Data collection is where your evaluation plan stops being an idea and becomes something a unit could actually do.
You can have a sharp PICOT question, strong evidence and a tidy implementation plan, yet still lose marks if this section only promises to "monitor outcomes". Faculty look for exactly what you will measure, how, by whom and when. This guide shows you how to write that level of detail.
What a Data Collection Plan States
A capstone data collection plan states the measure, the method (chart audit, direct observation, survey or existing dashboard data), the tool, the collector, the time periods and how data will be stored. Collect a baseline before the change and the same measure after it.
Write it in operational language, use validated instruments where they exist, and only use identifiable patient data with written authorization from the site.
What Data Collection Does in an EBP Capstone
In most BSN and MSN evidence-based practice capstones you are not running primary research. You are proposing, and in some programs piloting, a practice change, then explaining how you would know whether it worked.
Data collection serves two purposes, and a credible plan needs both:
- Baseline: showing the size of the current problem before anything changes.
- Follow-up: measuring the same outcome after the intervention, in the same way, to see whether it improved.
If the baseline and follow-up use different definitions, tools or time windows, the comparison falls apart. Decide the measure once and keep it fixed.
Data Collection Methods by Project Type
The right method follows from your outcome. Most capstones use one main method plus a process measure, such as protocol adherence, to show the change was actually delivered.
| Method | Best for | Data produced | What to specify |
|---|---|---|---|
| Chart or EHR audit | Falls, infection rates, medication errors, readmissions, documentation compliance | Counts and rates, often per 1,000 patient-days | Audit period, fields extracted, who audits, access approval |
| Direct observation with a checklist | Hand hygiene, bundle adherence, SBAR use at handoff, risk assessment completion | Compliance rate: compliant observations divided by total, times 100 | Observation window, shifts, observer, the checklist itself |
| Survey or questionnaire | Nurse knowledge, confidence, attitudes, safety culture, patient experience | Likert scores, test scores, pre and post comparisons | Instrument, validity evidence, paper or online, timing, anonymity |
| Existing quality dashboard | Units already tracking NDNQI, Press Ganey or CMS measures | Monthly or quarterly rates the organization already collects | Exact metric, data pull dates, that the data are aggregate |
| Patient outcome tracking | HbA1c, blood pressure, weight, length of stay | Clinical values from the record | QI or IRB determination, data use agreement if identifiable |
Chart Audits: The Most Common Capstone Data Source
A chart audit, also called a record review, is a systematic check of patient records to see whether an event happened, a protocol was followed or an outcome was reached. It suits BSN capstones well because it uses data that already exist and asks nothing of patients.
A complete chart audit description answers six questions:
- Which records? For example, all discharges from a 36-bed medical-surgical unit.
- Which period? A 12-week baseline, then a 12-week follow-up that starts after a settling-in period.
- Which data fields? Fall events from the safety reporting system, injuries from incident reports, risk tool completion from the assessment flowsheet.
- Who collects? The quality coordinator pulls aggregate data; the student audits a set number of random charts each month.
- Which tool? A standard extraction form with defined fields and yes or no responses.
- How is data stored? De-identified, in a password-protected file that only named people can open.
Weak: "Data on falls will be collected and compared before and after the intervention."
Strong: "The unit quality coordinator will extract all fall events and fall-related injuries recorded in the safety event system for the 36-bed medical-surgical unit during a 12-week baseline period. The student will audit 20 randomly selected charts per month for completion of the fall risk assessment in the admission flowsheet, using a standard extraction form. The same measures will be repeated for 12 weeks, beginning four weeks after go-live. De-identified data will be kept in a password-protected spreadsheet."
The strong version names the records, period, fields, collectors, tool and storage, so a reader could repeat it.
Direct Observation and Audit Checklists
Observation suits process measures, where the question is whether staff did something rather than what happened to the patient. Hand hygiene and bundle checks are the classic examples.
Define each item on the checklist so two observers would score the same moment the same way. State how many observations you plan per period, which shifts and days you will cover, and who observes. Rotating times reduces the chance that staff only comply when they know the auditor is around.
Staff behave differently when watched. Name this observation effect as a limitation and, where possible, use unannounced or rotating audits.
Surveys and Validated Instruments
When you measure knowledge, confidence or attitudes, use a validated instrument wherever one exists. Validated means prior peer-reviewed work has shown it is reliable (it measures consistently) and valid (it measures what it claims to).
| What you measure | Common instruments | Notes |
|---|---|---|
| Safety culture | AHRQ Surveys on Patient Safety Culture, hospital version | Free and widely used, so results can be compared |
| Pain | NRS, VAS, FLACC, CPOT | Choose by population: CPOT for ventilated or nonverbal adults, FLACC for young children |
| Burnout | Maslach Burnout Inventory, Oldenburg Burnout Inventory, Mini-Z | The MBI needs a license; check terms for the others |
| Patient experience | HCAHPS | Run by the organization; use existing scores rather than your own survey |
| Topic knowledge | Usually a quiz built from your teaching content | Say clearly that it has not been independently validated |
For any survey, state how it is given (paper or online), when (before teaching, straight after, and perhaps at 30 days), and how responses stay anonymous. If you need to pair pre and post scores, use a self-generated code rather than names.
What Counts as Pilot Data
Some programs ask you to collect a little real data during the capstone. This is a small test of the change before you propose full rollout, not a full study.
- A few weeks of baseline data before the change.
- A small first rollout with one shift or one care team, with early outcome data.
- Pre and post knowledge scores from a single teaching session.
- Compliance audits from the first two weeks of a new protocol.
Pilot data does not need to be statistically conclusive. It shows that your measurement approach works and that the change shows early signs of the expected effect. Report it descriptively with counts, percentages and means, and say that the small sample limits what you can conclude.
How Much Data Is Enough?
A QI capstone proposal rarely needs a formal power calculation. It needs a window long enough to capture a meaningful number of events or observations, which for most unit projects means several weeks to a few months.
- Rare events such as falls or CLABSI need longer windows, because a few weeks may show almost none.
- Compliance audits need enough observations per period that one missed item does not swing the rate wildly.
- Surveys need a realistic response target based on how many eligible staff you have.
Whatever you choose, explain why it is feasible on your unit, and acknowledge that a convenience sample limits how far results generalize.
Putting It on a Timeline
Faculty like to see data collection fitted into the project schedule. A simple table makes the sequence obvious.
| Phase | Data activity | Who |
|---|---|---|
| Weeks 1 to 12 | Baseline outcome and process data | Quality coordinator, student |
| Weeks 13 to 14 | Staff education, pre and post knowledge test | Student, nurse educator |
| Weeks 15 to 18 | Settling-in period, weekly adherence spot checks | Charge nurses |
| Weeks 19 to 30 | Follow-up outcome and process data, same tools | Quality coordinator, student |
| Weeks 31 to 32 | Analysis and report to unit leaders | Student |
Protecting Patient Data
Even in a QI project, extracting names, record numbers or other identifiers needs explicit permission from the facility. That usually comes through the quality department, a data use agreement or a QI determination from the IRB.
De-identified aggregate data, such as total falls in a month or an overall compliance rate, generally carries fewer requirements. Confirm your site's policy before you collect anything. Describing identifiable data collection without documented authorization is both an integrity problem and a privacy risk. See the IRB approval guide for the QI versus research question.
How NurseCapstone Helps with Data Collection Plans
Send your PICOT question, your setting, the outcomes you care about and your rubric. A writer matched to your nursing level drafts a model data collection section that names each measure, method, tool, collector, timeframe and storage step, and fits it to your evaluation plan.
Use it as a reference for your own section. Your writer will not invent data or results. Free revisions within the scope of your original request are included.
Want Your Data Collection and Evaluation Sections Done Together?
Send your PICOT question, setting and rubric. You get a model section that names every measure, tool, collector and timeframe.
Start My Evaluation SectionFree revisions · Full refund if late · Written from scratch for your order
Frequently Asked Questions
Write the section in the future tense and describe exactly what would be collected, from which records, by whom and over what period. Faculty grade whether the plan is specific and feasible, not whether you ran it.
Usually not for a QI proposal. Choose a collection window long enough to capture a meaningful number of events or observations, explain why it is feasible, and note that a convenience sample limits generalization.
A tool that published research has shown measures consistently and measures what it claims to. Using one makes your results more credible than a home-made questionnaire.
Yes, if no validated tool fits your topic. Build it from your teaching content and state plainly that it has not been independently validated.
Yes. The measure, definition, tool and collection method must stay the same, or the before and after figures cannot be compared.
It depends on your site and program. Many QI projects get a QI determination rather than full review, but any use of identifiable patient data needs written authorization from the facility.
Outcome data shows what happened to patients, such as the fall rate. Process data shows whether staff delivered the change, such as the percentage of completed risk assessments.
Descriptively, with counts, percentages and means in a short table, and with a clear note that the small sample limits conclusions.