, and explain the limits of the evidence. A method is valuable when it produces reliable information that people can use.
Global Capacity Lab supports researchers, M&E teams, product and strategy leaders, and organisations that need to turn evidence into action. Learn more about GCL’s research and insights services or contact the team about research support and training.
Suggested related links: