Customer Discovery and Empathic Research

Introduction

Product management begins with an uncertainty: whether a meaningful customer problem exists, for whom it matters, and under what circumstances it occurs. Customer discovery is the disciplined process of reducing that uncertainty through direct engagement with people, observation of their contexts, and interpretation of the evidence generated by their behaviour and accounts. Empathic research is a related orientation that seeks to understand customers as situated human beings rather than merely as sources of requirements or market statistics.

The purpose of empathic research is not to make researchers feel what participants feel. Rather, it is to develop a sufficiently grounded understanding of participants’ goals, constraints, practices, emotions, social relationships and meanings to make better product decisions. This distinction matters because customers may be unable to articulate latent needs directly. They may describe an idealised process, omit routine workarounds, or adapt their behaviour to organisational norms. Effective discovery therefore combines what people say, what they do, what they use, and what their environment makes possible or difficult.

For product managers, qualitative research is particularly valuable in the early stages of opportunity identification, problem framing, segmentation, service design and feature prioritisation. It is also useful after launch when teams need to explain adoption barriers, failure points or unexpected uses. Qualitative findings do not automatically represent the prevalence of a problem. Their principal contribution is explanatory depth: they help teams understand how and why a problem occurs and what a credible response might look like.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Explain the purpose and limits of customer discovery and empathic research.
2. Select among interviews, observation, contextual inquiry, diary studies and immersion.
3. Develop a research question, discussion guide and ethical consent process.
4. Construct a defensible qualitative sampling strategy.
5. Identify explicit, latent and unarticulated needs without over-interpreting evidence.
6. Apply triangulation, reflexivity and bias controls to improve credibility.
7. Translate qualitative findings into product opportunities while preserving the distinction between evidence and interpretation.
8. Manage privacy, safety, accessibility and data-governance obligations in research.

1. Customer Discovery as a Product Management Capability

A product hypothesis normally contains several linked assumptions: a target user experiences a problem, the problem has sufficient importance, existing alternatives are inadequate, a proposed intervention is feasible, and users or buyers can adopt it. Discovery research primarily investigates the first two assumptions and helps clarify the third. It should not be confused with asking customers to design the product or with seeking approval for a solution already chosen.

A useful discovery question is specific enough to guide inquiry but open enough to permit surprise. For example, “How do small retailers reconcile online and offline inventory during a normal trading week?” is more productive than “Would you use our inventory dashboard?” The former investigates current practice; the latter invites speculation and social desirability bias.

Explicit, latent and unarticulated needs

An explicit need is directly stated, such as “I need to export this report every Friday.” A latent need is not usually stated as a requirement but can be inferred from recurring behaviour, workarounds or emotional responses. For example, repeated manual checking may indicate a need for confidence or control, not merely a need for automation. An unarticulated need may be difficult for participants to express because the practice is habitual, socially sensitive or outside their awareness.

These categories are analytical, not psychological diagnoses. A researcher should avoid claiming certainty about an unspoken need. The appropriate language is provisional: “The evidence suggests that reducing uncertainty may matter more than reducing the number of steps.” Such claims should be tested through further research or product experiments.

Research questions and decision alignment

A research plan should begin with a product decision. Examples include:

• Should the team pursue this problem space?
• Which user situations create the greatest difficulty?
• What alternatives do people currently use?
• Which trust, privacy or workflow concerns could prevent adoption?
• What should be explored in a prototype test?

The research question should specify the population, situation and phenomenon of interest. A broad objective such as “understand customer pain points” is insufficient. A stronger formulation is: “Understand how first-time parents select overnight childcare information, what makes sources credible, and where decisions become difficult.”

2. Theoretical Foundations

Interpretivism and situated meaning

Many empathic studies are informed by an interpretivist view: reality is experienced and interpreted through social, cultural and organisational contexts. The same product feature may have different meanings for different users. A notification may represent convenience to one participant and surveillance to another. Consequently, researchers study meaning in context rather than assuming that behaviour has a universal explanation.

This does not mean that every interpretation is equally valid. Interpretations should be supported by transparent evidence, comparison across cases, reflexive analysis and, where appropriate, participant feedback.

Phenomenology and lived experience

A phenomenological orientation focuses on how a phenomenon is experienced. It is useful when the product concerns emotionally significant events such as chronic illness, financial stress or professional identity. The researcher explores the participant’s account of time, emotion, bodily experience and meaning rather than reducing the experience to a list of features.

Ethnographic sensibility

Ethnographic research examines practices within their social and material settings. A product manager need not conduct a full anthropological study to use an ethnographic sensibility. The practical implications are to observe routines, tools, language, informal rules, power relationships and exceptions rather than studying isolated user actions.

Jobs-to-be-Done and progress

Jobs-to-be-Done approaches frame demand around progress a person is trying to make in a circumstance. The “job” is not simply a demographic attribute or feature request. For example, a commuter may be trying to arrive predictably while preserving time for family, not merely trying to buy a transport ticket. This framing helps researchers examine triggers, desired outcomes, existing alternatives and anxieties surrounding change. It should be used as a questioning lens, not as a substitute for evidence.

The double-diamond logic

A common design process distinguishes divergent discovery from convergent definition, followed later by divergent development and convergent delivery. In discovery, the team broadens its understanding before narrowing to a problem framing. Premature convergence—selecting a solution after the first interview—is a major source of confirmation bias.

3. Choosing Qualitative Methods

No single method reveals the whole customer experience. Method selection should reflect the question, sensitivity of the topic, participant burden, access constraints and required level of contextual detail.

Semi-structured interviews

An interview guide provides consistency without turning the interaction into a questionnaire. Begin with recent, concrete episodes: “Tell me about the last time you had to resolve this issue.” Probe sequence, actors, tools, decisions, feelings and consequences. Useful probes include “What happened next?”, “How did you know?”, “What did you try instead?” and “What made that difficult?”

Avoid leading questions such as “How much would you like an automated solution?” Avoid hypothetical questions unless the study specifically concerns expectations. Instead of asking whether participants value security, ask them to describe the last time they assessed whether a service was safe.

The interviewer should listen more than speak, tolerate pauses, and distinguish a participant’s words from their own interpretation. Interviews should normally be recorded only with explicit permission. If recording is not appropriate, use two-person note-taking or detailed notes written immediately after the session.

Observation

Observation focuses on what occurs in practice. A researcher may record actions, sequence, duration, interruptions, artefacts, physical arrangement and interactions with other people. Observation should not be treated as a neutral camera: the researcher’s presence and selection of what to record shape the account.

A structured observation template can include: setting; actors; task; tools; triggers; visible workarounds; errors; interruptions; decisions; emotional or bodily cues; and questions for later clarification. Interpretation should be separated from description. “Participant repeatedly looked at a handwritten card” is observation; “participant lacks trust in the system” is an interpretation requiring further evidence.

Contextual inquiry

Contextual inquiry combines observation with brief questions during or immediately after activity. The researcher asks the participant to explain what they are doing without forcing them into an artificial demonstration. A useful principle is partnership: the participant is the expert in their work, while the researcher contributes curiosity and analytical attention.

Contextual inquiry is especially useful for business products, healthcare services, industrial tools and administrative workflows. It can reveal hidden dependencies, local terminology and exceptions that interviews miss. However, the researcher must avoid disrupting safety-critical or confidential work.

Diary studies

Diary studies capture experiences close to the time they occur. They may use paper, a secure application, voice notes or scheduled prompts. A diary protocol should define the event that triggers an entry, the expected frequency, the approximate time required and the support available to participants.

Prompts should be concise and neutral: “Describe the most recent time you encountered difficulty with X. What were you trying to accomplish, and what did you do?” Diary studies are useful for fluctuating symptoms, recurring service interactions, consumption, commuting and workplace interruptions. They require careful attention to accessibility, participant fatigue, incomplete entries and privacy risks in real-time records.

Immersion

Immersion involves sustained engagement with a customer environment, service journey or community. It can include shadowing, participating in frontline work, using an existing service repeatedly, or spending time in the physical and social setting where the problem occurs. Immersion can generate systemic insight, but it does not make the researcher a member of the community. Claims of “becoming the user” can erase differences in power, identity and lived experience.

4. Sampling for Qualitative Discovery

Qualitative sampling is usually purposive rather than statistically representative. Participants are selected because they can illuminate the research question. Relevant dimensions may include role, experience level, context of use, frequency of the problem, accessibility needs, geography, organisational size and relationship to the purchase decision.

A sampling matrix makes assumptions visible:

Researchers often begin with criterion sampling, recruit participants who meet defined conditions, and then use maximum-variation sampling to explore meaningful differences. Snowball sampling can reach hidden populations but may over-represent connected networks. Convenience sampling is fast but should be treated as a source of exploratory insight, not a basis for broad claims.

Saturation and information power

Saturation is commonly described as the point at which additional data no longer produce meaningful new themes. It is not a fixed number of interviews. Saturation depends on the specificity of the research question, homogeneity of the sample, quality of dialogue and complexity of the phenomenon. A narrow, well-defined study may reach useful saturation earlier than a broad study of diverse users.

A practical stopping decision considers: whether major participant groups have been heard; whether new interviews add properties to existing themes or entirely new themes; whether contradictory cases have been investigated; and whether the findings are sufficient for the product decision. Teams should document this reasoning rather than use “we conducted ten interviews” as a universal rule.

5. Ethics, Consent and Participant Welfare

Ethical research is not merely a compliance step. It protects participants and improves evidence by establishing trust and appropriate boundaries.

Informed consent

Consent should be voluntary, informed, specific and revocable. Participants should understand the purpose of the study, what participation involves, expected duration, recording arrangements, foreseeable risks, compensation, data use, confidentiality limits and whom to contact with concerns. Consent is an ongoing process; researchers should re-check permission when the activity changes or sensitive material emerges.

A participant’s agreement to join a study does not automatically permit the collection of unrelated personal data. Researchers should follow data minimisation: collect only what is necessary for the stated purpose. Sensitive information requires stronger safeguards, restricted access and, where applicable, review by an institutional ethics committee or equivalent governance body.

Power and vulnerability

Employees may feel unable to refuse an employer-sponsored study. Patients, children, financially dependent people, migrants and people with disabilities may face additional vulnerability. Recruitment should avoid coercive language, and compensation should be fair without becoming an undue inducement. Researchers should provide accessible materials, interpreters or support where appropriate.

In observation, consent must account for bystanders and incidental data. In workplaces, a manager’s permission does not replace individual participant consent. In digital research, screenshots, usernames, location traces and metadata can identify people even after obvious identifiers are removed.

Privacy and data governance

A responsible data plan specifies collection, storage, retention, access, transcription, analysis, deletion and reporting. Use pseudonyms in working files and remove indirect identifiers where feasible. Secure recordings and transcripts; do not place confidential data into unapproved generative AI tools. If automated transcription is used, verify accuracy and understand whether the provider retains or trains on submitted content.

6. Conducting and Analysing the Study

A disciplined study typically proceeds through the following sequence:

1. Define the product decision and research questions.
2. Review existing evidence, including support logs, analytics and prior studies.
3. Choose methods and define inclusion and exclusion criteria.
4. Prepare recruitment materials, consent documents and a discussion or observation guide.
5. Pilot the protocol with one or two participants.
6. Conduct sessions and write reflexive notes after each session.
7. Transcribe or summarise consistently, preserving relevant context.
8. Code evidence, compare cases and examine contradictions.
9. Form themes and connect them to product decisions.
10. Validate interpretations and communicate limitations.

Coding and thematic analysis

Coding assigns concise labels to meaningful segments of data. Initial codes may be descriptive—“switches between systems”—or interpretive—“seeks reassurance before committing.” Researchers then compare codes, group related codes into candidate themes, and test whether the themes are coherent across the dataset.

A theme is not simply a frequently repeated topic. It is a patterned, meaningful idea relevant to the research question. A less frequent experience may be strategically important if it reveals a severe safety risk or a failure for an underserved group.

A useful evidence record contains:

The final product artefact should distinguish raw evidence, interpretation, opportunity and solution. “Customers need a dashboard” is usually a solution claim. “Coordinators cannot determine whether records are current without manual cross-checking” is a problem statement grounded in evidence.

7. Triangulation and Bias Control

Triangulation increases credibility by comparing different sources, methods, researchers or theoretical perspectives. For example, interviews may suggest that users rarely encounter errors, while support tickets and observation reveal frequent workarounds. The discrepancy is not a nuisance to be averaged away; it is a prompt to investigate context, memory and meaning.

Types of triangulation include:

• Data triangulation: compare participant groups, settings or time periods.
• Method triangulation: combine interviews, observation, diaries or artefacts.
• Researcher triangulation: involve more than one analyst or reviewer.
• Theory triangulation: examine data through more than one conceptual lens.

Triangulation does not guarantee truth. Methods may share the same blind spot, and apparent contradiction may reflect genuinely different experiences.

Common biases and controls

Confirmation bias occurs when researchers notice evidence supporting an existing hypothesis. Countermeasures include documenting assumptions before fieldwork, asking disconfirming questions and actively seeking negative cases.

Leading-question bias results from wording that suggests a preferred answer. Use neutral prompts and ask about actual episodes.

Recall bias arises when participants reconstruct past events inaccurately. Prioritise recent examples, artefacts and diaries.

Social desirability bias occurs when participants present themselves favourably. Emphasise that there are no right answers and avoid conducting research in front of authority figures.

Sampling bias occurs when recruitment excludes relevant groups. Compare the sample with the intended population and state exclusions explicitly.

Interviewer effects arise from tone, identity, expertise or reactions. Use training, consistent prompts, reflexive notes and, where appropriate, multiple researchers.

Survivorship bias appears when only current or successful users are studied. Include churned, failed, inactive or rejected users when ethically and practically possible.

Interpretive overreach occurs when researchers present a plausible explanation as established fact. Use calibrated language and link claims to evidence.

Reflexivity is the systematic examination of how the researcher’s assumptions, identity, incentives and relationship with the organisation shape the study. A reflexive journal can record expectations before fieldwork, moments of surprise, discomfort, changes in interpretation and unresolved questions.

8. From Findings to Product Decisions

Research findings become useful when connected to decisions without being reduced to simplistic personas or isolated quotations. A product team can synthesise findings into a problem framing that identifies the user, situation, unmet outcome, barriers and consequences. It should also specify what remains unknown.

An opportunity statement might read: “When independent retailers update stock across multiple sales channels, they need a trustworthy way to identify which records require attention, because manual reconciliation consumes time and creates fear of overselling.” This statement is more actionable than “retailers want better inventory management.”

Prioritisation should consider severity, frequency, strategic relevance, feasibility and equity. Qualitative research can illuminate severity and context but generally cannot estimate frequency for a whole market. Quantitative research, behavioural data or experiments may be required before making scale claims.

Personas should be treated as evidence-based communication tools, not fixed identities. A persona that says “Alex is a busy millennial who loves convenience” offers little analytical value. A stronger representation describes a recurring situation, goals, constraints, coping strategies and relevant differences among users.

9. Technology and AI in Empathic Research

Digital tools can support recruitment, scheduling, transcription, coding, diary prompts and repository management. Generative AI may help summarise or suggest codes, but it should not replace researcher judgement. Automated systems can omit context, flatten ambiguity, reproduce stereotypes, invent quotations or expose confidential data.

Good governance includes approved tools, access controls, de-identification, human verification and an audit trail of automated assistance. Researchers should not treat model-generated themes as findings unless they can trace them back to original evidence. Participants should be informed if AI-mediated interaction or analysis materially affects the study, especially when it changes privacy expectations.

Remote research improves access for some participants but can exclude people with limited connectivity, private space, digital literacy or assistive technology. Method choice should therefore consider equity rather than assuming that the most convenient technology is the most inclusive.

10. Illustrative Case: Discovering the Real Problem in Small-Business Cash Flow

This is a fictional illustrative case.

Background

A product team planned a mobile application that would provide small-business owners with automated cash-flow forecasts. The initial assumption was that owners lacked financial information and therefore needed more sophisticated dashboards.

Problem

Early interviews revealed that owners frequently said they wanted “better visibility.” However, the phrase was vague. The team needed to understand when cash-flow uncertainty became consequential and how owners currently managed it.

Analysis and approach

The team conducted twelve semi-structured interviews with owners and bookkeepers, selected across retail, professional services and hospitality. It then conducted contextual inquiry with six businesses during weekly payment and invoice routines and invited eight owners to complete a two-week diary after making financial decisions.

The researchers obtained consent for recording, separated business identifiers from transcripts, and allowed participants to decline observation of sensitive transactions. They compared interview claims with observed practices and diary entries.

Interviews initially suggested that owners wanted prediction accuracy. Observation showed a different pattern: owners often had adequate accounting data but could not determine which invoices were genuinely collectible, which expenses could be delayed, and which supplier relationships would be damaged by late payment. Several maintained parallel spreadsheets because they did not trust the status labels in existing software. Diary entries showed that uncertainty was most stressful immediately before payroll and tax deadlines.

The team coded evidence under information gaps, trust, timing, relationships, workarounds and consequences. A negative case—a business with highly accurate forecasts but persistent anxiety—challenged the assumption that accuracy alone solved the problem. Further inquiry showed that the owner needed confidence that decisions were defensible to a business partner and accountant.

Outcome

The opportunity was reframed from “provide a more detailed forecast” to “help owners make and explain near-term cash decisions using transparent assumptions and confidence indicators.” The first prototype therefore displayed data sources, uncertain invoices, upcoming obligations and scenario explanations rather than only a single forecast number.

The team did not claim that all small businesses shared this need. It identified the finding as strongest for owner-managed firms with irregular receivables and limited financial support, and recommended a later quantitative study to estimate segment size.

Learning points

1. Stated demand for visibility did not reveal the underlying decision context.
2. Observation exposed workarounds and trust issues that interviews alone missed.
3. Diaries captured the timing and emotional intensity of the problem.
4. A negative case improved the problem framing.
5. The research informed a prototype direction without pretending to prove market prevalence.

Challenges and Limitations

Qualitative research is vulnerable to access limitations, participant fatigue, selective memory and researcher interpretation. Customers may not represent non-users, silent users or people who have already abandoned a category. Organisational research may be distorted by gatekeepers who select participants or restrict observation. Sensitive topics may require specialist safeguarding and cannot be explored through casual product interviews.

There is also a managerial risk of “insight theatre,” in which vivid quotations create enthusiasm without changing a decision. To prevent this, every major finding should state its evidence base, affected group, confidence level, product implication and remaining uncertainty. Research should be proportionate: not every decision requires an elaborate ethnography, but important or high-risk decisions should not rely on a few convenient conversations.

Best-Practice Checklist

Before research, define the decision, question, population, risks and stopping logic. During research, obtain meaningful consent, ask about concrete experiences, observe context, document surprises and protect participant dignity. During analysis, separate evidence from interpretation, compare cases, investigate contradictions and involve more than one perspective when feasible. After research, communicate limitations, avoid prevalence claims unsupported by the method, translate findings into testable opportunities and retain only the data that governance rules permit.

Chapter Summary

Customer discovery and empathic research help product teams understand problems in the situations where they occur. Interviews reveal narratives and meanings; observation reveals visible practice; contextual inquiry links action to explanation; diary studies capture experiences over time; and immersion develops insight into wider systems and cultures. Strong research begins with a product decision and a focused research question rather than a preferred solution.

Qualitative sampling is purposive and should include meaningful variation in roles, contexts, experience and outcomes. Saturation is a reasoned judgement, not a universal participant count. Ethical research requires voluntary and ongoing consent, data minimisation, privacy protection, accessibility and attention to power and vulnerability.

Credible interpretation depends on systematic coding, reflexivity, triangulation and active controls for confirmation, recall, social desirability, sampling and interviewer bias. Findings should distinguish evidence from interpretation and opportunity from solution. Qualitative research explains experiences and mechanisms but does not by itself establish market prevalence. Product managers should therefore combine empathic research with appropriate quantitative evidence and experiments. Technology and AI can support research operations, but human judgement, confidentiality and accountability remain essential.

Key Terms

Customer discovery: A structured process for reducing uncertainty about customer problems, contexts and alternatives.

Empathic research: Research designed to understand people’s experiences, meanings, goals and constraints in context.

Explicit need: A need or difficulty directly stated by a participant.

Latent need: An underlying need inferred from behaviour, workarounds or recurring patterns.

Unarticulated need: A relevant need that participants may not consciously express or easily describe.

Semi-structured interview: An interview guided by planned topics while allowing flexible probing.

Observation: Systematic study of behaviour, interactions and environments as they occur.

Contextual inquiry: Observation combined with questions asked in the participant’s real activity context.

Diary study: A longitudinal method in which participants record experiences close to the time they occur.

Immersion: Extended engagement with a customer’s environment, service or practice.

Purposive sampling: Selecting participants because they meet characteristics relevant to the research question.

Maximum-variation sampling: Deliberately including diverse cases to explore differences.

Saturation: A judgement that additional data are producing little meaningful new insight for the study purpose.

Informed consent: Voluntary agreement based on understandable information about participation and data use.

Triangulation: Comparing evidence across data sources, methods, researchers or theories.

Reflexivity: Examination of how the researcher’s position and assumptions shape the research.

Coding: Labelling meaningful segments of qualitative data for systematic analysis.

Theme: A patterned and meaningful idea developed from related evidence.

Negative case: Evidence that contradicts or qualifies an emerging explanation.

Interpretive overreach: Presenting a plausible interpretation as more certain than the evidence permits.

Review Questions

Short Answer Questions

1. Distinguish between an explicit need and a latent need, giving one example of each.
2. What is the principal difference between observation and contextual inquiry?
3. Why is qualitative sampling usually purposive rather than statistically representative?
4. List four elements that should be addressed in informed consent.
5. What is triangulation, and why can contradiction between methods be useful?

Long Answer Questions

1. Explain how a product manager should move from a product decision to a focused qualitative research plan.
2. Compare interviews, observation, contextual inquiry, diary studies and immersion, including the situations in which each is most appropriate.
3. Discuss saturation, maximum-variation sampling and negative-case analysis in the development of credible qualitative findings.
4. Analyse the ethical challenges of conducting contextual inquiry in a workplace or healthcare environment.
5. Explain how coding, thematic analysis, reflexivity and triangulation work together to control interpretive bias.

Analytical / Application Questions

1. A team believes that users need a faster checkout process. Design a discovery study that tests this assumption without leading participants toward a speed-focused solution.
2. Interviewees claim that they rarely make errors, but observation shows repeated manual checking and correction. Develop at least three possible explanations and propose a triangulation plan.
3. Using the fictional cash-flow case, write a revised opportunity statement and identify which claims require further quantitative validation before product investment.

Multiple Choice Questions with Answers

Question 1

Which research question is most appropriate for early customer discovery?

A. Would you buy our proposed dashboard?

B. How do you currently reconcile inventory across sales channels?

C. Which colour should the dashboard use?

D. Would you recommend our future product?

Correct Answer: B. How do you currently reconcile inventory across sales channels?

Question 2

What is a principal strength of contextual inquiry?

A. It eliminates all observer effects.

B. It measures market prevalence precisely.

C. It links observed behaviour with explanations in the real context.

D. It requires no participant consent.

Correct Answer: C. It links observed behaviour with explanations in the real context.

Question 3

Which sampling approach deliberately includes diverse cases to explore differences?

A. Maximum-variation sampling

B. Pure random sampling

C. Quota-free convenience sampling

D. Single-case sampling

Correct Answer: A. Maximum-variation sampling

Question 4

Which statement about qualitative saturation is correct?

A. It always occurs after exactly ten interviews.

B. It means every member of the market has been studied.

C. It is a reasoned judgement that further data add little meaningful new insight.

D. It proves that a problem is widespread.

Correct Answer: C. It is a reasoned judgement that further data add little meaningful new insight.

Question 5

Which practice best reduces leading-question bias?

A. Describe the preferred solution before asking questions.

B. Ask participants about a recent concrete episode.

C. Ask participants to agree with the product hypothesis.

D. Use only yes-or-no questions.

Correct Answer: B. Ask participants about a recent concrete episode.

Question 6

What does data triangulation involve?

A. Using only one highly detailed interview

B. Comparing evidence from different participant groups, settings or times

C. Allowing software to select the final themes

D. Removing all contradictory evidence

Correct Answer: B. Comparing evidence from different participant groups, settings or times

Question 7

Which statement best describes informed consent?

A. A manager’s permission is always sufficient.

B. Consent is a one-time administrative form unrelated to data use.

C. Participants voluntarily agree after receiving understandable information about the study.

D. Consent permits collection of any data discovered during research.

Correct Answer: C. Participants voluntarily agree after receiving understandable information about the study.

Question 8

Why are diary studies particularly useful?

A. They capture experiences repeatedly and close to when they occur.

B. They guarantee complete participant compliance.

C. They eliminate recall and interpretation entirely.

D. They are suitable only for single-session tasks.

Correct Answer: A. They capture experiences repeatedly and close to when they occur.

Question 9

In qualitative analysis, what is a negative case?

A. A participant who gives an unfavourable product rating

B. Evidence that contradicts or qualifies an emerging explanation

C. A data record that must always be deleted

D. A method for calculating sample size

Correct Answer: B. Evidence that contradicts or qualifies an emerging explanation

Question 10

Which is the most responsible use of generative AI in qualitative research?

A. Upload identifiable transcripts to any free tool.

B. Accept AI-generated themes without checking source evidence.

C. Use an approved tool with safeguards and verify outputs against the original data.

D. Allow AI to obtain consent on behalf of the research team.

Correct Answer: C. Use an approved tool with safeguards and verify outputs against the original data.