Qualitative research is an approach used to develop an in-depth understanding of experiences, perceptions, meanings, behaviours, relationships, processes, and contexts.
Instead of focusing primarily on numerical measurement, qualitative research often seeks to understand how people interpret situations and why particular experiences or behaviours occur.
Qualitative research is widely used in:
The appropriate qualitative design depends on the research question, participants, setting, and type of understanding the study seeks to develop.
Qualitative research investigates phenomena using non-numerical or primarily textual, visual, observational, or experiential evidence.
Common sources of qualitative data include:
The objective is often to understand meaning, experience, process, context, or interpretation.
Qualitative research does not mean that numbers can never appear in a study.
For example, a qualitative project might report that:
12 participants described concerns about data privacy.
The important feature is that the study’s primary analytical purpose is not statistical estimation but understanding the nature and meaning of the participants’ experiences.
Qualitative research can be particularly useful when the researcher wants to understand:
It may be less suitable when the primary objective is to estimate a population proportion or statistically test a relationship between numerical variables.
Qualitative research questions often focus on:
Examples include:
How do employees experience the introduction of artificial intelligence into their daily work?
How do small business owners perceive cybersecurity risks?
What factors shape students’ experiences of online learning?
How do managers describe the challenges associated with remote teams?
The exact wording should reflect the research purpose.
Interviews are only one qualitative method.
A qualitative study may involve:
The researcher should select methods that can produce the type of information required to answer the research questions.
Several qualitative approaches are commonly used.
These include:
The terminology and boundaries between these approaches can vary across disciplines.
The important issue is to understand what the chosen design is intended to accomplish.
A case study examines a defined case in depth.
The case may be:
A case must have reasonably clear boundaries.
For example:
The implementation of an AI-assisted customer-service system within one selected company during a defined period.
The researcher can then examine the case using several sources of evidence.
Phenomenological approaches generally focus on people’s lived experiences of a phenomenon.
For example, a study might explore:
How do newly qualified nurses experience their transition into professional practice?
The emphasis is on understanding the experience and the meanings participants attach to it.
Researchers should use a phenomenological approach only when it is genuinely appropriate to the research purpose.
Grounded theory approaches aim to develop an explanatory framework or theory from systematically collected and analysed data.
The research process may involve an iterative relationship between:
The precise procedures vary between grounded theory traditions.
It should not be described simply as “coding interviews.”
Ethnographic research generally seeks to understand people, practices, behaviours, and culture within a social setting.
It may involve:
Ethnographic research can require substantial engagement with the research environment.
Narrative research examines stories and how people describe experiences through narratives.
Researchers may be interested in:
The analytical focus is on how experiences are narrated and interpreted.
Qualitative descriptive approaches aim to provide a relatively direct and comprehensive account of participants’ experiences or perceptions.
They can be appropriate when the research objective is primarily descriptive rather than developing a formal theory or philosophical interpretation.
The design should still involve systematic data collection and analysis.
Qualitative research generally uses purposeful participant selection rather than attempting to obtain a statistically representative sample.
The researcher identifies participants who can provide relevant information about the research problem.
Selection criteria may include:
The criteria should be connected to the research questions.
Purposive sampling involves deliberately selecting participants because they meet characteristics relevant to the study.
For example, a study of AI adoption among business managers might recruit managers who have direct experience implementing or using AI tools.
The purpose is information relevance rather than statistical representativeness.
Convenience sampling involves recruiting participants who are readily accessible.
It can be practical, but it may introduce important limitations.
For example, people who are easiest to recruit may differ systematically from people who are harder to reach.
If convenience sampling is used, its implications should be acknowledged.
Snowball sampling involves participants helping researchers identify other potential participants.
It can be useful when researching populations that are:
However, recruitment through existing networks can also create selection bias.
Qualitative research does not have one universal sample-size rule.
The appropriate number of participants depends on:
A small qualitative sample can be appropriate for an in-depth study.
A larger sample may be necessary when the research covers diverse groups or multiple contexts.
Researchers may consider whether the collected information is sufficiently rich and relevant to address the research question.
Factors affecting adequacy can include:
Do not use a predetermined number as the only justification for qualitative sample size.
Interviews are one of the most widely used qualitative methods.
They allow researchers to explore:
The quality of an interview depends heavily on the researcher’s preparation and interviewing technique.
Structured interviews use a predetermined set of questions presented in a consistent format.
They can improve comparability between participants.
However, highly structured interviews may provide less opportunity to explore unexpected issues.
Semi-structured interviews combine consistency with flexibility.
The researcher prepares an interview guide but can:
This format is widely useful for exploratory qualitative research.
Unstructured interviews provide substantial flexibility.
The researcher may begin with a broad topic and allow the participant’s responses to shape the discussion.
This can produce rich information but requires considerable interviewing skill.
Good qualitative interview questions should generally be:
For example:
Weak:
Don’t you agree that the new system has improved productivity?
Better:
How has the new system affected your work?
The second question allows the participant to express positive, negative, or mixed experiences.
Leading questions can influence responses.
Instead of:
How beneficial was the training?
consider:
How would you describe your experience of the training?
The second formulation does not assume that the training was beneficial.
Useful follow-up questions include:
Follow-up questions can reveal details that a predetermined questionnaire might miss.
Focus groups involve guided discussions with multiple participants.
They can help researchers understand:
The interaction between participants can produce information that individual interviews may not reveal.
Potential challenges include:
The researcher should consider group composition and facilitation carefully.
Observation can provide information about behaviour in context.
It can be useful when people:
Observation should be systematic rather than simply watching events without a plan.
Before collecting observational data, define:
A structured observation schedule may be appropriate in some projects.
Field notes may capture:
Separate direct observations from interpretations where possible.
For example:
Observation:
The participant paused for approximately ten seconds before answering the question.
Interpretation:
The participant may have been uncertain about how to respond.
Keeping these distinctions clear can improve analytical transparency.
Qualitative research can analyse existing documents.
Potential sources include:
Researchers should establish why particular documents were selected and how their authenticity and relevance were assessed.
Depending on the research question and ethics requirements, qualitative research may also examine:
Researchers should consider:
Public availability does not automatically mean that every form of online information can be used without ethical consideration.
Researchers may record interviews when appropriate and when participants have provided the required consent.
Recording can improve accuracy during transcription and analysis.
Researchers should have a clear procedure for:
Transcription converts recorded speech into a written form.
The required level of detail depends on the analytical purpose.
A basic transcript may capture spoken words.
A more detailed transcript may also record:
The researcher should use a consistent transcription approach.
Coding involves assigning labels to meaningful segments of data.
For example, interview excerpts concerning AI adoption might be coded:
Codes help organise large quantities of qualitative information.
Initial coding involves identifying potentially meaningful ideas in the data.
Researchers should avoid forcing every statement into predetermined categories unless the research design specifically calls for a deductive framework.
Unexpected ideas may be analytically important.
Related codes can be grouped into broader categories.
For example:
Codes
could contribute to:
Category: Training and capability barriers
This process reduces complexity while retaining meaningful distinctions.
Themes are broader patterns of meaning that help answer the research question.
For example:
Codes
may contribute to:
Theme: Organisational readiness
A theme should represent something analytically meaningful, not simply a topic mentioned frequently.
Thematic analysis is a widely used approach for identifying and interpreting patterns in qualitative data.
A practical workflow may involve:
The exact process should reflect the chosen analytical framework.
Inductive analysis allows patterns and concepts to emerge from the data rather than beginning entirely from predetermined categories.
This can be useful when:
Inductive analysis does not mean that researchers have no prior assumptions.
Researchers always bring some perspective to the research process.
Deductive analysis begins with concepts, theories, categories, or questions established before or outside the immediate dataset.
For example, a researcher might analyse interviews according to an established theoretical framework.
Deductive analysis can provide structure while still allowing new observations to emerge where appropriate.
A study can combine inductive and deductive processes.
For example:
This can be useful for complex research questions.
Constant comparison involves comparing:
The objective is to identify similarities, differences, patterns, and exceptions.
This approach is especially relevant to several qualitative analytical traditions.
A negative case is evidence that does not fit an emerging pattern.
For example, if most participants report that management supports AI adoption but several participants describe strong management resistance, those contrasting accounts should not automatically be discarded.
Negative cases can improve the depth of analysis by challenging premature conclusions.
The term saturation is used in several qualitative research traditions.
In broad terms, it may refer to a point at which additional data collection is no longer producing sufficiently new information for the purpose of the study.
Researchers should be precise about what type of saturation they mean.
Saturation should not be treated as a magic numerical threshold.
Researchers can influence qualitative research through:
Reflexivity involves considering these influences.
A reflexive researcher may maintain:
The objective is not to pretend that the researcher has no perspective, but to make relevant influences more visible.
Positionality refers to the researcher’s relationship to the research context and participants.
Relevant factors can include:
The importance of positionality varies by research design and discipline.
Qualitative researchers commonly consider:
These concepts address different aspects of research quality.
Credibility concerns whether the interpretation is sufficiently supported by the collected evidence.
Potential strategies include:
Transferability concerns whether readers can determine whether findings may be relevant to another context.
Researchers can support this by providing sufficient contextual detail.
The objective is not necessarily to claim statistical generalisation.
Dependability concerns the consistency and transparency of the research process.
Researchers can improve transparency by documenting:
Confirmability concerns the extent to which findings are grounded in the data rather than being solely the researcher’s preferences or assumptions.
Useful practices can include:
Triangulation can involve combining:
For example, a study could compare:
Triangulation should be used meaningfully rather than simply to increase the volume of data.
Member checking involves sharing certain interpretations or findings with participants for feedback where appropriate.
It may help identify:
However, participant agreement does not automatically determine whether an academic interpretation is correct.
The appropriateness of member checking depends on the research design and methodological tradition.
Qualitative research can involve sensitive and personal information.
Important considerations include:
Ethical requirements should be established before data collection.
These concepts are related but not identical.
Anonymity generally means the participant’s identity is not known or cannot reasonably be linked to the data.
Confidentiality means identifying information may be known to the researcher but is protected and not improperly disclosed.
Researchers should describe their procedures accurately.
Researchers may use:
Be careful when reporting quotations.
A quotation can reveal identity indirectly if it contains distinctive information about the participant or organisation.
Research data may include:
Researchers should establish secure storage and access procedures.
Where institutional or legal requirements apply, follow the applicable requirements for retention and deletion.
A qualitative findings chapter should not simply reproduce interview transcripts.
Instead, organise findings around:
Use selected quotations to illustrate important findings.
A useful quotation should:
Avoid filling the chapter with long quotations.
The researcher’s analysis remains central.
The findings section generally presents what the analysis revealed.
The discussion interprets those findings in relation to:
The exact division varies by discipline.
Qualitative findings should not automatically be presented as universal.
Instead of:
Employees do not trust AI.
a more defensible formulation might be:
Participants in this study frequently expressed concerns about trusting AI-generated outputs.
The second statement accurately reflects the study’s evidence and scope.
Qualitative research can produce important explanations and insights without requiring statistical generalisation.
However, researchers should not make numerical population claims that their design cannot support.
For example:
Most employees in the country believe…
would require evidence capable of supporting that broader claim.
Qualitative research can complement quantitative research.
For example:
Quantitative stage
A survey identifies that employee AI adoption is lower among certain groups.
Qualitative stage
Interviews explore why adoption differs.
The qualitative findings can provide context for interpreting the numerical results.
AI tools may assist with some research tasks, such as:
However, researchers should not assume that automated outputs are accurate.
Particular care is required when using AI for:
Researchers remain responsible for the final analysis.
Do not upload confidential participant information to an AI service unless the use is authorised and appropriate safeguards are in place.
A topic such as:
People’s experiences of technology.
may be too broad.
Define the:
Questions that assume a particular answer can distort the data.
Use neutral wording.
A theme should represent a meaningful analytical pattern.
Not every frequently mentioned topic is necessarily a theme.
Coding should contribute to answering the research questions.
Do not create hundreds of labels that have no analytical value.
Differences between participants can be important findings.
Do not discard information simply because it does not fit the dominant pattern.
A qualitative study should contain analysis, not merely participant statements.
Reporting:
Six participants mentioned training.
is descriptive.
Asking:
What does the participants’ discussion of training reveal about organisational readiness?
moves toward analysis.
A qualitative study usually does not justify broad numerical population claims.
Explain the scope of the evidence accurately.
The researcher’s position can affect qualitative research.
Relevant influences should be considered and documented.
Suppose the research question is:
How do employees experience the introduction of AI-assisted software in their workplace?
A possible design could be:
Approach
Qualitative.
Participants
Employees who have directly used the software.
Sampling
Purposive sampling based on direct experience.
Data collection
Semi-structured interviews.
Analysis
Thematic analysis.
Potential themes
Ethics
Informed consent, confidentiality, secure data storage, and appropriate handling of potentially sensitive workplace information.
The exact design would need to be justified in the context of the actual research project.
Identify exactly what you want to understand.
Frame questions around experiences, meanings, processes, or perceptions where appropriate.
Choose an approach that matches the purpose.
Determine who can provide relevant information.
Select participants using an appropriate approach.
Prepare interview questions, observation schedules, or other tools.
Identify problems before the main data collection.
Conduct interviews, observations, focus groups, or other planned activities.
Store recordings, transcripts, notes, and other materials securely.
Read and review the material carefully.
Identify meaningful segments and assign appropriate labels.
Group related findings into analytically useful patterns.
Look for contradictory or negative cases.
Check whether the themes are supported by the data.
Ensure the analysis addresses the actual purpose of the study.
Interpret findings in relation to previous research where appropriate.
Explain relevant limitations concerning participants, setting, data, and interpretation.
Before completing a qualitative study, ask:
The purpose is generally to develop a detailed understanding of experiences, meanings, perceptions, processes, behaviours, or contexts.
There is no universal number.
The appropriate sample depends on the research design, research question, participant characteristics, information richness, and analytical requirements.
Qualitative research involves interpretation, but that does not mean it should be arbitrary.
Researchers can improve rigor through systematic procedures, transparent documentation, reflexivity, appropriate sampling, careful coding, and evidence-based interpretation.
Yes.
Numbers may be used descriptively, such as reporting the number of participants who mentioned a particular issue.
The key issue is how the evidence is analysed and what claims are made from it.
Yes, where interviews are appropriate to the research question and design.
However, the researcher should justify why interviews provide the evidence required.
Not necessarily.
Structured interviews require greater consistency, while semi-structured interviews allow appropriate follow-up questions.
The level of flexibility should be consistent with the research design.
Thematic analysis is an approach for identifying, organising, analysing, and interpreting meaningful patterns or themes within qualitative data.
A code is generally a more specific label applied to a segment of data.
A theme is a broader analytical pattern that brings related codes and ideas together.
Saturation can refer to a point where additional data collection is no longer producing sufficiently new information for the analytical purpose of a study.
The precise meaning depends on the methodological tradition.
Yes.
Rigorous qualitative research requires appropriate design, systematic data collection, careful analysis, transparent reasoning, ethical practice, and conclusions that are supported by the evidence.
Continue exploring the Lamtas academic research resources:
Explore research planning, qualitative research support, data analysis, research reports, evidence-based projects, and consulting services.
Explore educational support, tutoring, training, learning resources, and research-related development.
Explore research writing, editing, reports, documentation, technical writing, and professional content services.
Explore artificial intelligence, data analysis, analytics, automation, and data-driven technology services.
Qualitative research is most valuable when the researcher needs to understand more than measurable outcomes.
It can reveal:
A strong qualitative study connects the entire process:
Research question → Qualitative design → Participant selection → Data collection → Coding → Themes → Interpretation → Conclusions
The strength of the research comes not from collecting the largest possible amount of information, but from collecting relevant evidence and analysing it carefully, transparently, and ethically.