Research design is the blueprint for how evidence is gathered and analyzed to answer a question.
In a nursing capstone it matters at two levels: the design of each study in your literature review, and the design of your own project's evaluation. Students who mix up the two, or misname a design, lose marks on methodology criteria that are otherwise straightforward.
The Main Research Designs at a Glance
Quantitative designs (systematic reviews, RCTs, quasi-experimental, cohort, case-control and descriptive studies) measure and test relationships. Qualitative designs (phenomenology, grounded theory, ethnography, qualitative description) explore experience and meaning, and mixed methods combine both.
Most BSN and MSN capstones evaluate a practice change with a quasi-experimental pre-post design, which you should name and describe as a limitation-aware quality improvement evaluation, not as a randomized trial.
Two Design Questions in Every Capstone
| Level | The question | Where it appears |
|---|---|---|
| The evidence you review | What design did each study use, and how strong is it? | Literature review and evidence table |
| Your own project | How will you know whether your practice change worked? | Methods, evaluation plan and limitations |
For the first, you identify each study's design and its level of evidence, then use that to judge how strong the case for your intervention is. For the second, you are usually designing a quality improvement or evidence-based practice evaluation, not new research.
Quantitative Designs
Quantitative designs use numerical data to describe, measure or test relationships between variables. Evidence hierarchies such as Melnyk and Fineout-Overholt's place them at most of the higher levels, especially systematic reviews and randomized trials.
| Design | Key features | Usual level (Melnyk) | Typical use in a capstone |
|---|---|---|---|
| Systematic review or meta-analysis of RCTs | Synthesizes all trials on a question; meta-analysis pools effect sizes statistically | I | Lead source in your review if one exists |
| Randomized controlled trial | Random allocation to intervention or control; strongest single-study design | II | Core primary evidence |
| Quasi-experimental | Intervention without randomization: pre-post, non-equivalent groups, interrupted time series | III | Common in nursing QI literature and in your own evaluation |
| Cohort or case-control | Observational; follows groups forward or looks back from outcomes | IV | Risk factors, such as for falls or infections |
| Descriptive, correlational or cross-sectional | Describes variables or relationships at one point, without intervention | VI | Needs assessments, baseline data, staff surveys |
Level V covers systematic reviews of descriptive and qualitative studies, and Level VII covers expert opinion. If your program uses the Johns Hopkins model instead, its levels run from I to V with separate quality ratings, so label studies with the hierarchy your rubric names.
Qualitative Designs
Qualitative designs produce narrative data about experience, meaning and perspective. They answer "what is it like" and "how does this happen" questions rather than "does X cause Y".
| Design | Question it answers | Data | Capstone example |
|---|---|---|---|
| Phenomenology | What is the lived experience of this? | In-depth interviews | Nurses' experience of burnout; families' experience of ICU |
| Grounded theory | What theory explains this social process? | Interviews, observation, constant comparison | How nurses develop clinical judgment |
| Ethnography | What are the cultural norms in this setting? | Extended observation and interviews | Unit safety culture |
| Qualitative description or content analysis | What themes appear in accounts or documents? | Interviews, open survey answers, records | Barriers named in incident reports |
A single qualitative study sits at Level VI in the Melnyk hierarchy. That reflects the kind of question it answers, not poor quality.
How to Identify a Study's Design When You Appraise It
Authors do not always label their design clearly, and some label it wrongly. Read the methods section and work through these questions in order.
- Did the researchers assign an intervention? If no, the study is observational (cohort, case-control, cross-sectional or descriptive) or qualitative.
- If they assigned it, was allocation random? Random allocation means an RCT. Without it, the study is quasi-experimental.
- If observational, which direction? Following people from exposure to outcome suggests a cohort study; starting from the outcome and looking back suggests case-control; one point in time suggests cross-sectional.
- Is the data numbers or words? Interviews, focus groups and themes signal a qualitative design.
- Does it synthesize other studies? Then it is a review: check for a systematic search and appraisal before calling it systematic.
Evidence table entry (illustrative): Design: quasi-experimental, pre-post without control group. Level: III. Sample: 112 nurses on four medical units. Key finding: hand hygiene compliance rose after visible feedback boards. Limitation: no comparison unit, so other changes may explain the result.
Mixed Methods Designs
Mixed methods studies combine quantitative and qualitative data in one project. Three configurations appear most often in nursing research.
- Explanatory sequential: numbers first, then interviews to explain the results. Common in intervention evaluations.
- Exploratory sequential: interviews first, then a quantitative phase that tests what emerged, often a new instrument.
- Convergent: both data types collected at the same time and merged at analysis.
A small capstone can borrow this logic. For example, a pre-post audit of compliance combined with a short open-ended staff survey explains both whether and why the change worked.
Your Own Project's Evaluation Design
Most BSN and MSN evidence-based practice capstones use a quasi-experimental pre-post design. You collect baseline data, implement the practice change, collect the same data afterwards and compare.
There is no randomization and usually no control group. That is acceptable for a quality improvement project, as long as you say so and treat it as a limitation on causal claims.
"This project will use a quasi-experimental pre-post design to evaluate the effect of a STRATIFY-based fall risk protocol on fall rates and fall-related injuries on a 36-bed medical-surgical unit. Baseline data will be drawn from the unit's fall log for the 12 weeks before implementation, and post-intervention data for the 12 weeks after full rollout.
Rates will be compared using descriptive statistics (falls per 1,000 patient-days). No randomization or control group will be used; this limits causal inference and is acknowledged as a design limitation."
The example is specific, uses correct terminology and is honest about its limits. Those are exactly the features methodology rubrics reward.
Ways to Strengthen a Pre-Post Design
You cannot turn a unit project into a trial, but small choices make your conclusions more credible.
- Use equal time periods before and after, so seasonal or staffing changes affect both.
- Track several points rather than one before and one after, for example monthly rates shown on a run chart.
- Add a process measure, such as protocol compliance, alongside the outcome.
- Add a balancing measure that would show unintended harm, such as restraint use in a fall project.
- Note other changes on the unit during the project that could explain results.
Matching the Design to Your Question
| If your question asks... | Look for, or use... |
|---|---|
| Does an intervention improve an outcome? | Systematic reviews and RCTs in the literature; pre-post evaluation for your project |
| What increases the risk of an outcome? | Cohort and case-control studies |
| How common is a problem on my unit? | Descriptive or cross-sectional data, such as an audit |
| Why do staff not follow a protocol? | Qualitative studies; a short staff survey or focus group |
| Did it work, and why? | Mixed methods logic: outcome data plus staff feedback |
QI, Research and IRB Review
Most BSN and many MSN capstones are quality improvement projects rather than human subjects research. They evaluate a practice change in one setting instead of producing generalizable knowledge.
Many such projects receive a QI determination or exemption rather than full IRB review, but that decision belongs to your school and clinical site, not to you. Always address the determination in your plan, even when the answer is that the project is QI.
How NurseCapstone Helps with Research Design
Send your PICOT question, project plan, setting details and rubric. A writer matched to your level drafts a model methods and evaluation section that names your design correctly, explains measures and data collection and states limitations honestly.
Use it to shape your own section. Writers never invent data or results, and free revisions within the scope of your original request are included.
Want Your Design Described Correctly?
Send your PICOT question, project plan and rubric. Your writer drafts a model methods and evaluation section with accurate design terminology.
Start My MethodologyFree revisions · Full refund if late · Written from scratch for your order
Frequently Asked Questions
Describe your project's evaluation design, which is usually a quasi-experimental pre-post design. Name it, then explain your baseline measurement, intervention period and post-measurement.
No. It has no random allocation and usually no control group, so describe it as quasi-experimental and note the limit on causal claims.
Yes, when they answer questions numbers cannot, such as why staff skip a protocol. Label them at the correct evidence level and use them to inform implementation, not to prove effectiveness.
A cohort study follows people with and without an exposure forward to see who develops the outcome. A case-control study starts with people who have the outcome and looks back at exposures.
Use the one your rubric names. Melnyk and Fineout-Overholt's seven levels and the Johns Hopkins model are both common in nursing programs.
It depends on your school and site. Many QI projects receive a QI determination rather than full review, but you must follow the formal process and document the outcome.
Yes, on a small scale. Pairing outcome data with brief staff feedback is a practical way to explain whether and why a change worked.
A measure that would show an unintended harm from your change, such as increased restraint use during a fall prevention project.
Check how participants were assigned. If allocation was random, it is an RCT; if the researchers assigned the intervention without randomization, it is quasi-experimental.