A one-to-one interview is a planned conversation in which one interviewer collects information from one participant. In research, it is commonly used to explore experiences, beliefs, decisions, or expert knowledge in depth. It may be structured, semi-structured, or unstructured and conducted face-to-face, by telephone, or online.
Introduction
One-to-one interviews are widely used in social science, education, healthcare, business, psychology, evaluation, user research, and many other fields. They allow a researcher to concentrate on one participant’s account without the influence of other participants.
However, conducting an interview involves more than preparing a list of questions. Researchers must decide who to interview, what degree of structure to use, how to obtain consent, how to record and protect the data, and how the answers will be analysed.
This guide explains:
- What a one-to-one interview means in research.
- How it differs from an in-depth interview and a focus group.
- The main types and delivery modes.
- How to select participants and develop an interview guide.
- How to conduct, transcribe, analyse, and report interviews.
- The method’s advantages, limitations, ethical risks, and common mistakes.
- How digital tools and artificial intelligence can be used responsibly.
Key takeaways
- “One-to-one” describes an interview involving one interviewer and one participant; it does not by itself determine whether the study is qualitative or quantitative.
- Semi-structured interviews are frequently used when researchers need both consistency and flexibility.
- The interview guide should be derived from the research questions and intended analytical method.
- There is no universal minimum number of interviews; sample adequacy must be justified methodologically.
- Interview data are co-produced through interaction, so interviewer influence and reflexivity must be considered.
- Recording, transcription, AI use, confidentiality, and data retention should be addressed in the ethics and data-management plan.
What Is a One-to-One Interview?
A one-to-one interview, also called an individual interview or one-on-one interview, is a data-collection format in which one researcher or trained interviewer communicates with one participant at a time. Its purpose is to obtain information relevant to a defined research question, evaluation objective, or practical decision.
The interview may be:
- Highly standardised, with identical questions and response choices.
- Semi-structured, with core questions and flexible follow-up probes.
- Open and exploratory, with only broad topics prepared in advance.
The conversation may take place in person, through a telephone call, by video conference, or through another synchronous communication platform.
Is a One-to-One Interview Always Qualitative?
No. A one-to-one interview is commonly qualitative, but it is not automatically qualitative.
A structured interviewer-administered questionnaire may produce numerical or categorical data suitable for quantitative analysis. A semi-structured interview usually produces qualitative text, although selected responses may also be categorised or counted. Some mixed-methods studies collect both forms of data in the same interview.
The type of data depends on:
- The research purpose.
- The questions asked.
- The available response format.
- The interviewer’s freedom to probe.
- The method used to analyse the answers.
One-to-One Interview Versus In-Depth Interview
A one-to-one interview describes how many people participate in the interaction, whereas an in-depth interview describes the depth and purpose of the inquiry.
| Feature | One-to-one interview | In-depth interview |
|---|---|---|
| Basic meaning | One interviewer speaks with one participant | The interviewer explores a participant’s experiences or meanings in substantial detail |
| Defining feature | Individual format | Depth, probing, and detailed exploration |
| Possible structure | Structured, semi-structured, or unstructured | Usually semi-structured or relatively unstructured |
| Possible data | Quantitative, qualitative, or mixed | Primarily qualitative |
| Typical purpose | Description, classification, evaluation, explanation, or exploration | Understanding experiences, motivations, interpretations, and processes |
| Relationship | May be brief and standardised | Usually requires rapport and extended probing |
Many qualitative in-depth interviews are conducted one-to-one. However, the terms should not be treated as exact synonyms.
For example, a researcher who asks every household member 20 fixed yes-or-no questions is conducting one-to-one interviews, but not necessarily in-depth interviews. A researcher who explores how a patient adapted to a chronic illness through open questions and detailed probes is conducting a one-to-one in-depth interview.
Main Characteristics of a One-to-One Research Interview
A well-designed individual research interview usually has the following characteristics:
A defined research purpose
The interview is conducted to answer a research question rather than simply to have an interesting conversation.
Direct interaction
The researcher can clarify a question, request an example, or explore an unexpected issue as it arises, provided the interview design allows this.
Individual attention
The participant does not need to compete with others for speaking time and is not directly influenced by a group discussion.
Verbal data
Most of the data consist of spoken answers. Depending on the study, researchers may also consider pauses, tone, gestures, drawings, timelines, photographs, demonstrations, or screen activity.
An interactional context
The resulting account is not produced by the participant alone. It is shaped by the questions, setting, researcher-participant relationship, social expectations, and the participant’s interpretation of the research situation.
Planned documentation
The conversation is normally documented through audio recording, video recording, field notes, a structured response form, or a combination of these methods.
Types of One-to-One Interviews
One-to-one interviews can be classified by their degree of structure, purpose, and communication mode.
Types Based on Structure
Structured interview
A structured interview uses predetermined questions in a fixed order. Wording and response options are usually kept consistent across participants.
This format is appropriate when the researcher needs:
- Strong comparability.
- Standardised administration.
- Limited interviewer discretion.
- Responses that can be coded numerically.
- A larger number of relatively brief interviews.
Its main limitation is that unexpected explanations may not be explored adequately.
Semi-structured interview
A semi-structured interview uses an interview guide containing planned topics and open-ended questions, while allowing the interviewer to ask probes, request examples, clarify meaning, and adjust the order when appropriate.
This is often the most practical format for research that needs both:
- Consistency across participants.
- Flexibility to explore individual experiences.
A systematic review by Kallio et al. (2016) emphasised that a semi-structured guide should be developed through a reasoned process rather than assembled as an informal list of questions.
Unstructured interview
An unstructured interview is guided by a general purpose or topic rather than a detailed schedule. Questions emerge through the conversation.
It is useful when:
- Little is known about the topic.
- Participants’ own categories and narratives are central.
- Prematurely imposing a detailed structure might restrict discovery.
“Unstructured” does not mean unplanned. The researcher still needs a clear purpose, ethical protocol, sampling rationale, and analytical strategy.
Types Based on Purpose
In-depth interview
Explores experiences, meanings, motivations, beliefs, or decision-making in detail.
Key-informant interview
Collects specialised knowledge from someone with a relevant position, role, or expertise—for example, a school principal, community organiser, clinician, policymaker, or industry specialist.
Narrative interview
Encourages the participant to construct an extended story, often about a life event, transition, identity, or sequence of experiences.
Critical-incident interview
Asks the participant to describe a particular event or situation that was especially effective, difficult, consequential, or revealing.
Cognitive interview
Investigates how a person understands, retrieves information for, judges, and answers a survey or assessment question. It is often used to test questionnaires.
Elite interview
Involves participants who hold influential, senior, or specialised positions. It may require extensive background preparation and careful management of power relations.
Phenomenological interview
Examines how participants experience and make sense of a phenomenon. The form of questioning should be aligned with the chosen phenomenological tradition.
Types Based on Delivery Mode
Face-to-face interview
The interviewer and participant occupy the same physical setting. This can support rapport and allow observation of visible contextual or nonverbal information.
However, physical presence can also intensify power differences, social pressure, accessibility barriers, or safety concerns.
Video interview
Video platforms allow geographically dispersed participants to take part and preserve some visual interaction. Remote interviewing can be convenient and effective, but researchers must consider connectivity, privacy, platform security, digital literacy, and the participant’s physical surroundings (Archibald et al., 2019).
Telephone interview
Telephone interviews are accessible for participants who lack stable video connections or prefer not to appear on camera. Visual cues are unavailable, but the reduced visibility may make some participants more comfortable.
Text-based synchronous interview
Real-time text chat can be useful for participants who communicate more comfortably in writing or require certain accessibility accommodations. The researcher must recognise that typed interaction produces a different form of data from speech.
When Should One-to-One Interviews Be Used?
One-to-one interviews are appropriate when the researcher needs a detailed individual account and direct group interaction is not necessary to answer the research question.
They are particularly useful for:
- Exploring personal experiences or interpretations.
- Discussing potentially sensitive or private topics.
- Interviewing participants who may be uncomfortable speaking in a group.
- Reaching geographically dispersed participants.
- Collecting specialised knowledge from experts.
- Understanding why people made particular decisions.
- Following up findings from surveys, observations, or documents.
- Examining differences among individual cases.
- Giving participants enough time to explain complex experiences.
Example
A researcher studying why first-generation doctoral students consider leaving their programmes may use semi-structured one-to-one interviews. The method allows each participant to discuss supervision, finances, belonging, family expectations, and institutional support without having to disclose these experiences in front of peers.
When Is a One-to-One Interview Not the Best Method?
The method may be unsuitable when:
- The research question concerns directly observable behaviour rather than reported experience.
- Group interaction, consensus, disagreement, or shared language is the phenomenon of interest.
- The study requires population estimates or statistical generalisation.
- Participants can answer the research question adequately through a brief questionnaire.
- The researcher cannot provide appropriate privacy or data protection.
- The topic creates risks that the research team is not trained or equipped to manage.
- Interviewing would place an unnecessary burden on participants.
- The study requires real-time observation of how people use a service, device, or environment.
In these situations, observation, a survey, document analysis, a focus group, diary methods, or a mixed-methods design may be more appropriate.
One-to-One Interview Versus Other Methods
| Method | Best suited to | Main strength | Main limitation |
|---|---|---|---|
| One-to-one interview | Individual experiences, interpretations, decisions, and expert knowledge | Depth and opportunity to probe | Resource-intensive and affected by interviewer-participant interaction |
| Focus group | Group norms, shared language, disagreement, and interaction | Multiple perspectives and interaction in one session | Confidentiality cannot be guaranteed and some participants may dominate |
| Survey | Measuring distributions, associations, or standardised responses | Broad coverage and efficient comparison | Limited depth and little opportunity for clarification |
| Observation | Behaviour, practice, context, and interaction as they occur | Captures what people do rather than only what they report | Motives and private experiences may remain unclear |
| Diary method | Experiences recorded over time | Reduces reliance on a single retrospective account | Requires sustained participant effort |
| Document analysis | Policies, records, media, reports, or historical material | Nonreactive access to existing material | Documents may be incomplete or produced for another purpose |
Researchers can combine methods when each contributes different evidence. For example, interviews may explain why a pattern appeared in survey results or why observed practice differs from an official policy.
How to Conduct a One-to-One Interview
A rigorous interview study can be organised into 10 steps.
Step 1: Define the research question
Clarify exactly what the interviews need to explain or explore.
A weak purpose is:
To learn about students.
A stronger purpose is:
To explore how international master’s students interpret and respond to academic feedback during their first semester.
The stronger version identifies the population, phenomenon, and context.
Step 2: Choose the methodological approach
Decide whether the study is descriptive, phenomenological, narrative, grounded-theory-oriented, case-based, realist, constructivist, or guided by another methodology.
This decision affects:
- What counts as relevant knowledge.
- How questions are phrased.
- The interviewer’s role.
- How detailed the transcript must be.
- Which analytical method is appropriate.
- What claims may be made from the findings.
Step 3: Select the interview type
Choose structured, semi-structured, or unstructured interviewing.
Semi-structured interviewing is often suitable when researchers need to address the same broad areas with every participant while preserving space for individual explanations.
Step 4: Develop a sampling strategy
Identify who can provide information relevant to the research question.
Qualitative interview studies commonly use purposive sampling. Possible strategies include:
- Criterion sampling.
- Maximum-variation sampling.
- Homogeneous sampling.
- Typical-case sampling.
- Critical-case sampling.
- Expert sampling.
- Snowball or chain-referral sampling.
Convenience alone is usually a weak justification unless access limitations are acknowledged and the sample remains appropriate to the study aim.
Step 5: Develop the interview guide
Map every major interview topic to the research question.
A strong guide usually moves from:
- Introduction and rapport-building.
- Factual or contextual questions.
- Broad experience questions.
- Focused follow-up questions.
- Sensitive or demanding topics.
- Reflection and recommendations.
- Final opportunity to add information.
Avoid including questions merely because they sound interesting. Each should have a clear purpose.
Step 6: Complete ethical and logistical planning
Before recruitment, determine:
- Whether ethics review or institutional approval is required.
- How informed consent will be documented.
- Whether recording is necessary.
- What will happen if a participant becomes distressed.
- How privacy will be protected.
- Where interviews will take place.
- How files will be named, transferred, stored, retained, and destroyed.
- Whether interpreters, transcription providers, or AI services will access the data.
- How withdrawal requests will be handled.
- Whether compensation is appropriate.
Step 7: Pilot the protocol
Conduct at least one realistic trial with someone who resembles the intended participant group when feasible.
A pilot can reveal:
- Confusing language.
- Leading or double-barrelled questions.
- Missing topics.
- Poor question order.
- Excessive interview length.
- Weak probes.
- Recording problems.
- Questions that cause avoidable discomfort.
Pilot data should not automatically be included in the main analysis. The protocol and consent arrangements should state whether inclusion is possible.
Step 8: Recruit and schedule participants
Provide clear information about:
- The study’s purpose.
- Why the person has been invited.
- What participation involves.
- Expected duration.
- Recording.
- Potential risks or benefits.
- Confidentiality limits.
- Compensation.
- Voluntary participation.
- Contact and withdrawal procedures.
Avoid pressure, particularly when the researcher holds authority over potential participants.
Step 9: Conduct and document the interview
At the beginning:
- Confirm consent.
- Request recording permission separately when required.
- Explain confidentiality and its limits.
- Remind the participant that they may decline a question or stop.
- Establish the expected duration.
- Check that the setting is sufficiently private.
During the conversation:
- Ask one clear question at a time.
- Listen without preparing the next question too early.
- Allow pauses.
- Request concrete examples.
- Use neutral probes.
- Avoid correcting, debating, diagnosing, or evaluating the participant.
- Monitor time without rushing important material.
- Observe and document relevant contextual information.
At the end:
- Ask whether anything important has been missed.
- Explain the next steps.
- Clarify whether the participant will receive a summary or follow-up.
- Thank the participant.
- Check the recording immediately.
Step 10: Transcribe, analyse, and report the data
Create a transparent path from the original recording to the published interpretation.
This includes:
- Preparing or checking transcripts.
- Removing or replacing identifiers where appropriate.
- Writing post-interview and reflexive memos.
- Applying the chosen analytical method.
- Comparing cases and examining contradictory evidence.
- Documenting decisions.
- Selecting quotations responsibly.
- Reporting the researcher’s role and study limitations.
One-to-One Interview Guide Template
The following template can be adapted to a dissertation, evaluation, user-research project, or qualitative study.
Administrative details
- Study title:
- Participant code:
- Interviewer:
- Date and time:
- Location or platform:
- Consent confirmed:
- Recording permission confirmed:
- Expected duration:
- Relevant contextual notes:
Opening script
Thank you for taking part. This study examines [brief purpose]. I am interested in your own experiences and views, and there are no right or wrong answers. You may decline any question or stop the interview at any time. With your permission, I will record the conversation so that I can represent your answers accurately. Do you have any questions before we begin?
Introductory questions
- Could you briefly describe your role or experience in relation to [topic]?
- How long have you been involved with [setting, activity, service, or experience]?
- What does [central concept] mean to you?
Main experience questions
- Can you tell me about your experience of [phenomenon]?
- Could you walk me through what happened?
- What aspects were most important to you?
- How did you respond at the time?
- What influenced that decision?
- How, if at all, did the experience change over time?
- What made the situation easier or more difficult?
Interpretation questions
- How do you understand that experience now?
- Why do you think it happened in that way?
- How did other people or institutions affect the experience?
- Was your experience different from what you expected?
Recommendation questions
- What should an organisation, practitioner, or policymaker understand about this issue?
- What changes would improve the situation?
- What advice would you give someone in a similar position?
Closing questions
- Is there anything important that I have not asked about?
- Is there anything you would like to clarify or correct?
- Are there any questions you would like to ask me?
Types of Interview Questions and Probes
Open-ended question
Encourages explanation in the participant’s own words.
How did you experience the transition to online learning?
Descriptive probe
Requests detail.
What did a typical day look like?
Elaboration probe
Invites the participant to continue.
Could you tell me more about that?
Example probe
Moves from an abstract claim to a concrete event.
Can you give me an example?
Clarification probe
Checks the intended meaning.
When you say the process was “unfair,” what does unfair mean in this situation?
Contrast probe
Explores differences.
How was that experience different from your previous workplace?
Timeline probe
Examines sequence and change.
What happened immediately after that?
Perspective probe
Explores how the participant interprets another viewpoint without assuming it is correct.
How do you think your supervisor understood the situation?
Reflection probe
Invites evaluation or meaning-making.
Looking back, has your understanding changed?
Silent probe
A short, attentive pause can allow a participant to think and continue. Silence should not be used manipulatively or prolonged in a way that causes unnecessary discomfort.
Questions to Avoid
Leading questions
Weak:
Don’t you think the new policy was confusing?
Better:
How, if at all, did the new policy affect your understanding of the process?
Double-barrelled questions
Weak:
How effective and fair was the assessment?
Better:
How effective was the assessment?
How fair or unfair did it seem to you?
Assumptive questions
Weak:
Why did the service fail to support you?
Better:
How would you describe the support you received?
Excessively abstract questions
Weak:
What is your epistemological relationship with institutional authority?
Better:
Can you describe a time when an institutional rule affected a decision you made?
Questions that request unnecessary sensitive data
Do not ask for identifiable, traumatic, illegal, medical, financial, or otherwise sensitive information unless it is necessary to the research question and covered by an appropriate ethical protocol.
How Many One-to-One Interviews Are Enough?
There is no universal number of interviews that is correct for every study. Sample adequacy depends on the study’s aim, sample specificity, theoretical basis, quality of the interviews, variation in the phenomenon, and analytical strategy.
Malterud et al. (2016) proposed information power as a way to reason about qualitative sample size. A sample may require fewer participants when:
- The research aim is narrow.
- Participants are highly specific to the aim.
- Relevant theory is well established and meaningfully applied.
- Interview dialogue is rich and focused.
- The analysis examines cases intensively.
A larger or more varied sample may be needed when:
- The aim is broad.
- The sample contains several substantially different groups.
- Experiences are highly diverse.
- The interviews produce limited detail.
- The analysis depends on comparing subgroups.
- The study seeks to develop a wide-ranging explanatory model.
Saturation
Saturation is often used to justify stopping data collection, but researchers should explain what they mean by it.
Hennink et al. (2017) distinguished:
- Code saturation: Few or no new topical codes are being identified.
- Meaning saturation: Existing concepts are sufficiently understood in their variation, context, and nuance.
A study can reach code saturation before it has developed adequate meaning. Researchers should not declare saturation merely because later interviews sound familiar.
A defensible sample-size statement
Participants were recruited purposively and data collection continued while the research team assessed the information power of the sample, variation across cases, and the depth of the developing analysis. Recruitment stopped when additional interviews no longer materially extended the interpretation of the principal concepts.
The precise wording should match what the research team actually did.
How Long Should a One-to-One Interview Last?
There is no mandatory duration. Many semi-structured research interviews are planned for approximately 30–90 minutes, but appropriate length depends on:
- The number and complexity of topics.
- Participant age, health, communication needs, and fatigue.
- Topic sensitivity.
- Whether visual or task-based activities are included.
- The study methodology.
- The participant’s available time.
- Whether follow-up interviews are planned.
A focused 25-minute interview can produce better data than a repetitive 90-minute interview. Researchers should prioritise relevance and participant welfare over duration.
Ethical Considerations
Ethical interviewing requires more than obtaining initial consent. Researchers remain responsible for participant welfare and data protection throughout recruitment, collection, analysis, reporting, archiving, and disposal.
Informed consent
Participants should understand:
- The study’s purpose.
- What they will be asked to do.
- Whether recording will occur.
- Foreseeable risks or discomforts.
- How the data will be used.
- Who may access the data.
- Whether quotations may be published.
- The limits of confidentiality.
- Their right to decline questions or withdraw, subject to clearly explained limits.
Consent should be treated as an ongoing process, especially in long, sensitive, or multi-stage studies.
Privacy and confidentiality
Conduct interviews where other people cannot overhear unless the participant knowingly chooses otherwise.
Researchers should distinguish:
- Privacy: Protection of the interview setting and personal space.
- Confidentiality: Restrictions on access to identifiable information.
- Anonymisation: Removal or alteration of identifying information so that a person cannot reasonably be identified.
- Pseudonymisation: Replacement of direct identifiers while retaining a separate re-identification key.
True anonymity may be impossible when the research team knows who participated or when a participant has a unique role. Avoid promising anonymity when only confidentiality or pseudonymisation is feasible.
Deductive disclosure
A quotation may identify someone even when their name is removed. A distinctive job title, location, event, demographic combination, or life history may allow readers to infer the person’s identity.
Researchers may need to:
- Generalise nonessential details.
- Combine contextual descriptions.
- Shorten or paraphrase a quotation cautiously.
- Seek additional permission.
- Omit a highly identifying passage.
Any alteration must preserve the intended meaning and be disclosed appropriately.
Sensitive topics and distress
Prepare a proportionate protocol describing:
- Signs that the interview should pause.
- How the participant can skip a question.
- When the interview should stop.
- What support information can be provided.
- What happens if serious risk or legally reportable information is disclosed.
- How the interviewer will access supervision or debriefing.
Researchers should not imply that participation provides therapy unless the activity is genuinely a clinical intervention.
Power relationships
Special care is needed when the interviewer is also the participant’s:
- Teacher.
- Supervisor.
- Employer.
- Clinician.
- Service provider.
- Immigration adviser.
- Community leader.
Recruitment and consent procedures should reduce the possibility that participation affects grades, employment, treatment, access to services, or social standing.
Recording and data security
Before recording:
- Obtain explicit permission.
- Explain whether audio, video, chat, or automated transcription will be retained.
- Confirm the approved recording device or platform.
- Avoid unnecessary collection of identifiable visual information.
- Test secure transfer and storage.
The research team should know where cloud recordings are stored, whether platform providers process the content, and whether data cross national borders.
Conducting Online One-to-One Interviews
Online interviewing can expand access, but it should not be treated as an identical substitute for in-person research.
Advantages
- Reduced travel.
- Wider geographical access.
- Easier scheduling.
- Inclusion of participants who prefer their own environment.
- Screen sharing or digital stimulus presentation.
- Built-in recording or transcription options, subject to approval.
Limitations
- Unstable internet or device access.
- Reduced visibility of body language and surroundings.
- Platform fatigue.
- Digital exclusion.
- Difficulty establishing a private setting.
- Risk of unannounced people entering the room.
- Dependence on a third-party platform.
- Security and data-location concerns.
Online-interview checklist
Before beginning:
- Confirm the participant is in a sufficiently private place.
- Ask whether headphones would improve privacy.
- Agree on what to do if the connection fails.
- Provide a telephone or alternative contact method.
- Disable unnecessary platform features.
- Confirm recording consent after the connection begins.
- Explain any automated transcription notice.
- Avoid displaying confidential participant information on screen.
- Save the recording to the approved location.
- Remove platform-generated files that are not authorised for retention.
Transcribing One-to-One Interviews
Transcription converts a recorded interaction into a written representation. It is not a purely mechanical step because every transcript reflects decisions about what to include.
Possible levels include:
Intelligent or cleaned verbatim
Removes repetitions, false starts, and filler words while preserving substantive meaning. It may be suitable for some applied thematic or content analyses.
Full verbatim
Includes the words spoken, repetitions, false starts, and relevant vocal features. It supports closer attention to how accounts are expressed.
Detailed interactional transcription
Records pauses, overlap, intonation, timing, emphasis, laughter, and other features. It may be necessary for conversation analysis or discourse-oriented work.
The transcription convention should match the analytical method. A researcher cannot conduct detailed interaction analysis from a heavily cleaned summary transcript.
Transcript checking
Automated or human-generated transcripts should be checked against the recording. Particular attention is needed for:
- Names and specialised terminology.
- Accents and dialects.
- Code-switching.
- Low-volume speech.
- Overlapping speech.
- Emotional or hesitant passages.
- Numbers and dates.
- Negation, where a missing word can reverse meaning.
How Are One-to-One Interviews Analysed?
Interview analysis should be selected according to the research question, methodology, and type of account collected.
| Analysis method | Main purpose | Suitable interview data |
|---|---|---|
| Thematic analysis | Develop patterns of shared meaning across a dataset | Broadly applicable semi-structured or open interviews |
| Qualitative content analysis | Organise and interpret content using inductive or deductive categories | Descriptive, applied, policy, health, or communication studies |
| Framework analysis | Compare cases and topics systematically through a matrix | Applied research with clear questions and stakeholder needs |
| Interpretative phenomenological analysis | Examine how individuals make sense of significant experiences | Detailed, relatively homogeneous experiential samples |
| Grounded-theory analysis | Develop an explanatory theory through iterative collection and analysis | Process-focused interviews combined with theoretical sampling |
| Narrative analysis | Examine how people construct and organise stories | Life histories, transitions, identity, and event narratives |
| Discourse analysis | Study how language constructs meanings, identities, or social realities | Interviews treated as situated language rather than transparent reports |
| Conversation analysis | Examine turn-taking and interactional organisation in detail | High-quality recordings and detailed transcripts |
| Template analysis | Apply and revise a hierarchical coding template | Studies needing structured but adaptable coding |
Basic thematic-analysis workflow
A practical thematic workflow may include:
- Read and become familiar with each transcript.
- Record initial observations and reflexive notes.
- Code relevant features of the data.
- Compare codes across interviews.
- Develop candidate themes.
- Examine whether themes are supported by the dataset.
- Refine boundaries and relationships among themes.
- Identify contradictory or negative cases.
- Select quotations that demonstrate the interpretation.
- Write an analytical account rather than a list of quotations.
Coding software can help organise data, but it does not decide what the data mean.
Ensuring Quality and Trustworthiness
Quality is produced through coherent design and transparent reasoning, not by adding a checklist after the analysis.
Methodological coherence
The research question, sampling, interview design, transcription, analysis, and claims should fit together.
Reflexivity
Researchers should examine how their identities, assumptions, experiences, relationships, and institutional positions influence:
- Recruitment.
- Rapport.
- Questions and probes.
- Interpretation.
- Attention to particular accounts.
- Decisions about what to publish.
A reflexive memo can be written immediately after each interview.
Audit trail
Maintain records of:
- Interview-guide revisions.
- Recruitment decisions.
- Sampling changes.
- Field notes.
- Coding versions.
- Theme development.
- Meetings and disagreements.
- Deviant-case analysis.
- AI prompts and outputs, when applicable.
- Reasons for excluding or altering data.
Negative-case analysis
Actively examine cases that challenge the developing explanation. Contradictory evidence should be analysed, not hidden.
Triangulation
Researchers may compare evidence across:
- Participants.
- Researchers.
- Methods.
- Data sources.
- Time periods.
- Theoretical perspectives.
Triangulation should have a clear purpose. Different sources do not necessarily need to produce identical findings.
Participant feedback
Participants may be invited to comment on a transcript, summary, interpretation, or proposed recommendation. This can be useful, but it is not a universal validation test. Participants may change their views, disagree with an interpretation, or feel uncomfortable reading spoken language in transcript form.
Reporting standards
The Consolidated Criteria for Reporting Qualitative Research, or COREQ, provides a 32-item checklist designed for interviews and focus groups (Tong et al., 2007). The Standards for Reporting Qualitative Research, or SRQR, provides 21 broader reporting standards (O’Brien et al., 2014).
Researchers should also follow the requirements of their discipline, institution, funder, ethics committee, and target journal.
Advantages of One-to-One Interviews
Detailed individual accounts
The participant has time to explain experiences, reasoning, and context without competing with a group.
Flexible clarification
In semi-structured and unstructured formats, the interviewer can ask what the participant means and request examples.
Greater conversational privacy
Some participants may discuss sensitive experiences more freely without peers, colleagues, or family members present.
Access to specialised knowledge
Experts and key informants can explain processes that are poorly documented or difficult to observe.
Adaptability
Interviews can be conducted in person, remotely, through telephone, or with accessibility adjustments.
Connection with other methods
Interviews can explain survey patterns, contextualise observations, assist questionnaire development, or evaluate a programme.
Attention to minority or unusual cases
Individual interviewing can create space for experiences that might be suppressed or outnumbered in a group setting.
Limitations of One-to-One Interviews
Time and resource requirements
Recruitment, scheduling, interviewing, transcription, checking, coding, and interpretation require substantial labour.
Interviewer influence
Question wording, tone, identity, reactions, assumptions, and follow-up choices may affect the participant’s account.
Social-desirability and recall problems
Participants may provide answers they consider acceptable, forget details, rationalise earlier decisions, or reconstruct events through their present perspective.
Limited observation of actual behaviour
Interviews reveal what people say in the research context. Reported practice may differ from behaviour in everyday settings.
Restricted statistical generalisation
Small purposive samples are generally not designed to estimate population percentages.
Inconsistent depth
Some participants are highly articulate while others provide brief or guarded answers. Differences in interview quality can complicate comparison.
Confidentiality risks
Rich contextual detail can make participants identifiable even after names are removed.
Emotional burden
Sensitive interviews may affect participants and researchers. Appropriate training, boundaries, supervision, and support procedures may be required.
Dependence on researcher skill
Poor listening, leading questions, excessive talking, or failure to probe can reduce the usefulness of the data.
Common Mistakes
Treating the interview guide as a questionnaire
Reading questions mechanically can prevent active listening and meaningful follow-up.
Asking questions unrelated to the research aim
Every topic increases participant burden and analytical work.
Beginning with the most sensitive issue
A gradual transition usually allows the participant to understand the interview process and decide how much to disclose.
Interrupting useful silence
Researchers may answer their own question or move on before the participant has finished thinking.
Asking several questions at once
The participant may answer only the final or easiest part.
Expressing approval or disapproval
Overt judgment can influence later answers.
Assuming rapport eliminates bias
A warm interaction may support communication, but it does not make the interview neutral or remove power differences.
Choosing a sample number without justification
“Other studies interviewed 15 people” is not a sufficient methodological rationale.
Coding before clarifying the analytical approach
Codes should be developed and used in a way that is consistent with the study’s conceptual and methodological assumptions.
Reporting quotations without analysis
A collection of quotations is not a finding. The researcher must explain the pattern, context, variation, and significance.
Uploading confidential data to unapproved software
Convenience does not override consent, ethics approval, contractual restrictions, or data-protection obligations.
Digital Research Tools
Digital tools can assist with organisation, but researchers remain responsible for methodological decisions.
Scheduling and communication
Calendar systems, secure email, research-management platforms, and approved video-conferencing services can support recruitment and scheduling.
Recording
Use an institutionally approved recorder or platform. A backup may be appropriate when it does not create additional privacy or consent problems.
Transcription
Automated speech recognition can produce an initial transcript. The researcher should check it against the recording and document how errors were corrected.
Qualitative-data analysis
Programs such as NVivo, ATLAS.ti, MAXQDA, Dedoose, and Taguette can support:
- Transcript organisation.
- Coding.
- Retrieval.
- Memo writing.
- Case comparison.
- Codebook management.
- Visual displays.
- Team collaboration.
Software does not establish validity and should not determine the methodology.
Artificial Intelligence in One-to-One Interview Research
AI may support interview preparation, transcription, translation, data organisation, code suggestions, and comparison of candidate themes. However, AI output should be treated as an auditable research aid rather than an independent analyst.
Recent methodological literature identifies both opportunities and risks in AI-assisted qualitative analysis (Jones, 2025; Nicmanis & Spurrier, 2025).
Potential uses
AI may help researchers:
- Identify duplicated or unclear interview questions.
- Generate possible neutral probes for expert review.
- Produce a first-pass transcript.
- Locate passages associated with a predefined codebook.
- Summarise a transcript for navigation.
- Compare human-developed themes with alternative groupings.
- Check whether draft reporting omits required methodological details.
Principal risks
- Disclosure of confidential or identifiable data.
- Inaccurate transcription.
- Fabricated interpretations or quotations.
- Loss of cultural, emotional, and linguistic nuance.
- Overconfident summaries.
- Unclear training-data or retention practices.
- Inconsistent outputs across model versions.
- Poor fit with the study’s epistemology.
- Automation bias, in which researchers accept plausible output without sufficient examination.
Responsible AI workflow
- Verify that AI use is permitted by the ethics approval, consent materials, data-management plan, institution, and applicable law.
- Prefer approved, contractually protected, or locally hosted tools for sensitive data.
- Remove unnecessary identifiers before processing.
- Define the AI’s limited role in advance.
- Keep prompts, settings, model information, and output versions.
- Compare outputs with the original recording or transcript.
- Require a human researcher to make final coding and interpretive decisions.
- Search actively for omissions and contradictory evidence.
- Do not use an AI-generated quotation unless it is verified word-for-word against the source.
- Disclose material AI use in the methods or transparency statement.
AI should not be represented as a neutral replacement for researcher interpretation.
Examples of One-to-One Interviews in Research
Education example
Research question: How do novice teachers experience their first year in rural schools?
Researchers could conduct semi-structured interviews at the beginning, middle, and end of the academic year. Questions may address professional identity, workload, community relationships, mentoring, and intentions to remain.
Healthcare example
Research question: How do patients understand changes to a long-term medication regimen?
Individual interviews may be preferable to focus groups because participants can discuss symptoms, uncertainty, cost, family influence, and communication with clinicians privately.
User-research example
Research question: Why do first-time users abandon an online application process?
The interviewer might combine a one-to-one interview with screen sharing or a usability task. The interview explores expectations and reasoning, while observation records actual interaction with the system.
Organisational example
Research question: How do remote employees experience informal access to managers?
Individual interviews may reveal differences associated with role, seniority, time zone, caring responsibilities, communication norms, and perceived psychological safety.
Policy-evaluation example
Research question: How do local service providers implement a newly introduced funding requirement?
Key-informant interviews can examine interpretation of the policy, resource constraints, unintended consequences, and local adaptation.
Example Methodology Description
The following model should be adapted rather than copied unchanged:
Semi-structured one-to-one interviews were used to explore participants’ experiences of academic feedback during their first semester. Participants were selected purposively because they had direct experience of the phenomenon under investigation. The interview guide was developed from the research questions and relevant literature, reviewed by two researchers, and piloted before data collection. Interviews were conducted through an institutionally approved video platform, with informed consent and separate permission for audio recording. Recordings were transcribed, checked against the audio, pseudonymised, and analysed using thematic analysis. Reflexive memos and an audit trail were maintained throughout data collection and analysis.
A complete methods section should also explain:
- The methodological orientation.
- Recruitment and inclusion criteria.
- Sample-size rationale.
- Interviewer characteristics relevant to the interaction.
- Interview setting and duration.
- Guide development.
- Recording and transcription conventions.
- Coding and theme-development process.
- Quality procedures.
- Ethical approval.
- Data protection.
- Material use of AI or external services.
Conclusion
A one-to-one interview is an individual data-collection format that can support structured measurement or detailed qualitative inquiry. Its value lies in focused interaction, clarification, and access to individual experience, but credible findings require more than asking good questions. The research question, sampling, interview design, ethics, transcription, analysis, reflexivity, and reporting must form a coherent methodological process.
Leave a Comment