Wits SOCL4028A: Research Methods for Health and Illness Honours Exam Guide

South Africa’s Honours-level research methods for health and illness sits at the point where theory, ethics, and methodological rigour meet real-world health systems. For SOCL4028A (Wits), exam success usually depends on more than memorising definitions: you need to show you can choose and justify methods, anticipate bias, apply ethics, and interpret findings in health and illness contexts. This study guide is designed to help you write high-scoring exam answers aligned with how honours markers typically assess reasoning, clarity, and methodological competence—especially within South African university and health research settings.

1) SOCL4028A Core Concepts: Health, Illness, and the Logic of Research Design (Wits)

At Honours level, “research methods” in health and illness is not merely a toolbox of techniques. It is an integrated way of thinking about what counts as evidence, how social processes shape health, and how we design studies that produce credible, ethical knowledge. SOCL4028A typically expects students to connect conceptual frameworks (health/illness theories, social determinants, power and inequality) with concrete design decisions (sampling, measurement, analysis, and trustworthiness).

1.1 Health vs illness: sociological framing that examiners value

A common weakness in exam scripts is treating health and illness as purely biological states. Instead, sociological research methods require you to demonstrate the difference between:

Health: often measured as indicators (morbidity, mortality, self-rated health, service access).
Illness: lived experience of symptoms and disability, interpreted through social meanings (stigma, gender norms, explanatory models, and help-seeking).

In health and illness research, the same “condition” can produce different outcomes depending on social position—such as employment status, housing stability, migration history, language access, and interaction with healthcare workers. This means research methods must capture not only incidence or severity but also processes like barriers to care, trust in systems, and pathways from symptom recognition to treatment.

1.2 Evidence and causality in social health research

Exams frequently test whether you can articulate the logic of causality without oversimplifying. In health and illness studies, you may encounter:

Associations (e.g., higher food insecurity linked with worse self-rated health).
Temporal ordering (did the exposure occur before the outcome?).
Mechanisms (how and why does the exposure influence health through intermediate steps?).
Confounding and selection effects (who enters the study? who drops out?).

A strong honours answer recognises that sociological health research often aims for plausible causal explanation, not just statistical significance. You should be able to explain how design choices strengthen causal inference even in complex social settings.

1.3 The research question as the organising principle

High marks are linked to how clearly you align research question → objectives → design → sampling → data collection → analysis → ethics.

A useful exam template is:

Main research question
Sub-questions
Key constructs (what exactly are you measuring or exploring?)
Population and setting (who, where, and under what health system context?)
Time frame (cross-sectional, longitudinal, retrospective, or prospective)
Feasibility and ethics (access, consent, risk, confidentiality)

Example of a health and illness research question that allows method alignment:

RQ: How do patients in Johannesburg experience barriers to accessing antiretroviral therapy (ART), and how do those barriers shape adherence and perceptions of health?
- Method fit: qualitative interviews + thematic analysis, or mixed methods to triangulate adherence barriers with service-use patterns.

This type of question pushes you to show you can justify whether you need breadth (quant) or depth (qual), and how you integrate them.

1.4 Research paradigms and what they mean for method choice

Honours exams often reward students who can explain paradigms in applied terms, not just list them.

You might be expected to understand (at a practical level):

Positivist / post-positivist orientations: knowledge through observable patterns; emphasis on measurement, reliability, and validity.
Interpretivist / constructivist orientations: knowledge through meaning-making; emphasis on context, reflexivity, and thick description.
Critical orientations: knowledge tied to power, inequality, and structural constraints; emphasis on emancipation, reflexivity, and revealing hidden mechanisms.

In health and illness research, different paradigms may justify different approaches:

A critical paradigm might treat stigma as a structural process affecting access and outcomes, making qualitative inquiry and intersectional analysis central.
A post-positivist paradigm might treat service access as a measurable construct requiring a survey with validated items.

A high-scoring answer demonstrates that paradigm choice affects:

how you define constructs (operationalisation),
what you consider “evidence,” and
how you argue credibility (validity/reliability vs trustworthiness/dependability/confirmability).

1.5 Design typologies you should know deeply

In SOCL4028A, you may be asked to compare:

Cross-sectional vs longitudinal (e.g., baseline vs follow-up)
Experimental/quasi-experimental vs observational (ethical constraints often limit RCTs in health sociology)
Qualitative designs: phenomenology, grounded theory, ethnography, case study
Mixed methods: concurrent vs sequential; triangulation; explanatory or exploratory sequencing

A strong exam script doesn’t just name designs—it explains the consequences.

Example: why longitudinal matters for health and illness

If you study adherence or perceptions of illness over time, cross-sectional data may miss:

how experiences change after diagnosis,
how side effects and stigma evolve,
whether barriers predict later outcomes.

Longitudinal follow-up can capture temporal pathways, though it introduces challenges:

attrition bias,
higher cost,
retention incentives and ethics.

2) Sampling, Measurement, and Data Quality in Health and Illness Research (Wits)

A major portion of research methods exams evaluates your ability to justify sampling and measurement decisions. In health and illness studies—particularly in South Africa where communities differ sharply in access, language, and health system navigation—sampling and data quality aren’t technical footnotes. They shape the entire meaning of your findings.

2.1 Sampling logic: probability and non-probability approaches

2.1.1 Probability sampling (when generalisation is the goal)

Probability sampling supports claims about population parameters. Common methods include:

Simple random sampling
Stratified sampling (e.g., by gender, age group, or province)
Cluster sampling (e.g., by clinics or communities)

In health research, stratification can be particularly useful where health behaviours and access vary across demographic groups. For example, if you expect different ART experiences by age group and gender, a stratified sample can reduce sampling variance and support subgroup comparisons.

2.1.2 Non-probability sampling (when depth, context, or theory-building is the goal)

Non-probability methods include:

Purposive sampling (selecting information-rich cases)
Maximum variation sampling
Snowball sampling (useful for hidden populations like certain forms of stigma-related care avoidance)
Theoretical sampling (common in grounded theory)

Non-probability sampling can be highly appropriate in illness research because lived experience is central. For instance, to explore stigma around tuberculosis (TB), you may need participants with different experiences:

those who successfully completed treatment,
those who interrupted treatment,
those who delayed seeking care.

A high-mark answer explains why the sampling method matches the question, not just that it exists.

2.2 Sample size: qualitative saturation vs quantitative power

Examiners may test knowledge of how sample size is justified differently in qualitative vs quantitative research.

2.2.1 Qualitative sample size and saturation

In qualitative work, “sample size” is often determined by:

theoretical saturation (new interviews stop adding new conceptual insights),
data saturation (themes become repetitive),
information richness and analytical demands.

You should be able to argue saturation carefully without pretending it is a mechanical number. For example, in interviews exploring illness narratives, you might reach saturation after varied participant inclusion, especially across key contexts (e.g., urban vs peri-urban clinics).

2.2.2 Quantitative sample size and power (and why it’s often context-specific)

For surveys, sample size is linked to statistical power and precision. Even if you don’t memorise formulae, you should show you understand:

how effect size assumptions affect required N,
how outcome prevalence influences sampling needs,
how attrition or non-response must be considered,
and how subgroup analyses require larger samples.

In South Africa, non-response can be significant due to:

distrust,
language barriers,
clinic workload constraints,
and uneven access to reliable contact information.

Thus, sample size planning in health research should include anticipated non-response and operational realities.

2.3 Operationalising constructs: from theory to measurable variables

Measurement is where conceptual clarity becomes empirical clarity. In health and illness research, key constructs like “stigma,” “access,” and “quality of life” require careful operationalisation.

2.3.1 Common construct-to-variable pathways

Stigma might be operationalised via:
- perceived discrimination items,
- internalised stigma scales,
- community norms measures.
Service access might be operationalised via:
- distance/time to clinic,
- appointment delays,
- transport costs,
- ability to obtain medications when needed.
Adherence might be operationalised via:
- self-reported adherence measures,
- pharmacy refill records,
- clinic appointment attendance,
- and (if feasible) biomarkers.

In exam responses, you should highlight that each operationalisation has trade-offs:

self-report is subject to social desirability bias,
clinic records may omit contextual reasons for missed appointments,
biomarkers have cost and ethical implications.

2.4 Reliability and validity: what examiners look for

2.4.1 Reliability

Reliability refers to consistency:

Test–retest reliability (same measurement across time under stable conditions)
Inter-rater reliability (consistent scoring across interviewers or coders)
Internal consistency (e.g., for multi-item scales)

For qualitative research, reliability becomes dependability—ensuring coding and interpretation are stable and transparent through auditing and coder agreement processes.

2.4.2 Validity

Validity refers to whether the instrument measures what it claims to measure. Key forms include:

Content validity (items cover the concept domain)
Construct validity (the construct behaves as theoretically expected)
Criterion validity (agreement with external criterion)

In health and illness contexts, construct validity is challenged by:

cultural differences in understanding symptoms,
language translation issues,
and social contexts that shape how people interpret health questions.

2.5 Measurement in multilingual South African contexts

South Africa’s multilingual realities make language a methodological issue, not just a logistical one. A strong exam answer may mention:

back-translation for survey instruments,
pilot testing,
interviewer training,
and how to handle literacy differences.

You should explain why translation affects validity:

if “stigma” items do not capture local meanings,
if response categories are misunderstood,
or if idioms for symptoms vary.

2.6 Bias and threats to data quality

A “data quality” section must go beyond listing biases. Examiners want you to link bias to design and mitigation strategies.

2.6.1 Selection bias

Selection bias occurs when participants differ systematically from those who did not participate.

Example: only patients who frequently attend clinics participate in a study on adherence, excluding those facing severe access barriers.

Mitigation:

broaden recruitment settings (multiple clinics),
document non-response characteristics if possible,
use weighting (quantitative) or diverse purposive sampling (qualitative).

2.6.2 Information bias (misclassification and reporting)

Information bias includes:

recall bias (especially for retrospective illness experiences),
social desirability bias (especially for sensitive topics like HIV disclosure or mental health).

Mitigation:

shorter recall periods,
neutral question wording,
ensuring privacy during interviews,
using validated scales,
triangulating self-report with records where ethically permitted.

2.6.3 Confounding

Confounding occurs when a third variable influences both exposure and outcome.

Example: socioeconomic status could affect both health service access and health outcomes.

Mitigation:

design strategies (restriction/matching),
analytical strategies (multivariable regression),
careful conceptual model building.

2.7 Trustworthiness in qualitative research: credibility, transferability, dependability, confirmability

Qualitative “validity” is often addressed via trustworthiness criteria:

Credibility: accuracy of interpretations from the participant viewpoint.
Transferability: extent findings apply to other contexts (supported via thick description).
Dependability: stability of the process (audit trails, consistent procedures).
Confirmability: minimising researcher bias through reflexivity and documentation.

To score well, include practical strategies:

triangulation (data/source/method),
member checking (where appropriate),
peer debriefing,
maintaining a coding diary,
and clear audit trails.

2.8 Concrete mini-case: designing a study on illness experiences in a Johannesburg clinic

Consider a hypothetical Honours project focused on illness experience and healthcare interactions in Johannesburg.

Purpose: Understand how patients interpret symptoms, experience stigma, and navigate healthcare systems.
Research question: How do patients describe the meaning of symptoms and how do these interpretations affect care-seeking and treatment experience?

A coherent methodological plan could be:

Design: qualitative case study across one clinic and surrounding community referral networks.
Sampling: purposive maximum variation sampling including:
- participants with recent diagnosis,
- participants with long-term illness,
- participants who delayed seeking care.
Data collection: semi-structured interviews in preferred languages with trained interviewers; include prompts on help-seeking pathways.
Analysis: thematic analysis with reflexive coding; create a codebook evolving across interviews.
Trustworthiness: audit trail, peer debriefing among supervisors/peers, and thick description of clinic context.

This mini-case demonstrates how method choices are justified by the conceptual logic of illness experience research.

3) Research Ethics, Reflexivity, and Governance in South African Health Research (Wits)

Ethics is a high-stakes area in health and illness research. In exams, you’re not only assessed on whether you know ethical principles, but whether you can apply them to realistic dilemmas—especially in South African contexts with vulnerability, power imbalances, and the complexity of healthcare systems.

3.1 Core ethical principles: aligning theory with practice

At honours level, you should demonstrate familiarity with foundational ethics such as:

Respect for persons (autonomy and dignity)
Beneficence (maximise benefits, minimise harm)
Justice (fair selection of participants)
Confidentiality and privacy
Informed consent (voluntary and informed)

A good exam answer also indicates what these principles mean operationally:

how consent is obtained,
how participants’ rights are protected,
how data is secured,
and how harm is handled.

3.2 Informed consent: more than a signature

Exams often award marks for describing consent processes that reflect real participant situations.

Key elements:

Information: participants understand purpose, procedures, risks, benefits, and alternatives.
Comprehension: language and literacy are considered; questions are asked to confirm understanding.
Voluntariness: refusal should not affect care; participants can withdraw without penalty.
Ongoing consent: participants can revise agreement as the study progresses.

In health research, informed consent must handle:

medical vulnerability (participants may feel pressured due to dependency on healthcare staff),
distress (interviews may trigger sensitive experiences),
and comprehension differences (multilingual settings).

3.3 Handling vulnerable populations ethically

Illness research often intersects vulnerability:

people living with chronic illness,
people experiencing mental distress,
refugees or migrants,
adolescents or guardianship contexts.

Ethical challenges include:

additional safeguards for comprehension,
avoiding coercion,
and ensuring support pathways if interviews cause distress.

A high-scoring exam response includes:

a plan for referral to care if participants experience distress,
guidance on interviewer conduct,
and how researchers prevent confidentiality breaches.

3.4 Confidentiality, anonymity, and data protection

Confidentiality is central in health and illness research because data can include:

HIV status disclosure dynamics,
mental health experiences,
experiences of stigma or discrimination,
illegal or highly sensitive exposures.

A strong response discusses:

Anonymisation in transcripts (removal of identifiers).
Secure storage (password-protected devices; encrypted drives where possible).
Access control (only authorised researchers).
Data retention policies (how long data is stored and when it is destroyed).
Handling quotes in publication (ensuring quotations cannot be traced to individuals).

In exam contexts, also mention the difference between:

anonymity (no one knows who the person is), and
confidentiality (identity is known to the research team but protected).

3.5 Ethical tensions and common dilemmas (and how to argue solutions)

Dilemma 1: Mandatory reporting vs confidentiality

Sometimes researchers encounter disclosure of harm or intent to harm. You should be able to argue a solution framework:

inform participants of limits of confidentiality during consent,
document disclosures carefully,
follow institutional and legal requirements,
and prioritise participant safety.

Dilemma 2: Researcher role conflict

If the researcher is also a clinician or associated with the clinic, participants may feel obliged to participate. Mitigation includes:

clarifying role boundaries,
ensuring recruitment is not done by clinicians who control care access,
and using independent recruitment channels where possible.

Dilemma 3: Compensation and voluntariness

Compensation can be ethical and appropriate, but it must not coerce. An exam answer should:

justify compensation as time/travel costs,
use transparent amounts,
avoid amounts so large that participation becomes financially coercive,
and ensure it is consistent for participants.

3.6 Reflexivity: recognising researcher position and its effects

Reflexivity is not optional in qualitative health research, and even in quantitative studies it matters through assumptions, interpretation, and interpretation of missing data.

A strong honours answer includes a structured reflexivity practice:

Identity and positionality: how the researcher’s social position (gender, class, language proficiency, health background) may affect participant responses.
Power dynamics: how health status, illness narratives, and institutional authority can shape interaction.
Interpretive stance: how the researcher’s assumptions guide coding and theme development.

Practical reflexivity strategies to mention:

reflexive journal during data collection,
memo-writing during analysis,
discussing disagreements with supervisors or co-coders,
documenting how interpretations shift over time.

3.7 Ethics in mixed methods: additional safeguards

Mixed methods add complexity:

multiple data sources (interviews + survey + records) increase re-identification risk.
integration requires careful linking of datasets.

Safeguards:

separate identifiers across datasets,
use coded linkages stored securely,
and ensure that integration methods do not compromise confidentiality.

3.8 Ethics governance in South Africa: institutions and approval flow

South African universities typically require approval via ethics committees before fieldwork begins. While you may be asked to outline the process generally, you can score by describing:

ethics proposal preparation (protocol, instruments, consent forms),
risk assessment,
recruitment procedures,
data management plan,
and reporting of amendments.

A coherent answer states that ethical approval ensures:

participants are protected,
risks are minimized,
and the study meets governance requirements.

Even if the specific committee name differs by institution, your exam response should show you understand the logic of ethics review and the typical components of a protocol.

3.9 Concrete ethical scenario: interviewing patients about ART adherence

Suppose your study involves interviewing patients about ART adherence experiences at a clinic in Johannesburg.

Potential ethical risks:

disclosure of missed doses may lead participants to fear blame or disciplinary action by staff,
participants may become distressed discussing side effects or stigma,
and data can be sensitive enough that small populations could make identification possible.

Ethical safeguards:

Recruitment neutrality: recruitment is done by research staff independent of treatment decision-making.
Consent clarity: explain that care will not be affected; confidentiality is assured.
Interview setting: private space where conversations cannot be overheard.
Distress protocol: pause/stop if distress occurs; provide referral options.
Data anonymisation: remove clinic identifiers in transcripts; store data securely.

Examiners often award marks when you show you can anticipate participant fears and design the ethics plan to address them.

4) Data Collection Methods and Analytic Strategies for Health and Illness Research (Wits)

Research methods in health and illness research are fundamentally about turning questions into data that can answer them—then turning data into credible insights. This section focuses on data collection and analysis, with emphasis on method selection, quality control, and analytic reasoning.

4.1 Choosing a data collection method: fit to question and context

4.1.1 Interviews (qualitative)

Semi-structured interviews are common in illness research because they:

allow exploration of meanings and interpretations,
capture narratives of help-seeking,
and enable probing on sensitive topics.

Key interview design components:

introduction and rapport building,
neutral and non-leading questions,
use of prompts (e.g., “Tell me about the first time you noticed…,” “What did you think it meant?”),
managing distress and boundaries.

A strong exam answer describes how interview guides connect to research objectives. If your question targets stigma, you include prompts on:

disclosure experiences,
perceived consequences,
coping strategies,
and interactions with healthcare workers or family.

4.1.2 Surveys (quantitative)

Surveys provide breadth and allow quantification of patterns.

Survey design considerations:

question clarity and response options,
validated scales where possible,
handling missing responses,
and pilot testing for comprehension.

For health and illness research, surveys might measure:

self-rated health,
social determinants (income, housing stability),
stigma scales,
service access experiences,
and perceived quality of care.

4.1.3 Observations and ethnography (qualitative)

Observation can uncover how interactions occur in practice—waiting rooms, clinic workflows, and informal communication patterns.

However, it raises ethical and practical challenges:

ensuring patient privacy,
obtaining permission from clinic management and participants,
and minimising disruption.

An exam answer should show when observation adds value beyond interviews:

when you want to study interactional dynamics (e.g., provider-patient communication).

4.1.4 Document and secondary data analysis

Secondary sources include:

clinic records,
public health reports,
policy documents,
and administrative datasets.

A strong answer includes that secondary data has constraints:

missing variables,
inconsistent measurement,
and potential data quality issues.
You must show how you would assess and address those limitations.

4.2 Mixed methods integration: not just “use both”

Mixed methods requires integration logic—how and why the methods inform each other.

Common designs:

Concurrent triangulation: collect qualitative and quantitative data at the same time; compare results.
Explanatory sequential: collect quantitative first, then qualitative to explain.
Exploratory sequential: qualitative first to develop or inform quantitative instruments.

In exams, markers often want to see:

a justification for the integration stage,
and a plan for how contradictions will be interpreted.

Example logic:

Quantitative results show a relationship between transport difficulty and missed appointments.
Qualitative interviews explain the mechanisms: transport insecurity interacts with job scheduling and anxiety about being seen.

This coherence helps show methodological maturity.

4.3 Data analysis: qualitative approaches you should master

4.3.1 Thematic analysis

Thematic analysis is widely used and can be flexible. A typical analytic flow:

familiarisation with data,
initial coding,
searching for themes,
reviewing themes,
defining and naming themes,
producing the report.

High-level expectations:

show how codes map to themes,
justify theme development,
and link themes to research questions and conceptual frameworks.

4.3.2 Grounded theory (for theory-building)

Grounded theory typically involves iterative data collection and analysis:

open coding,
constant comparison,
memoing,
selective coding.

In health and illness research, grounded theory fits when:

you want to build a conceptual model of processes (e.g., pathways from symptom recognition to treatment engagement).

A strong exam answer explains how sampling may evolve with emerging theory and how saturation is achieved.

4.3.3 Discourse analysis

Discourse analysis examines how language and meaning structures shape health experiences and social relations. It is particularly useful for:

examining stigma in narratives,
analysing policy discourse and its impact on practice,
understanding how “normal” and “deviant” illness are constructed.

Markers may assess whether you can connect linguistic patterns to social meaning and power.

4.4 Quantitative analysis: core analytic reasoning

Even if your exam question is qualitative, you may be asked to compare quantitative approaches.

Common quantitative steps:

data cleaning and screening,
descriptive statistics,
inferential analysis,
model selection and interpretation,
sensitivity checks and assumptions.

4.4.1 Descriptive statistics

Descriptive analysis includes:

frequencies and percentages,
means and standard deviations,
medians and IQR if data are skewed.

In health and illness research, you should interpret descriptives in context:

Are differences clinically meaningful or just statistically significant?
Do patterns reflect structural inequities?

4.4.2 Regression and associations

Regression models allow adjustment for confounders. A high-scoring answer includes:

identifying plausible confounders from theory,
specifying the outcome and exposure,
discussing assumptions (linearity, independence, multicollinearity),
and interpreting effect sizes in meaningful terms.

4.5 Triangulation: where and how it strengthens claims

Triangulation can mean:

data source triangulation (different participant groups),
method triangulation (interviews + survey),
analyst triangulation (multiple coders),
theory triangulation (different interpretive lenses).

Exams might ask you to evaluate how triangulation affects credibility:

If interview narratives align with survey measures of service access barriers, claims become more robust.
If they diverge, you should interpret divergence as potentially meaningful (e.g., survey measures may miss contextual nuances captured in narratives).

4.6 Analytic rigour: coding audits and quantitative checks

4.6.1 Qualitative coding quality

Strategies include:

codebook development,
inter-coder agreement (where appropriate),
audit trails,
and memoing.

In exams, you should explain that rigour is not about mechanical agreement alone; interpretive clarity and transparency matter.

4.6.2 Quantitative data quality

Quality steps include:

handling missing data (listwise deletion vs imputation, with justification),
checking outliers,
assessing measurement reliability (scale reliability),
and ensuring assumptions are met.

4.7 Concrete analytic plan: linking stigma, access, and illness outcomes

A plausible honours analytic plan could include:

Qualitative component: Thematic analysis of interview data on stigma and care-seeking.
- Output: themes describing mechanisms (e.g., anticipated judgement → delayed clinic visits → interrupted treatment).
Quantitative component: Survey measures of perceived stigma and service access difficulties.
- Output: associations between stigma scores and service access indicators.

Integration:

Use qualitative themes to interpret quantitative associations.
If quantitative association is weak but qualitative narratives are strong, argue that the survey instrument may not capture culturally specific stigma dynamics; propose instrument refinement.

This approach demonstrates that analysis is methodologically reasoned rather than purely descriptive.

5) Writing, Exam Technique, and Full-Length Answer Frameworks for SOCL4028A (Wits)

Honours exams reward not only methodological knowledge but exam-writing strategy: coherence, conceptual accuracy, and explicit justification. This final section provides a practical framework for turning your study into high-scoring exam responses, with institutionally relevant examples grounded in South African health and illness research contexts.

5.1 Understanding what exam questions typically test

While actual exam wording can vary, SOCL4028A-style questions often fall into categories such as:

Compare and justify two research designs for a specific health/illness topic.
Critically discuss sampling strategies for a vulnerable or hard-to-reach population.
Explain ethical issues and propose mitigation strategies.
Describe a data collection instrument and assess validity/reliability.
Outline an analytic method and argue trustworthiness/rigour.
Interpret methodological choices in the context of South African healthcare access and inequality.

High marks usually come from demonstrating:

command of terminology,
ability to link theory to method,
application to a plausible health setting,
and critical reflexivity.

5.2 The “assessment rubric mindset”: marks depend on explicit justification

Markers often reward answers that clearly show:

Which method you would use,
Why you would use it (matching to constructs and research question),
How you would implement it (steps, procedures),
What can go wrong (bias/limitations),
How you would mitigate issues (design/analysis/ethics),
What you would produce (expected outputs and interpretive claims).

Use this structure in almost every exam response. Even if the question asks only “describe,” marks rise when you include justification, risks, and mitigation.

5.3 A full-length answer template you can reuse (and customise)

Use the following exam structure. It is designed to be adaptable to qualitative, quantitative, or mixed-methods questions.

Step-by-step template (adapt per question)

Direct answer to the question (1–2 paragraphs)
Contextualise the health/illness problem (the setting and conceptual constructs)
Choose design(s) and justify paradigm fit
Sampling plan (who, how, and why)
Data collection method(s) (what, where, tools, procedure)
Measurement and quality (validity/reliability or trustworthiness)
Ethics and governance (consent, confidentiality, risk mitigation)
Analysis strategy (coding/quant models + rigour steps)
Limitations and mitigation
Expected contribution (what knowledge the study generates)

This template prevents common exam errors:

giving only definitions,
failing to justify choices,
or ignoring ethics.

5.4 Example exam question A: design and justify a study on illness narratives and stigma

Possible question: “Discuss an appropriate research design to investigate how stigma affects healthcare-seeking among people with a chronic illness in South Africa. Include sampling, data collection, analysis, and ethical considerations.”

A high-scoring answer might proceed as:

Design choice: qualitative semi-structured interviews using a purposive sampling strategy (maximum variation: age, gender, treatment stage).
Paradigm: interpretivist/critical—stigma as socially produced and structurally reproduced.
Sampling: purposive recruitment from clinic or community referral networks; include participants at different stages (new diagnosis vs long-term).
Data collection: interview guide with prompts about symptom interpretation, disclosure decisions, encounters with healthcare workers, and coping strategies.
Analysis: thematic analysis; develop codebook; ensure credibility via peer debriefing and audit trails.
Ethics: informed consent in participants’ preferred language; privacy during interviews; confidentiality regarding sensitive stigma experiences; distress protocol.
Limitations: selection bias if only clinic attendees participate; mitigation through recruiting from multiple community touchpoints.
Contribution: produce a mechanism-based understanding of stigma pathways affecting health and illness outcomes.

If you include a concrete link—like how anticipated judgement leads to delayed care—you demonstrate you understand not just methods but social mechanisms.

5.5 Example exam question B: comparing quantitative and qualitative approaches for ART adherence

Possible question: “Compare quantitative and qualitative methods for studying ART adherence. Which would you choose for an honours study and why?”

A top answer includes:

Quantitative strengths: measurement of adherence patterns, estimating associations between predictors (e.g., perceived access barriers) and adherence outcomes.
Quantitative weaknesses: risk of missing mechanisms; adherence measurement may be inaccurate via self-report.
Qualitative strengths: understanding lived barriers (stigma, side effects, social support, clinic experience), capturing narrative mechanisms.
Qualitative weaknesses: limited generalisability, reliance on interpretive rigour.

A justified honours choice might be mixed methods:

quantitative survey to map prevalence and correlates of missed doses or appointment attendance,
qualitative interviews to explain how barriers operate in everyday life.

Then you explain integration logic:

explanatory sequential design: survey results guide interview sampling to explain patterns.

5.6 Example exam question C: validity and trustworthiness in multilingual settings

Possible question: “How would you ensure validity and reliability (or trustworthiness) when conducting research interviews or surveys in multilingual South African communities?”

A strong answer should explicitly mention:

Translation: back-translation, pilot testing, interviewer training.
Comprehension: checking understanding, using neutral wording, adjusting response categories if misunderstood.
Quality checks: for quant—reliability of scales; for qual—coding audit trails, peer debriefing, reflexive journalling.
Ethical considerations: ensure participants are not pressured if they cannot fully understand.

You get marks by showing you know that language impacts measurement validity and ethical inclusion.

5.7 Writing style that scores: clarity, signposting, and methodological precision

Exam writing should be structured and precise:

Use headings or signposting phrases: “First, … Second, … Finally, …”
Define key terms exactly once (then use them consistently).
Use comparative language clearly: “more appropriate because…” “less suitable due to…”
When discussing limitations, propose mitigations.
Avoid over-general claims. Tie arguments to the question context.

5.8 Common mistakes that lower marks (and what to do instead)

Only definitions, no application
- Fix: always attach definitions to a specific design decision.
Ethics listed but not operationalised
- Fix: describe consent, confidentiality, risk management steps.
Sampling mentioned without justification
- Fix: link sampling to research aim and inclusion/exclusion constraints.
No mention of bias or threats to validity/trustworthiness
- Fix: include at least two threats and mitigation.
Confusing qualitative validity with “no structure”
- Fix: emphasise trustworthiness criteria and audit processes.

5.9 A condensed checklist you can use in the exam (practice)

Before writing your final paragraphs, scan this checklist:

Research question aligned with method?
Paradigm coherent with data type?
Sampling justified and feasible?
Measurement: operationalisation and quality?
Data quality: reliability/validity or trustworthiness?
Ethics: consent, confidentiality, harm mitigation?
Analysis: clear and appropriate?
Limitations: acknowledged with mitigations?

South Africa-focused final synthesis: what examiners expect you to demonstrate

Although SOCL4028A is a Wits Honours course, what matters in answers is your ability to situate research methods within South African realities: health inequalities, multilingual contexts, governance and ethics procedures, and complex healthcare navigation. A top-performing script shows methodological rigour and social understanding:

You define health and illness as socially shaped realities, not purely biological states.
You build research questions that generate design consequences.
You select sampling strategies that reflect population access realities and conceptual needs.
You operationalise constructs carefully and justify measurement choices.
You address bias, confounding, and/or qualitative trustworthiness.
You apply ethics in operational terms: consent, confidentiality, vulnerability, and governance.
You analyse data with transparency, linking outputs back to the original question.
You write with coherent structure, methodological signposting, and critical balance.

A research methods honours exam is fundamentally a test of reasoned methodological judgment. This guide is organised to help you practice that judgment—so that your final answers read like a credible honours researcher’s plan rather than a list of memorised concepts.