Observation, Documents and Multimodal Data

Introduction

Qualitative health research investigates how people experience illness, organise care, interpret professional practice and negotiate institutional systems. Interviews are valuable, but they cannot reveal everything. What people say may differ from what they do; important routines may be taken for granted; and meanings may be embedded in documents, images, digital traces, sounds and material arrangements. Observation, document analysis and multimodal inquiry therefore extend the researcher’s ability to study health care as it is experienced and enacted.

These methods are particularly important for health-care managers. Service quality is expressed not only in policy statements but also in waiting-room interactions, handovers, informal workarounds, signage, electronic records and patient-generated accounts. A policy may prescribe one pathway while staff and patients create another in practice. Rigorous qualitative inquiry makes such discrepancies visible without reducing them to simple compliance failures.

This chapter examines participant and non-participant observation, fieldwork, field notes, diaries, documents, photographs, audio-visual materials, digital ethnography and patient-generated content. It emphasises methodological fit, reflexivity, ethical governance, data management and analytic integration.

Learning objectives

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

1. Distinguish participant, non-participant, overt and covert observation.
2. Design fieldwork and produce analytically useful field notes and diaries.
3. Evaluate documents as social, organisational and political artefacts rather than neutral sources.
4. Apply ethical and practical principles to photographs, audio-visual materials, digital ethnography and patient-generated content.
5. Integrate multimodal data with interviews and other qualitative evidence.
6. Assess quality, reflexivity, confidentiality, data security and representational risk.
7. Translate findings into defensible implications for health-care management and service improvement.

1. Observation as situated inquiry

Observation is the systematic study of activities, interactions, settings, objects and embodied practices in their social context. It is not merely watching. The researcher makes theoretically informed decisions about what to attend to, where to position themselves, how long to remain in the field, what to record and how to interpret what is observed.

Observation is useful when the research question concerns:

• routines and workflows;
• communication and interaction;
• spatial organisation and accessibility;
• informal norms and workarounds;
• embodied or material practices;
• differences between formal policy and lived practice; and
• experiences that participants may find difficult to articulate.

Observation produces situated knowledge. A researcher may learn, for example, that a hospital’s stated appointment process is technically available but practically inaccessible because patients must navigate unclear signage, multiple queues and repeated identity checks. Such findings are not simply descriptions of behaviour; they illuminate how organisational systems distribute time, effort, dignity and opportunity.

1.1 Major forms of observation

Participant observation involves some degree of participation in the setting while observing social life. Participation may range from limited involvement, such as attending team meetings, to substantial involvement, such as joining a community health programme in an observer role. Participation can facilitate rapport and insider understanding, but it may also alter the setting and blur professional boundaries.

Non-participant observation involves observing without taking part in the activity. It can support analytic distance and reduce role confusion, although complete detachment is rarely possible. The researcher’s presence, identity and recording practices still influence the field.

Overt observation is disclosed to those being observed. It is normally the ethical default in health research because informed consent, privacy and accountability are central. Covert observation conceals the researcher’s identity or purpose. It is generally difficult to justify in clinical settings because patients and staff have heightened privacy interests, and because meaningful consent may be impossible. Any exceptional use requires strong ethical justification, minimal risk, independent review and compliance with applicable law and institutional policy.

Structured observation uses predefined categories or schedules, such as recording interruptions during handovers. Unstructured or semi-structured observation allows concepts to develop through field engagement. Postgraduate research often combines both: an initial open phase maps the setting, followed by focused observation guided by emerging analytic questions.

1.2 The observer’s position

The observer is not an invisible recording device. Age, gender, professional status, institutional affiliation, language, ethnicity and perceived authority can affect access and interaction. A clinician-researcher may be treated as an insider by staff but as an authority figure by patients. An external researcher may obtain greater distance but require more time to understand local meanings.

Reflexivity requires the researcher to examine how their position shapes access, attention, interpretation and relationships. Reflexive practice may include a positionality statement, regular supervision, a reflexive journal and explicit discussion of contradictory evidence. Reflexivity does not mean treating all interpretations as equally valid; it means making the conditions of knowledge production visible.

2. Designing fieldwork

Strong observation begins before entering the field. The researcher should define the phenomenon, setting, participants, time period and boundaries of observation. A fieldwork protocol commonly includes:

1. research questions and sensitising concepts;
2. site selection and rationale;
3. access arrangements and gatekeepers;
4. inclusion and exclusion criteria;
5. observation schedule and duration;
6. consent procedures;
7. safety and escalation procedures;
8. data-recording methods;
9. confidentiality and data-security arrangements; and
10. plans for analysis, feedback and dissemination.

2.1 Access and gatekeeping

Health-care organisations are layered fields. Formal access may be granted by a chief executive, research office or ethics committee, while practical access depends on ward managers, clinicians, receptionists, community leaders and patients. Gatekeepers can facilitate entry, but they may also shape who is visible to the researcher. Staff nominated by managers may present an unusually favourable account, while dissenting or marginalised voices remain inaccessible.

Researchers should negotiate access at multiple levels and clarify that participation is voluntary. They should avoid allowing managers to know precisely who declined unless disclosure is necessary and authorised. In clinical settings, observation must never disrupt care, delay treatment or expose confidential information unnecessarily.

2.2 Sampling observations

Observation is often purposive rather than statistically representative. Sampling may seek variation across:

• time: day, night, weekday and weekend;
• location: reception, consultation room, ward and community setting;
• activity: routine, crisis and transition points;
• role: patient, family member, clinician, manager and support worker; and
• case: high-performing, problematic or contrasting sites.

The aim is analytic richness and variation. Researchers should explain why particular observations were selected and what remains unseen. Claims must not exceed the fieldwork’s scope.

2.3 The observation guide

An observation guide provides focus without turning qualitative observation into a rigid checklist. Useful domains include:

The guide should be piloted. Overly broad guides produce undirected description; overly narrow guides prevent discovery.

3. Field notes, diaries and analytic memos

Field notes are the primary record of observation. They should distinguish what was seen or heard from what the researcher inferred. A practical structure contains four layers:

1. Descriptive notes: people, place, sequence, speech, objects and observable actions.
2. Verbatim or near-verbatim records: memorable phrases, with quotation marks used only when wording is sufficiently accurate.
3. Analytic notes: possible meanings, patterns, contrasts and questions.
4. Reflexive notes: the researcher’s emotions, assumptions, role and influence.

Notes should be written or expanded as soon as feasible. Memory tends to preserve striking events while losing routine context. Time stamps, pseudonyms, location codes and event identifiers support later reconstruction. Researchers should avoid recording unnecessary clinical identifiers, exact diagnoses or information that could identify a patient.

A field diary records the broader research process: access negotiations, ethical decisions, changes in sampling, personal reactions, emerging relationships and methodological problems. An analytic memo develops an interpretation. For example, repeated observations of staff apologising for delays may lead to a memo about “repair work,” connecting individual politeness to structural capacity constraints.

A useful note format is:

Field notes are data, not merely administrative records. Editing them can improve clarity, but the researcher should preserve an audit trail showing when and why substantive changes were made.

4. Documents as qualitative data

A document is a materially or digitally recorded text, image or artefact produced, circulated or used in a social setting. Examples include policies, guidelines, care plans, referral forms, complaint letters, meeting minutes, audit reports, training materials, posters, websites, electronic messages and patient-held records.

Documents can be studied in at least three ways:

• As evidence of content: What does the document say?
• As evidence of practice: How is it used, interpreted or ignored?
• As social action: What does producing the document accomplish, for whom and under what authority?

A policy is therefore not simply a container of information. It may legitimise a managerial priority, allocate responsibility, define a problem or create an auditable account of organisational activity.

4.1 Evaluating documents

Document analysis should address provenance, purpose, audience, production conditions, circulation, authority, completeness and silences. Questions include:

• Who created the document and whose voice is absent?
• When and why was it produced?
• What decision, event or institutional process does it record?
• Is it an original, revised, summarised or copied version?
• What terms are defined, and what assumptions remain implicit?
• How is compliance, risk or quality represented?
• What categories are made visible or invisible?
• How does the document compare with observed practice and participant accounts?

Authenticity and credibility are distinct. A document may be authentic but strategically selective. Conversely, an informal message may lack formal authority but reveal how work is actually coordinated.

4.2 Document sampling and analysis

Document sampling can be purposive, chronological, contrastive or theoretically driven. A researcher might compare policy versions before and after a service reorganisation, or compare official guidance with patient information materials and complaint correspondence.

Analysis commonly proceeds through:

1. cataloguing and contextualising documents;
2. close reading and initial coding;
3. identifying themes, categories, omissions and contradictions;
4. comparing documents across time, groups or purposes;
5. relating documents to observations and interviews; and
6. interpreting consequences for practice and power.

A document matrix can improve transparency:

5. Photographs, audio-visual materials and material culture

Images and recordings can document spaces, gestures, objects, interactions and sensory dimensions that text may miss. A photograph of a reception area may reveal barriers, surveillance arrangements or unequal access to seating. A video may show turn-taking, interruptions and non-verbal communication. Audio can preserve tone, pauses and overlapping speech.

However, visual and audio data are not transparent reproductions of reality. Framing, camera position, timing, editing and selection shape what is represented. A photograph may show a crowded corridor but not the people who left before the camera arrived. A recording captures a performance in the presence of a researcher and device, not an untouched event.

5.1 Ethical principles for visual and audio data

Visual and audio materials can identify people even when names are removed. Consent must address:

• what will be recorded;
• who may appear incidentally;
• where the material will be stored;
• who will view it;
• whether it may be used in teaching, publications or presentations;
• whether withdrawal is possible after dissemination; and
• whether faces, voices, tattoos, uniforms, locations or metadata create identification risks.

In busy clinical environments, incidental capture is a major concern. Researchers should use restricted recording zones, signage where appropriate, active consent procedures and immediate deletion protocols for accidental capture. Images of patients who are unconscious, distressed, children or cognitively impaired require heightened safeguards and legally valid consent or authorisation.

Photographs should not be treated as neutral proof of poor practice. Researchers should record context, date, purpose and selection rationale, and should invite participants to discuss images where appropriate. Participant-led photography can generate valuable insight but requires clear ownership, privacy and publication agreements.

6. Digital ethnography and online health spaces

Digital ethnography studies social life mediated through online platforms and connected technologies. Relevant settings include patient forums, social media groups, telehealth encounters, online support communities, health-app interfaces, discussion boards and digital advocacy networks.

Researchers must distinguish between public accessibility and ethical permission. A post visible without a password may still be experienced by its author as belonging to a bounded community. Quoting it verbatim can make it searchable and identifiable. Ethical decisions should consider sensitivity, vulnerability, expectations of privacy, platform terms, consent feasibility and the consequences of exposure.

Digital data also have distinctive properties:

• content can be edited, deleted, reposted or algorithmically amplified;
• usernames may be pseudonyms rather than anonymous identities;
• bots and commercial accounts may resemble patients;
• platform interfaces shape visibility and interaction; and
• online communities may cross national legal jurisdictions.

Researchers should archive data lawfully, minimise collection, document capture dates and preserve relevant context. When reporting, paraphrasing may be safer than verbatim quotation, especially for sensitive posts. If direct quotation is methodologically essential, the researcher should assess searchability and consider consent.

Digital ethnography should not assume that online communication is less real than face-to-face interaction. It may be a primary site of care coordination, identity work, advocacy and emotional support. At the same time, online participants may not represent all patients; people without connectivity, digital confidence or safe access can be systematically excluded.

7. Patient-generated content

Patient-generated content includes diaries, symptom logs, photographs, audio reflections, wearable data, app entries, online narratives, drawings and videos. It can shift the balance of interpretive authority by enabling patients to document experiences in their own time and preferred medium.

Patient-generated methods are especially useful for fluctuating symptoms, transitions between services, home-based care and experiences that are difficult to recall in an interview. A patient may record the practical consequences of fatigue, medication timing or transport barriers more accurately in the moment than weeks later.

The method also creates risks. Participants may feel pressure to produce data, monitor themselves excessively or disclose information about relatives and clinicians without their consent. Digital tools may privilege participants with devices, literacy, reliable connectivity and confidence. Researchers must provide accessible instructions, alternatives to digital participation and clear boundaries about whether submitted information will trigger clinical action. A research diary is not an emergency communication channel.

Ownership and control should be negotiated. Participants should know whether they can review, amend or withdraw contributions and how extracts will be selected. Compensation should not become coercive, particularly where participants are economically vulnerable.

8. Integrating multimodal evidence

Multimodal research combines different forms of data while respecting their distinctive properties. The goal is not to make every source say the same thing. Difference can be analytically productive. A policy may promise shared decision-making, observations may show clinician-led conversations, and patient diaries may describe uncertainty after discharge. These sources illuminate different layers of the same phenomenon.

Useful integration strategies include:

• triangulation: comparing sources to assess convergence and divergence;
• complementarity: using one source to elaborate another;
• expansion: examining dimensions unavailable in a single method;
• temporal sequencing: linking documents, observations and accounts across time; and
• critical contradiction analysis: asking why accounts differ and what institutional interests shape each representation.

A joint display can support analysis:

Integration should occur during design, data collection and interpretation, not only at the writing stage. Researchers may use qualitative software to link transcripts, photographs, documents, audio segments and codes, but software does not perform interpretation. File naming, version control, access permissions and a data dictionary are essential for large multimodal projects.

9. A worked applied scenario: improving outpatient chemotherapy navigation

This is an illustrative fictional case.

Background

A regional cancer centre reports missed appointments and repeated patient complaints about confusing arrival procedures. Management initially assumes that patients need clearer written instructions.

Problem and design

A qualitative team asks: “How do patients and staff experience the journey from arrival to chemotherapy administration?” The team uses non-participant observation in the entrance, registration area and infusion unit; analysis of appointment letters, signage and standard operating procedures; patient travel diaries; and interviews conducted after observation. Recording in clinical treatment areas is prohibited unless separately authorised.

Observation is sampled across morning and afternoon sessions and includes routine and high-volume days. Researchers use field notes rather than photographs because patients could be identifiable. Patient diaries include time of arrival, route taken, points of uncertainty, interactions and emotional responses.

Analysis

Documents present the pathway as a three-step process: check in, complete blood tests and proceed to the unit. Observations show that patients often encounter a fourth step: locating a temporary registration desk whose position changes during construction. Staff use informal directions such as “go upstairs and turn left after the old pharmacy,” although the pharmacy has moved. Diaries reveal that anxiety is intensified when patients fear being late for treatment or believe that asking questions will delay others.

The team compares sources rather than treating one as definitive. The written pathway is accurate from an administrative perspective, but it omits temporary spatial arrangements, emotional work and the time required to resolve uncertainty.

Approach and solution

The team recommends a co-designed navigation intervention: consistent signage, a map using landmarks familiar to patients, a clearly identified welcome point, staff scripts that invite questions, and a process for updating information when construction changes. The team also recommends monitoring whether the intervention works for people with limited English, visual impairment and mobility limitations.

Outcome

In this illustrative case, the centre reports fewer navigation complaints after implementation, but the qualitative evaluation finds that some patients still rely on family members because written instructions remain linguistically complex. The intervention is therefore revised to include translated and audio formats.

Learning points

1. Administrative documents describe intended pathways, not necessarily lived pathways.
2. Observation reveals spatial and interactional work that may be invisible in interviews.
3. Patient-generated diaries capture experiences between institutional encounters.
4. Multimodal findings can identify both operational and emotional causes of service problems.
5. Improvement should be co-designed and evaluated for equity, not judged only by aggregate complaints.

10. Quality, ethics and governance

Quality in observational and multimodal research involves credibility, dependability, confirmability, transferability and ethical integrity. Practical strategies include prolonged engagement, persistent observation, comparison across cases, negative-case analysis, participant feedback where appropriate, peer debriefing, an audit trail and thick description.

Member checking should not be treated as a compulsory test of truth. Participants may disagree with an interpretation because they have changed their views, occupy a different role or wish to protect the organisation. Researchers should explain how feedback was used rather than claiming that participant approval automatically validates findings.

Governance requirements commonly include research ethics approval, site permission, data-protection compliance, consent documentation, secure storage, retention schedules and procedures for adverse events. Health data are often sensitive personal data. De-identification reduces risk but does not guarantee anonymity, especially in small teams, rare conditions, distinctive voices or recognisable locations.

Researchers must plan for unexpected findings. Observation may reveal unsafe practice, abuse, serious neglect or imminent risk. The consent process should explain relevant limits to confidentiality, and the protocol should specify escalation routes. Researchers should not promise absolute secrecy when professional or legal duties require action.

Power deserves specific attention. Patients may believe participation affects treatment; employees may fear managerial retaliation; and community members may feel obliged to cooperate with a respected researcher. Recruitment should separate research participation from clinical or employment decisions wherever possible.

11. Technology, AI and contemporary developments

Mobile recording, secure cloud systems, electronic health records, wearable devices and online platforms expand the possibilities of qualitative inquiry. They also multiply risks concerning data linkage, cybersecurity, ownership and secondary use.

Artificial intelligence may assist with transcription, translation, image organisation, search, redaction and preliminary pattern discovery. Its use requires human oversight. Automated transcription can misrecognise accents, clinical terms, code-switching and emotionally significant pauses. Translation can erase culturally meaningful expressions. Image-recognition systems may reproduce demographic bias. Commercial tools may retain uploaded data or use it for model training.

Researchers should document the tool, version, purpose, data sent to it, quality checks and human decisions. Sensitive recordings should not be uploaded to an unapproved service. AI-generated coding is an aid, not a substitute for contextual interpretation, reflexivity or accountability. Participants should be informed when automated processing materially affects the handling or interpretation of their data, in accordance with institutional and legal requirements.

12. Best-practice checklist

Before fieldwork, confirm that the question requires observation or multimodal evidence, access is ethically authorised, recording boundaries are clear, and participants can decline without disadvantage. During fieldwork, protect care delivery, write notes promptly, distinguish observation from interpretation, monitor changing consent and record reflexive decisions. During analysis, preserve provenance, compare sources, attend to silences and contradictions, and avoid treating visual or digital material as self-evident. Before dissemination, assess re-identification risk, obtain appropriate permissions, represent participants fairly and report limitations. Finally, translate findings into feasible actions while retaining the complexity of the evidence.

Chapter Summary

Observation, document analysis and multimodal methods reveal dimensions of health care that interviews alone may miss. Participant and non-participant observation provide situated accounts of routines, interactions, space and material practice, while field notes, diaries and analytic memos preserve description, interpretation and reflexivity. Documents should be examined for provenance, purpose, authority, use, omissions and consequences rather than treated as neutral facts. Photographs, audio-visual materials, digital ethnography and patient-generated content can enrich inquiry, but they require careful attention to consent, privacy, ownership, searchability, accessibility and data security.

Multimodal analysis is strongest when sources are compared critically. Convergence can strengthen an interpretation, while contradiction may expose differences between policy, practice and lived experience. High-quality research requires transparent sampling, ethical governance, reflexive positioning, secure data management, negative-case analysis and appropriately cautious claims. Technology and AI can support handling and analysis, but they do not replace human judgement. For health-care managers, these methods offer practical ways to understand service pathways, identify inequities and design improvements grounded in patients’ and workers’ everyday experiences.

Key Terms

Observation: Systematic study of activities, interactions, settings and practices in context.

Participant observation: Observation involving some degree of participation in the setting.

Non-participant observation: Observation without taking part in the activity being studied.

Overt observation: Observation in which the researcher’s identity and purpose are disclosed.

Covert observation: Observation in which the researcher’s identity or purpose is concealed.

Fieldwork: Planned engagement with a social setting to collect and interpret qualitative data.

Field notes: Written records of observed events, speech, context and researcher interpretation.

Reflexivity: Critical examination of how the researcher shapes knowledge production.

Field diary: Ongoing record of research decisions, access, emotions and methodological developments.

Analytic memo: A written development of emerging concepts, relationships or interpretations.

Document analysis: Systematic interpretation of recorded texts, images or artefacts.

Provenance: The origin, authorship and production history of a document or data item.

Multimodal data: Data conveyed through more than one mode, such as text, image, sound and video.

Digital ethnography: Ethnographic study of social life mediated through digital technologies.

Patient-generated content: Data created by patients, including diaries, posts, photographs and app entries.

Triangulation: Comparison of multiple sources, methods or perspectives in inquiry.

Member checking: Seeking participant responses to interpretations or findings.

Re-identification risk: The possibility that a supposedly de-identified person can be recognised.

Gatekeeper: A person or organisation controlling or facilitating access to a research setting.

Negative-case analysis: Examination of evidence that challenges an emerging interpretation.

Review Questions

Short Answer Questions

1. Distinguish participant observation from non-participant observation.
2. Identify the four principal layers commonly included in field notes.
3. Why should documents be analysed as social actions as well as sources of information?
4. State two ethical risks associated with visual or audio data.
5. What is meant by reflexivity in qualitative health research?

Long Answer Questions

1. Explain how a researcher would design an observation study in a hospital ward, including access, sampling, consent, field notes and data protection.
2. Discuss a systematic approach to analysing a health-care policy document, including provenance, purpose, audience, silences and relation to practice.
3. Critically evaluate the opportunities and risks of digital ethnography in online patient-support communities.
4. Explain how patient-generated diaries can complement interviews and observation in research on chronic illness.
5. Discuss how multimodal data can be integrated without assuming that all sources should produce the same account.

Analytical / Application Questions

1. A clinic’s policy states that patients receive shared decision-making, but observations show rushed consultations and diaries show uncertainty at home. Design an analytic strategy for interpreting this divergence.
2. You are asked to photograph a crowded emergency-department waiting room to study accessibility. Identify the ethical, methodological and governance issues and propose a safer alternative design.
3. A hospital wants to use an AI transcription service for recorded patient interviews. Develop a decision framework addressing accuracy, confidentiality, bias, consent, human oversight and reporting.

Multiple Choice Questions with Answers

Question 1

Which feature most clearly distinguishes participant observation from non-participant observation?

A. Participant observation always uses a structured checklist.

B. Participant observation involves some degree of involvement in the setting.

C. Participant observation does not require field notes.

D. Participant observation can only be conducted online.

Correct Answer: B. Participant observation involves some degree of involvement in the setting.

Question 2

Which item is an example of a reflexive field note?

A. The clinic opened at 08:00.

B. The receptionist directed three patients to the same queue.

C. The researcher’s professional status may have encouraged staff to present the process positively.

D. A patient waited beside the registration desk.

Correct Answer: C. The researcher’s professional status may have encouraged staff to present the process positively.

Question 3

When analysing a health-care policy document, provenance refers primarily to the document’s:

A. Length and formatting.

B. Origin, authorship and production history.

C. Number of citations in academic literature.

D. Reading difficulty for patients.

Correct Answer: B. Origin, authorship and production history.

Question 4

Why is public accessibility insufficient by itself to establish that online patient content may be used without ethical consideration?

A. Public content is always inaccurate.

B. Online content cannot be archived.

C. Users may still reasonably expect privacy within a community context.

D. Only clinicians can interpret online content.

Correct Answer: C. Users may still reasonably expect privacy within a community context.

Question 5

Which practice best reduces the risk of identifying patients in visual research?

A. Publishing all images without alteration.

B. Recording only in restricted areas with appropriate consent procedures.

C. Removing the file name but retaining recognisable faces.

D. Assuming that blurred images are always anonymous.

Correct Answer: B. Recording only in restricted areas with appropriate consent procedures.

Question 6

A document says that discharge explanations are provided, while patient diaries describe confusion at home. The most appropriate interpretation is to:

A. Discard the diaries because the policy is official.

B. Assume that patients misunderstood the policy.

C. Examine how formal completion differs from meaningful understanding.

D. Conclude that the policy is fabricated.

Correct Answer: C. Examine how formal completion differs from meaningful understanding.

Question 7

Which is a major limitation of patient-generated digital data?

A. They can never capture experiences over time.

B. They may exclude people without reliable devices, connectivity or digital confidence.

C. They are always less detailed than interviews.

D. They cannot be ethically governed.

Correct Answer: B. They may exclude people without reliable devices, connectivity or digital confidence.

Question 8

What is negative-case analysis?

A. Deleting data that contradicts the main theme.

B. Examining evidence that challenges an emerging interpretation.

C. Counting only negative patient experiences.

D. Comparing two statistical samples.

Correct Answer: B. Examining evidence that challenges an emerging interpretation.

Question 9

What is the most appropriate role of AI-generated transcription in qualitative research?

A. It replaces researcher interpretation.

B. It is accepted without checking because software is objective.

C. It can assist processing but requires human quality checks and accountability.

D. It eliminates the need for participant consent.

Correct Answer: C. It can assist processing but requires human quality checks and accountability.

Question 10

Why can contradictions between policy, observation and patient accounts be analytically valuable?

A. They prove that one data source must be false.

B. They may reveal differences between intended systems, enacted practice and lived experience.

C. They prevent any defensible conclusion.

D. They show that qualitative research is unreliable.

Correct Answer: B. They may reveal differences between intended systems, enacted practice and lived experience.