Nursing Dissertation Guide

Secondary Research Nursing Dissertation Guide: Designs, Data and Analysis

Plan a secondary research nursing dissertation using evidence synthesis, existing datasets or documentary analysis with clear methods, ethics and reporting.

Written by Moses Updated August 25, 2026 10 min read
Secondary Research Nursing Dissertation Guide: Designs, Data and Analysis

A secondary research nursing dissertation uses existing studies, datasets or documents to answer a new nursing research question without recruiting new participants. It can be methodologically rigorous, but only when the source, design, analysis and conclusions are aligned. Secondary research is not simply an easier alternative to primary research and it should not become a sequence of article summaries.

Three common routes are evidence synthesis, secondary analysis of an existing dataset, and documentary or policy analysis. Each route answers a different type of question and has different requirements for searching, governance, appraisal, analysis and reporting.

Choose the secondary design before analysing anything

  • Evidence synthesis: asks what existing research shows about an intervention, experience, association or service issue.
  • Existing-dataset analysis: asks what patterns or associations can be examined in previously collected numerical or qualitative data.
  • Documentary or policy analysis: asks how a nursing issue is framed, governed or represented across documents.
  • The available source must fit the research question rather than forcing the question to fit whatever data happen to be available.
  • Ethics and data-governance requirements still apply where relevant.
  • Conclusions must remain within what the underlying design can establish.

Start with a question that existing material can answer

Begin with the knowledge gap, population and intended contribution, then confirm that appropriate evidence or data actually exist. Choosing “secondary research” only because recruitment is difficult can produce a weak dissertation if the available material does not answer the question.

For evidence synthesis, PICO may suit intervention questions while PICo or SPIDER may suit experience questions. For quantitative secondary analysis, define the population, predictor or exposure, outcome and time period. For documentary analysis, specify the policy issue, jurisdiction, document type and analytical lens.

A useful feasibility statement identifies the proposed source, access conditions, relevant variables or evidence types, likely missing information and analytical method. If any one of these remains uncertain, resolve it before finalising the title.

Distinguish the three main secondary-research routes

Route Best suited to Typical material Main risk
Evidence synthesis What is known about an intervention, experience, association or service issue? Published qualitative, quantitative or mixed-methods studies Selective searching or weak synthesis
Existing-dataset analysis What patterns, associations or changes appear in previously collected data? Survey files, cohort data, registries, service data or archived interviews Variables or sampling that do not fit the new question
Documentary/policy analysis How is a nursing problem framed, governed or addressed? Policies, standards, strategies, consultations or implementation reports Treating official documents as neutral evidence of implementation

For example, a review of how hospital-at-home services affect nursing workload is not the same study as analysing a workforce dataset. A policy analysis of hospital-at-home guidance answers another question again. The design should follow the decision you want to inform.

Check whether the source is fit for purpose

Secondary material was usually created for another purpose. Published studies reflect their own eligibility criteria; service datasets reflect operational coding; archived interviews contain the questions originally asked; policies reflect organisational priorities and intended audiences.

For each source, examine provenance: who created it, why, when, how, for whom and under what definitions. A dataset may contain “readmission” but define it differently from the dissertation. An archive may mention family involvement only incidentally. A policy may use “community nursing” without distinguishing professional roles.

Create a source-to-question matrix listing each concept in the question, its operational definition, the available variable or text, missing information and the effect on interpretation. If a core concept cannot be represented defensibly, narrow the question or choose another source.

Evidence synthesis: label the review accurately

A systematic review uses predefined searching, selection, appraisal and synthesis to answer a focused question. A scoping review maps the range and characteristics of evidence, often for a broader question. Some university programmes use structured or integrative literature reviews with different expectations.

PRISMA 2020 is a reporting guideline for systematic reviews and meta-analyses; following its checklist does not by itself make a review systematic (Page et al., 2021). The method still needs appropriate eligibility criteria, transparent searching, screening and synthesis.

Use our systematic review search-strategy guide when the project requires reproducible database searching.

Quantitative secondary-data analysis: understand the dataset before the test

Secondary quantitative analysis may use survey, cohort, registry or service data to estimate prevalence, compare groups, assess trends or explore associations. The design remains observational unless the underlying data came from an experiment.

Before selecting statistics, understand the sampling process, weighting, clustering, missing data, coding conventions, measurement periods and any changes in data collection over time. Repeated observations are not independent, and routine coding changes can sometimes look like clinical change.

STROBE supports transparent reporting of observational studies, while the RECORD extension is relevant to studies using routinely collected health data.

Qualitative secondary analysis: preserve context

Archived interviews, focus groups, diaries or field notes can be reanalysed for a new question, but the new question must remain close enough to what participants originally had the opportunity to discuss.

Consider what contextual information is missing, how the original interviewer shaped the accounts and whether anonymisation removed details important to interpretation. Reflexivity still matters because the secondary analyst brings a new question and interpretive position to material generated in another context.

Documentary and policy analysis: do not confuse policy with implementation

Documentary analysis examines content, framing, assumptions, omissions or development across a defined set of documents. State the jurisdictions, organisations, date range, document types and inclusion criteria.

Explain whether the method uses qualitative content analysis, thematic analysis, framework analysis, discourse analysis or another defensible approach. Most importantly, distinguish policy intention from implementation evidence. A strategy that recommends a nursing intervention does not prove that the intervention was adopted or effective.

Plan searching, access and selection transparently

Evidence reviews require documented database and supplementary searching. Dataset projects require catalogue searching, access arrangements and detailed documentation. Policy projects require a reproducible route through organisational websites, archives or repositories.

Record the source, search date, search terms or filters, results and selection decision. Save document versions where online content may change. For datasets, read the codebook, questionnaire, sampling documentation, licence and user guide before describing the methodology.

Define inclusion and exclusion criteria before examining results in depth. If criteria change, document what changed and why.

Secondary research still has ethics and governance requirements

Using existing material does not automatically remove ethical responsibilities. A review of public journal articles may not require research-ethics review, while person-level health data, restricted survey files or archived qualitative accounts may require formal approval, data-use agreements or both.

The UK Data Service emphasises legal, ethical and contractual responsibilities when data are reused. For health and care information, establish who controls the data, the lawful access route, whether consent covered reuse and whether the material is anonymous, pseudonymous or identifiable.

The Information Commissioner’s Office data-protection principles include lawfulness, fairness, transparency, purpose limitation, data minimisation, accuracy, storage limitation, security and accountability. UK health projects should also use current Health Research Authority guidance to determine which approvals or decisions apply.

Write an analysis plan before inspecting the results

An analysis plan reduces the temptation to change methods after seeing interesting findings. For a review, predefine appraisal and synthesis. For quantitative data, specify variables, coding, descriptive analysis, confounders, assumptions, missing-data handling and planned sensitivity analyses. For qualitative or documentary data, explain coding, theme development, reflexivity and the steps used to support analytical rigour.

Avoid a long list of statistical tests with no conceptual model. Software is a tool, not the methodology. SPSS, R or NVivo cannot compensate for a poorly defined question or a weak source.

Structure the dissertation around the chosen design

Section Main task
Introduction Define the nursing problem, gap, question, aim and why secondary research is appropriate
Background/literature context Position the topic without pre-empting the analysed findings
Methodology Name and justify the design, source, access, selection, analysis, ethics and limitations
Results Present the synthesis, dataset findings or document patterns transparently
Discussion Interpret the findings against relevant literature and source limitations
Conclusion Answer the research question without introducing new evidence or overstating causation

Worked example: continuity and unplanned hospital use

Consider a dissertation asking whether continuity with a community nursing team is associated with unplanned hospital use among adults receiving home-based care. A suitable anonymised service dataset might contain referral dates, nursing contacts, patient characteristics and hospital-use outcomes.

“Continuity” must be operationalised rather than assumed. It could be represented by the proportion of contacts delivered by the most frequently seen team or another justified continuity measure. Hospital use needs a defined observation period. Potential confounders might include age, multimorbidity, previous admissions, deprivation, referral reason and intensity of nursing input.

Before analysis, the student should examine missingness, impossible dates, duplicate episodes and coding changes, then explain how the analytic sample was constructed.

If greater continuity is associated with fewer unplanned admissions, the conclusion remains observational. Sicker patients may experience both more fragmented care and higher admission risk, while staffing, geography and social factors may influence both variables. The result can support further evaluation but does not automatically establish causation.

Common secondary-research mistakes

  • Calling every literature review systematic.
  • Choosing a dataset before defining a coherent question.
  • Ignoring why and how the source was originally created.
  • Assuming publicly accessible data are unrestricted for research use.
  • Treating software as the methodology.
  • Hiding missing data or missing contextual information.
  • Writing results source by source rather than synthesising around the question.
  • Using PRISMA for a dataset project simply because it is a familiar research guideline.
  • Making causal claims from observational secondary data.
  • Assuming official policy wording demonstrates implementation.

Final secondary-research checks

  • The design is named accurately.
  • The question can be answered with the selected material.
  • Source provenance is explained.
  • Access and licence conditions are clear.
  • Selection decisions are transparent.
  • The analysis plan aligns with the objectives.
  • Ethics, privacy and disclosure risks are addressed.
  • Missingness or missing context is evaluated.
  • Quality appraisal influences interpretation where relevant.
  • Results answer the question rather than merely describe sources.
  • Limitations explain how bias could affect the result.
  • Recommendations stay within what the design can support.

Frequently asked questions

Is every literature review secondary research?

Literature reviews use existing publications, but systematic, scoping, integrative and narrative approaches have different purposes and levels of reproducibility. Use the label your methodology actually supports.

Do I need ethics approval for secondary data?

Possibly. Requirements depend on identifiability, sensitivity, consent, licence terms, access arrangements and institutional policy. Obtain a formal decision rather than assuming approval is unnecessary.

Can existing NHS or hospital data be used for a student dissertation?

Only through an authorised route with the required governance and approvals. Employment or placement access does not automatically permit research use of service data.

Can secondary data prove cause and effect?

Usually not when the underlying data are observational. Confounding, selection, measurement and temporal ambiguity may limit causal interpretation.

Should PRISMA be used for a secondary dataset dissertation?

No. PRISMA is designed for systematic-review reporting. Use guidance appropriate to the actual study design.

Related Nursing Guides

Conclusion

A strong secondary research nursing dissertation begins with a question that existing material can genuinely answer. Its credibility depends on source provenance, an accurately named design, appropriate governance, a planned analysis and conclusions that remain within the evidence.

If you need help deciding whether your project is a review, dataset analysis or documentary study, use our nursing dissertation services or contact page.

References

Related articles

Browse all