Qualitative and Mixed-Methods Evaluation

Introduction and context

Health-care evaluation is concerned not only with whether an intervention produces an outcome, but also with how, why, for whom and under what conditions that outcome occurs. Quantitative measures may show that waiting time fell, hospital readmissions declined or patient satisfaction increased. They may not explain how staff changed their practice, why some patients benefited more than others, how implementation was experienced, or which contextual conditions shaped the result. Qualitative and mixed-methods evaluation addresses these questions by examining meaning, experience, behaviour, relationships, processes and context alongside measurable outcomes.

Qualitative evaluation is particularly valuable where an intervention is complex, delivered through social interaction, or implemented in variable organisational settings. Mixed-methods evaluation deliberately combines qualitative and quantitative evidence within a single evaluation design. Its purpose is not merely to place interview quotations beside statistical tables. Rather, it is to produce an integrated explanation that is more useful than either form of evidence alone.

For health-care managers, this approach supports service redesign, quality improvement, implementation, equity analysis and accountable decision-making. For researchers, it requires methodological coherence: the evaluation questions, philosophical assumptions, sampling, data collection, analysis and integration strategy must fit together.

Learning objectives

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

1. Explain the purposes, assumptions and practical uses of qualitative evaluation in health care.
2. Design and conduct interviews, focus groups, observation and document analysis responsibly.
3. Develop a credible sampling and data-management strategy.
4. Undertake thematic analysis and distinguish description from interpretation.
5. Use triangulation to examine convergence, complementarity and discrepancy.
6. Compare major mixed-methods designs and select an appropriate design for an evaluation question.
7. Integrate qualitative and quantitative findings to explain outcomes, context and stakeholder experience.
8. Assess rigour, ethics, governance and limitations in qualitative and mixed-methods evaluation.

1. Foundations of qualitative evaluation

1.1 What qualitative evaluation investigates

Qualitative evaluation examines how people interpret and experience a programme, service or policy and how social, organisational and material conditions influence implementation and outcomes. It can address questions such as:

• How do patients experience a new virtual consultation service?
• How do professionals adapt a clinical pathway in practice?
• Why is an evidence-based intervention adopted in one clinic but resisted in another?
• What unintended consequences are experienced by marginalised groups?
• How do formal policy requirements differ from work as actually performed?

The central object of inquiry may be an experience, process, mechanism, relationship, culture, implementation context or change in meaning. Qualitative evaluation does not imply an absence of structure or analytical discipline. It uses systematic procedures to produce an evidence-based interpretation.

1.2 Philosophical positions

Qualitative evaluation is often associated with interpretivism, which holds that social reality is understood through the meanings people construct in particular contexts. A patient’s account of “access” may include transport, digital confidence, trust, language, appointment flexibility and previous experiences of discrimination. These dimensions cannot always be reduced to a single indicator without losing important meaning.

Critical approaches add attention to power, inequality and whose knowledge is treated as authoritative. A participatory evaluation may involve patients or community members in defining questions, interpreting findings and deciding how results should be used. Realist approaches ask how mechanisms operate in particular contexts to produce outcomes. These perspectives are not interchangeable, but they can inform complementary evaluation questions.

A mixed-methods study often adopts pragmatism or a pluralist position. This does not mean that methods are combined without philosophical reflection. It means that the evaluation question determines the combination of evidence, while the evaluator remains explicit about assumptions concerning knowledge, causation, values and context.

1.3 Qualitative evidence and generalisation

Qualitative studies commonly use purposive rather than probability sampling. The objective is usually to obtain information-rich cases and variation in experience, not to estimate the prevalence of an opinion in a population. Qualitative findings may support:

• Transferability: readers judge whether findings apply to another setting, based on a detailed account of context.
• Analytical generalisation: findings contribute to a concept, explanation or theory that may be examined elsewhere.
• Theoretical generalisation: evidence refines propositions about how a mechanism works under specified conditions.

A small sample is not automatically rigorous, and a large sample is not automatically representative. Adequacy depends on the research question, diversity of relevant perspectives, quality of data and depth of analysis. Researchers should avoid claiming population prevalence from purposively selected interviews unless supported by appropriate quantitative evidence.

2. Designing a qualitative evaluation

2.1 From evaluation question to method

Method selection should follow the question. Interviews are suited to personal experience, reasoning and sensitive topics. Focus groups are useful for shared norms, language and interaction among participants. Observation examines behaviour and workflow in context. Document analysis reveals formal expectations, institutional memory and recorded decisions. Most complex evaluations benefit from more than one source.

A useful design sequence is:

1. Define the decision or learning need.
2. Identify stakeholders, processes, outcomes and contextual influences.
3. Formulate qualitative questions about experience, implementation and mechanism.
4. Select participants, settings and documents that can answer those questions.
5. Choose data-collection methods and develop flexible topic guides.
6. Establish ethical, governance and data-security procedures.
7. Plan analysis before collection, while allowing inductive discoveries.
8. Specify how qualitative findings will connect to quantitative results or decisions.

2.2 Interviews

A semi-structured interview uses a consistent set of topics while allowing participants to explain issues in their own terms. A good topic guide moves from broad questions to focused probes. For example:

• “Please describe your experience of the new discharge process.”
• “What happened when you first used it?”
• “What made it easier or more difficult?”
• “Did the process work differently for different patients?”
• “What would you change?”

Leading questions, premature assumptions and technical jargon can distort accounts. Probes such as “Can you tell me more?”, “What happened next?” and “How did that affect you?” encourage detail. Interviews may be conducted face to face, by telephone or online, but the mode can affect disclosure, accessibility and rapport.

The evaluator should record relevant contextual information, including role, setting, timing and interruptions. Audio recording and verbatim transcription are generally preferable when consent has been obtained. Field notes should document impressions, emerging interpretations and possible reflexivity concerns, but impressions must not be confused with evidence supplied by participants.

2.3 Focus groups

A focus group generates data through interaction. Participants may clarify, challenge or build on one another’s accounts, revealing shared expectations and disagreement. It is appropriate for exploring service language, collective experiences, acceptability and potential solutions.

Composition requires careful judgement. A group that combines senior clinicians and junior staff may inhibit criticism. Groups may therefore be organised around relevant power, role, age, language or experience differences. The moderator should establish ground rules, invite quieter participants, prevent domination and monitor distress. Focus-group data are not simply a collection of individual interviews: agreement, silence, humour, disagreement and changes in position are analytically meaningful.

Sensitive disclosures, highly unequal power relationships or a need for detailed personal narratives may make individual interviews safer and more suitable.

2.4 Observation

Observation examines what happens in practice rather than relying solely on what people recall or report. It may be non-participant, where the evaluator observes without taking an operational role, or participant, where the evaluator becomes involved in the setting. Structured observation uses predefined categories; ethnographic observation is more open and attentive to routines, relationships, artefacts and unexpected events.

An observation protocol should specify the setting, activities, actors, time period, events of interest and approach to note-taking. The evaluator should distinguish:

• Descriptive notes: what was seen or heard.
• Reflective notes: possible interpretations, questions and positionality.
• Analytic notes: emerging connections to the evaluation framework.

Observation can identify workarounds, interruptions, informal decision rules and discrepancies between policy and practice. It also creates heightened privacy risks, particularly in clinical environments. Consent, signage, exclusion procedures and protection of identifiable information are essential.

2.5 Document analysis

Documents include policies, clinical guidelines, meeting minutes, referral forms, electronic messages, audit reports, complaints, training materials and patient information. They may be the object of analysis or a source of evidence about implementation.

The evaluator should examine authorship, purpose, audience, date, completeness, circulation and power. A policy document records an organisational expectation; it does not prove that the expectation was implemented. Similarly, complaints may illuminate important failures but cannot by themselves establish their frequency in the wider population. Document analysis can compare official accounts with interviews and observations, revealing gaps between formal and lived systems.

2.6 Sampling and saturation

Purposive sampling selects participants because they possess relevant experience. Maximum-variation sampling seeks contrasting cases, such as urban and rural clinics or new and experienced staff. Criterion sampling includes all cases meeting a defined condition. Typical-case, critical-case and theoretical sampling serve different purposes.

Sample size should be justified by information needs, heterogeneity, method and resources. “Saturation” is used in several ways. Code saturation means that few new codes are appearing; meaning saturation means that the dimensions and nuances of existing themes are well understood. A study should state which concept is intended and how the judgement was made. Researchers should not stop solely because a predetermined number has been reached if important perspectives remain absent.

3. Data management and thematic analysis

3.1 Preparing the dataset

A credible analysis begins with an organised dataset. Each item should have a unique identifier and associated metadata, such as participant group, setting, date and method. Identifiers should be removed or replaced with pseudonyms. A data-management plan should address transcription, translation, storage, access, retention and destruction.

The evaluator should read the data repeatedly, record initial observations and maintain a reflexive journal. Reflexivity involves examining how the evaluator’s professional role, assumptions, identity, relationship with participants and institutional interests may shape data production and interpretation. It does not require pretending to be neutral; it requires making influence visible and subjecting interpretations to scrutiny.

3.2 Thematic analysis

Thematic analysis identifies patterned meaning across a dataset. It can be primarily inductive, allowing themes to develop from the data, or deductive, using an existing framework or evaluation model. In practice, many evaluations use a hybrid approach.

A practical analytic sequence is:

1. Familiarisation: read transcripts, notes and documents closely.
2. Initial coding: label relevant segments concisely. Codes may describe content, action, emotion, process or underlying meaning.
3. Code development: compare codes, clarify definitions, merge overlaps and identify exclusions.
4. Theme construction: group related codes into candidate themes that express a patterned idea.
5. Review: test themes against coded extracts and the whole dataset.
6. Definition and naming: state each theme’s central organising concept and boundaries.
7. Interpretation and reporting: explain significance, variation, relationships and implications.

A theme is more than a topic. “Communication” is a topic; “communication was treated as task completion rather than shared decision-making” is an interpretive theme. Themes should be internally coherent, distinct enough to be useful and supported by sufficient evidence.

A codebook may include the code name, definition, inclusion criteria, exclusion criteria and example. Software can assist with retrieval, coding, memoing and audit trails, but it does not perform interpretation automatically. The unit of analysis may be a phrase, incident, episode, document or whole case; it should be selected consistently with the research question.

3.3 Reporting qualitative findings

Strong reporting combines interpretation with evidence. A theme should be explained, illustrated with carefully selected quotations or observations, and related to context and competing interpretations. Quotations should not be overused or selected merely because they are vivid. The report should indicate whether an account was common, unusual, contested or associated with a particular subgroup, without implying statistical frequency where it was not measured.

Negative cases are especially valuable. A negative case is an instance that does not fit the emerging explanation. Investigating it can refine a theme, identify boundary conditions or expose an unjustified assumption. Member reflection, in which participants comment on emerging interpretations, may improve credibility, although participants cannot be expected to validate every analytic conclusion and disagreement should not automatically be treated as researcher error.

4. Triangulation and quality

Triangulation involves examining an evaluation question through multiple perspectives, methods, sources, investigators or theoretical lenses. It is not a rule that all sources must agree. Divergence may reveal different realities, measurement limitations, power relations or context-dependent effects.

Common forms include:

Quality is multidimensional. Credibility concerns the plausibility and evidential support of interpretations. Dependability concerns the transparency and consistency of the analytic process. Confirmability concerns the extent to which findings are grounded in data rather than undisclosed preference. Transferability is supported by thick description of context, participants and processes.

Useful practices include an audit trail, reflexive memos, peer debriefing, deviant-case analysis, transparent reporting of sample limitations and explicit links between raw data, codes, themes and conclusions. Inter-coder agreement can be useful in some team projects to identify inconsistent code interpretation, but a high percentage of agreement is not by itself proof of validity. Qualitative quality should not be reduced to a mechanical reliability score.

5. Mixed-methods evaluation

5.1 Purpose and logic

Mixed-methods evaluation integrates qualitative and quantitative approaches so that one form of evidence informs, explains, tests, develops or expands the other. Integration should occur at one or more of four points:

• Design: one strand determines sampling or questions in the other.
• Methods: instruments or participants are linked across strands.
• Analysis: findings are transformed, compared or combined.
• Interpretation: conclusions are drawn from the relationship between strands.

A study is not genuinely mixed methods merely because a report contains both a survey and interviews. The evaluator must show how the strands were connected and what additional insight resulted.

5.2 Major designs

Convergent design collects qualitative and quantitative data during a similar period, analyses them separately and compares the findings. It is useful when the evaluator wants a balanced understanding of outcomes and experience.

Explanatory sequential design begins with quantitative results and follows with qualitative inquiry. Interviews may explain an unexpected outcome, subgroup difference or implementation pattern. For example, a survey may show lower satisfaction among older patients, after which interviews explore accessibility, communication and digital barriers.

Exploratory sequential design begins qualitatively and uses findings to develop a survey, intervention, measure or implementation strategy. It is useful when concepts are poorly understood or existing instruments do not reflect local experience.

Embedded design places one method within a larger study. Qualitative observation may be embedded in a controlled evaluation to explain fidelity and adaptation, while quantitative outcomes remain the primary endpoint.

Multiphase design links several studies over time, such as qualitative development, quantitative testing and qualitative implementation analysis.

Design choice depends on timing, priority, resources, intervention complexity and the decision to be supported. Sequential designs require sufficient time between phases; convergent designs require a clear plan for resolving discordant findings.

5.3 Integration tools

A joint display is a table or visual arrangement that places qualitative themes and quantitative results together. For example:

Other integration techniques include connecting samples, building one instrument from another strand, merging datasets for case comparison and producing a meta-inference. A meta-inference is a conclusion drawn from the integrated evidence rather than from either strand alone.

5.4 Interpreting convergence and discrepancy

Convergence occurs when different strands support a similar conclusion. Complementarity occurs when strands illuminate different aspects of the same phenomenon. Expansion occurs when one strand broadens the scope of the other. Discordance occurs when findings conflict.

Discordance should be investigated, not concealed. Possible explanations include:

• different populations or time periods;
• differences between perceived and observed behaviour;
• social desirability in self-report;
• limitations in measurement validity;
• implementation variation between sites;
• genuine subgroup differences; or
• an intervention that improved one outcome while worsening another.

The evaluator should state which explanation is best supported and what uncertainty remains. Integration is a reasoning process, not an obligation to manufacture a single harmonious conclusion.

6. Applied case: evaluating a hospital-to-home medicines service

The following is an illustrative case, designed to demonstrate evaluation reasoning rather than report a real institution.

Background and problem

A regional hospital introduced a pharmacist-led medicines-reconciliation service for adults discharged after medical admission. The programme aimed to reduce medication discrepancies and prevent avoidable readmissions. After six months, administrative data showed a modest reduction in 30-day readmissions, but the improvement varied substantially between wards. Managers wanted to know whether the variation reflected implementation quality, patient experience, case mix or unobserved harms.

Evaluation design

The evaluation used an explanatory sequential mixed-methods design. Quantitative analysis compared readmission rates and documented medication discrepancies across wards, while monitoring patient characteristics and service reach. The quantitative findings showed the strongest improvement on wards where pharmacists completed reconciliation before discharge and contacted primary-care teams. One ward showed little improvement despite high recorded completion.

The qualitative phase purposively sampled patients with and without readmission, ward nurses, pharmacists, doctors, carers and primary-care professionals. Interviews explored discharge communication and medication understanding. Observation examined the discharge process on all three wards. Documents included reconciliation forms, discharge summaries, escalation policies and telephone-contact logs.

Analysis

Thematic analysis produced four themes:

1. Reconciliation as a clinical conversation: patients valued explanation and opportunity to ask questions, not merely a corrected list.
2. Completion recorded before coordination occurred: staff sometimes marked tasks complete while community communication remained delayed.
3. Informal work sustained continuity: nurses and carers often compensated for missing information through personal calls.
4. Complexity was unevenly distributed: patients with cognitive impairment, language barriers or multiple prescribers required more time and tailored support.

Observation supported the second and third themes. Documents showed that the completion indicator measured whether a form was signed, not whether the receiving professional had confirmed the plan. Interviews also revealed that some patients were reluctant to disclose confusion because discharge interactions felt rushed.

Integrated interpretation

The quantitative result suggested that the service was effective where reconciliation was linked to cross-boundary coordination. Qualitative evidence explained why recorded completion on one ward did not produce the same outcome: the process had become an administrative endpoint rather than a communication pathway. The integrated conclusion was therefore conditional, not simply “the programme worked” or “the programme failed.”

Approach and outcome

Managers revised the indicator set to distinguish medication review, patient teach-back, communication to primary care and resolution of outstanding discrepancies. The service introduced a risk-based pathway, interpreter access and a follow-up call for selected patients. Subsequent monitoring was planned, but the evaluation did not claim that the changes alone caused future improvements. The revised design included repeated quantitative monitoring and targeted observation to assess implementation.

Learning points

The case demonstrates that outcome data may identify variation without explaining it. Interviews reveal meaning and barriers; observation tests whether reported processes occur; documents expose the assumptions embedded in measures. Integration produced a more actionable finding than any single method. It also showed that an apparently simple process measure can hide important differences in quality and equity.

7. Ethics, governance and power

Qualitative evaluation often deals with identifiable stories, professional criticism and sensitive health information. Ethical review and organisational governance should address informed consent, voluntary participation, confidentiality, data minimisation, secure storage, withdrawal, distress and conflicts of interest. In focus groups, the evaluator cannot guarantee that other participants will preserve confidentiality; this limitation must be explained.

Patients may feel obliged to participate when approached by their care team. Recruitment should separate clinical care from evaluation decisions where possible. Staff may fear managerial consequences if they criticise a service. Reporting should avoid identifiable combinations of role, location and quotation, especially in small teams.

Consent for observation requires attention to everyone who may be encountered, including patients who are not research participants. If obtaining individual consent is impracticable, the evaluator must use an ethically justified procedure approved through appropriate governance and provide opt-out opportunities where possible.

Equity should be designed into sampling rather than added at the reporting stage. People with limited English, disabilities, unstable housing, low digital access or insecure immigration status may be systematically absent from conventional recruitment. Accessible materials, interpreters, flexible modes and fair compensation can improve inclusion. Researchers should also consider extractive practices: communities should receive findings in accessible forms and have meaningful opportunities to influence action.

8. Technology, AI and contemporary developments

Computer-assisted qualitative data-analysis software can support coding, retrieval, memoing, version control and joint displays. It cannot determine whether an interpretation is credible. Automated transcription saves time but may misrecognise accents, clinical terms, multilingual speech or emotionally significant pauses. Human checking is required.

Generative artificial intelligence may assist with administrative tasks such as proposing preliminary code labels or summarising large text collections, but its use creates risks of confidentiality breaches, fabricated summaries, loss of context, algorithmic bias and opaque decision-making. Sensitive health data should not be entered into an unapproved system. If AI is used, the evaluator should document the tool, purpose, data handling, human verification and limits. Final coding and interpretation remain accountable to the evaluation team.

Digital traces, electronic health records, portal messages and online communities create new documentary sources. They may provide longitudinal evidence but often reflect unequal participation and changing platform practices. Their use requires attention to consent, reasonable expectations of privacy, ownership, linkage error and the possibility that apparently public information remains ethically sensitive.

9. Challenges and limitations

Qualitative and mixed-methods evaluation requires time, skilled interviewing, careful management and interpretive judgement. Access can be difficult when services are busy or organisations fear unfavourable findings. Participants may offer socially desirable accounts, and the evaluator’s position may influence what is said or observed. Translation can alter meaning, while retrospective accounts may be shaped by current outcomes.

Mixed-methods studies can become expensive and analytically incoherent when both strands are conducted as separate projects. Integration may be superficial if researchers lack expertise in one method. Quantitative and qualitative strands may also have different definitions of the intervention, outcome or population. These differences should be reconciled explicitly.

No method eliminates uncertainty. Qualitative evidence can show plausible mechanisms and important experiences without estimating their prevalence. Quantitative evidence can estimate patterns without revealing their meaning or implementation conditions. The strongest evaluation states what each strand can support, what it cannot support and how the combined evidence changes confidence in the conclusion.

10. Best-practice checklist for managers and evaluators

A high-quality evaluation should:

• begin with a decision-relevant question rather than a preferred method;
• include stakeholders who experience, deliver and govern the service;
• sample for relevant variation and document who is missing;
• use topic guides that permit unexpected accounts;
• separate descriptive records from interpretation;
• maintain a clear audit trail and reflexive record;
• analyse discrepant and negative cases;
• protect confidentiality in small and powerful organisations;
• connect findings to implementation, equity and outcomes;
• specify the point and purpose of mixed-methods integration;
• use joint displays or equivalent tools to make integration visible; and
• communicate actionable findings without overstating causality or generalisability.

Chapter Summary

Qualitative evaluation explains experience, behaviour, implementation, mechanism and context. Interviews, focus groups, observation and document analysis answer different questions and should be selected deliberately. Purposive sampling seeks relevant variation and information-rich cases, while saturation should be justified rather than treated as a fixed numerical rule.

Thematic analysis involves familiarisation, coding, theme development, review, definition and interpretation. Rigour is strengthened through reflexivity, audit trails, peer challenge, negative-case analysis, transparent reporting and thick description. Triangulation compares sources, methods, investigators, theories or settings; agreement is useful, but divergence may reveal important contextual or measurement differences.

Mixed-methods evaluation integrates qualitative and quantitative evidence through design, methods, analysis and interpretation. Convergent, explanatory sequential, exploratory sequential, embedded and multiphase designs serve different purposes. Joint displays and explicit meta-inferences help evaluators move beyond parallel reporting. Ethical practice requires attention to consent, confidentiality, power, inclusion, data governance and responsible use of technology. The ultimate value of the approach lies in producing explanations that are credible, context-sensitive and useful for improving health-care decisions.

Key Terms

Qualitative evaluation: Systematic study of meaning, experience, process and context in order to assess or improve a programme or service.

Mixed-methods evaluation: An evaluation that intentionally integrates qualitative and quantitative evidence.

Interpretivism: A perspective emphasising socially constructed meanings and context-dependent understanding.

Pragmatism: An approach that selects and combines methods according to the evaluation problem and intended use of findings.

Purposive sampling: Selection of participants or cases because they possess characteristics relevant to the evaluation question.

Saturation: A judgement that further data are unlikely to add sufficient new codes, meanings or theoretical insight for the study purpose.

Semi-structured interview: An interview guided by common topics while allowing flexible probing and participant-led accounts.

Focus group: A facilitated group discussion in which interaction among participants is itself a source of data.

Observation: Systematic study of activities, behaviour and context as they occur in practice.

Document analysis: Critical examination of written, visual or electronic records as evidence of policy, practice or experience.

Reflexivity: Examination of how the evaluator’s position, assumptions and relationships influence the evaluation.

Code: A label applied to a segment of data to represent a relevant idea, action, experience or meaning.

Theme: A patterned and interpretive concept that organises important meaning across data.

Triangulation: Comparison of evidence from multiple sources, methods, investigators, theories or settings.

Joint display: A table or visual arrangement that integrates qualitative themes with quantitative findings.

Convergence: Agreement between evidence strands concerning a finding or interpretation.

Discordance: A meaningful difference or conflict between evidence strands.

Meta-inference: A conclusion generated from the integrated interpretation of qualitative and quantitative findings.

Transferability: The extent to which readers can judge whether findings apply in another context.

Review Questions

Short Answer Questions

1. What distinguishes a qualitative evaluation question from a quantitative outcome question?
2. Give two advantages and one limitation of focus groups in health-care evaluation.
3. What is the difference between a descriptive note and a reflective note in observation?
4. Define triangulation and name two forms of triangulation.
5. What is the purpose of a joint display in mixed-methods evaluation?

Long Answer Questions

1. Explain how an evaluator would design a thematic analysis from data collection through reporting.
2. Compare convergent, explanatory sequential and exploratory sequential mixed-methods designs.
3. Discuss how purposive sampling can support credible conclusions without producing statistical representativeness.
4. Explain why discordant qualitative and quantitative findings should be investigated rather than eliminated.
5. Critically assess the ethical challenges of observing care delivery in a hospital setting.

Analytical / Application Questions

1. A survey finds that patients report high satisfaction with a digital appointment system, but attendance remains lower among people over 75. Design a mixed-methods evaluation to explain this pattern, including sampling, data collection and integration.
2. Staff report that a new safety checklist is used consistently, while observation shows frequent omissions. Develop a triangulation strategy and identify at least three plausible explanations for the discrepancy.
3. An evaluation team has interviews from patients, nurses and managers, ward observations and readmission data. Propose a joint display and explain how it could support a defensible managerial decision without overstating causality.

Multiple Choice Questions with Answers

Question 1

Which method is most directly suited to examining how participants respond to and build on one another’s views?

A. Focus group

B. Document analysis

C. Structured observation

D. Routine data extraction

Correct Answer: A. Focus group

Question 2

In thematic analysis, what is the primary purpose of a codebook?

A. To calculate the intervention’s effect size

B. To define and organise how codes are applied

C. To replace participant recruitment

D. To guarantee statistical generalisability

Correct Answer: B. To define and organise how codes are applied

Question 3

Which statement best describes transferability?

A. The probability that a result occurred by chance

B. The percentage of codes agreed by two analysts

C. The reader’s judgement about applicability to another context

D. The number of participants who completed a survey

Correct Answer: C. The reader’s judgement about applicability to another context

Question 4

An evaluation collects outcome data first and then interviews selected participants to explain an unexpected subgroup result. Which design is this?

A. Convergent

B. Explanatory sequential

C. Exploratory sequential

D. Purely ethnographic

Correct Answer: B. Explanatory sequential

Question 5

Why can discordance between interview findings and administrative data be valuable?

A. It proves that one dataset must be discarded

B. It may reveal differences in context, perception or measurement

C. It eliminates the need for ethical review

D. It demonstrates statistical significance

Correct Answer: B. It may reveal differences in context, perception or measurement

Question 6

Which is an example of maximum-variation sampling?

A. Selecting only participants who agree with the intervention

B. Selecting participants from contrasting sites and stakeholder groups

C. Selecting every tenth record from a database

D. Selecting participants solely because they are easiest to contact

Correct Answer: B. Selecting participants from contrasting sites and stakeholder groups

Question 7

Which statement about qualitative themes is most accurate?

A. A theme is simply a broad topic heading

B. A theme is a patterned and interpretive concept supported by data

C. A theme must occur with identical frequency in every subgroup

D. A theme can be created without reviewing the dataset

Correct Answer: B. A theme is a patterned and interpretive concept supported by data

Question 8

What is the main purpose of a joint display?

A. To conceal contradictory findings

B. To place qualitative and quantitative findings together for integration

C. To convert all interview data into numerical scores

D. To replace the evaluation report

Correct Answer: B. To place qualitative and quantitative findings together for integration

Question 9

Which practice best protects confidentiality in a small clinical team?

A. Publishing every participant’s job title and exact quotation

B. Using identifiable quotations to increase realism

C. Removing or combining details that could identify individuals

D. Sharing raw transcripts with all managers

Correct Answer: C. Removing or combining details that could identify individuals

Question 10

What is a key limitation of automated transcription in health-care evaluation?

A. It always prevents data analysis

B. It may misrecognise accents, clinical terms or multilingual speech

C. It makes consent unnecessary

D. It guarantees researcher neutrality

Correct Answer: B. It may misrecognise accents, clinical terms or multilingual speech