Your evaluation plan answers the question every reviewer is asking: how will you know whether this worked?
"Outcomes will be monitored" earns little credit. A strong plan names the metric, the tool, the baseline, the data source, the timing and the target. This guide walks through each part, using one fall prevention project as a running example.
The Layers of an Evaluation Plan
A capstone evaluation plan has at least two layers: process evaluation (was the change delivered as designed?) and outcome evaluation (did the PICOT outcome change?). MSN and DNP plans usually add impact evaluation for wider system effects.
Write each measure as a SMART outcome with a data source, compare it with a baseline collected the same way, and set measurement points on a timeline long enough for the outcome to move.
The Three Types of Evaluation
Rubrics often ask for "an evaluation plan" without saying which type. Knowing the three types stops you from mixing them up or missing one your rubric expects.
| Type | What it measures | Question answered | When to include |
|---|---|---|---|
| Process | Fidelity, reach and dose of the intervention | Was the protocol followed, for the right patients, as often as planned? | Always; it explains whether weak results reflect a weak intervention or weak delivery |
| Outcome | Change in the targeted clinical or nursing outcome | Did falls fall, compliance rise or knowledge improve? | Always; it answers your PICOT |
| Impact | Wider effects on the organization, costs or population | Did costs drop, did the practice spread, did culture shift? | Optional at BSN; usually expected at MSN and DNP |
A complete plan covers process and outcome at minimum. Adding a short impact section shows system-level thinking at graduate level.
Building a Logic Model
A logic model maps the chain from what you invest, through what you do, to what results. Some programs want it as a table or diagram in an appendix; others just expect the plan to follow its logic.
| Inputs | Activities | Outputs | Short-term outcomes | Long-term outcomes |
|---|---|---|---|---|
| Nurse education time, a free validated risk tool, an EHR flowsheet update, pocket cards, charge nurse audit time | Staff education, EHR set-up, protocol launch, risk assessment each shift, weekly audits | Percentage of nurses trained, assessment completion rate, care plans created | Completion rate at target, higher knowledge scores, more high-risk patients flagged | Lower fall rate per 1,000 patient-days, fewer injurious falls, protocol in unit policy |
Read it left to right and every column should explain the next. If an outcome has no activity feeding it, either add the activity or drop the outcome.
Writing SMART Outcomes
When a rubric asks for SMART outcomes, each one must pass five tests.
- Specific: names exactly what changes, such as the fall rate rather than "patient safety".
- Measurable: a number or percentage you can verify from a named data source.
- Achievable: realistic for your intervention and timeframe, informed by what similar studies report.
- Relevant: the same outcome your PICOT asks about.
- Time-bound: measured at a stated point, such as 12 weeks after go-live.
| Weak | SMART |
|---|---|
| Falls will decrease | The unit fall rate will fall from a baseline of 3.8 to 3.0 or fewer per 1,000 patient-days at 12 weeks after implementation, measured from the safety event reporting system |
| Nurses will use the tool | Risk assessment completion will reach at least 90% for patients aged 65 and over within four weeks of launch, measured by monthly chart audit |
| Patient safety will improve | The fall-with-injury rate will be lower in the 12 weeks after launch than in the 12-week baseline, measured from incident report classifications |
| Knowledge will increase | Mean score on a 10-item knowledge test will reach at least 80% immediately after the education session, compared with the pre-test |
Baseline figures in these examples are illustrative. Use your own unit's data, and base your targets on what the evidence suggests is realistic.
Matching Each Measure to a Data Source
Every measure needs a named source and a named collector. A short table keeps this tidy and makes gaps obvious.
| Measure | Type | Data source | Collected by |
|---|---|---|---|
| Nurses trained | Process | Attendance records | Nurse educator |
| Assessment completion | Process | Chart audit of the assessment flowsheet | Student, charge nurses |
| Fall rate per 1,000 patient-days | Outcome | Safety event reports and census data | Quality coordinator |
| Fall-with-injury rate | Outcome | Incident report classification | Quality coordinator |
| Fall-related costs or spread to other units | Impact | Finance estimates, leadership reports | Nurse manager |
For more on audits, observation tools and surveys, see the data collection guide.
The Evaluation Timeline
State when each measurement happens. A timeline stops you from measuring too early, before the change has had time to work, or too late, after the pilot has ended.
| Timepoint | What to measure | Source |
|---|---|---|
| Before launch (weeks 1 to 4, plus earlier data) | Baseline fall rate, baseline completion, pre-education knowledge | EHR audit, event log, pre-test |
| Launch (week 5) | Staff trained, EHR template live | Attendance, IT confirmation |
| Early implementation (weeks 5 to 8) | Completion rate, early barriers | Weekly audit, staff feedback |
| Full implementation (weeks 9 to 16) | Completion rate, fall rate, injury rate | Fortnightly audit, event log |
| After implementation (week 17 on) | Fall rate compared with baseline, knowledge retention, staff views | EHR audit, repeat test, short survey |
Handling Rare Outcomes Such as Falls
Falls, pressure injuries and infections happen too rarely on one unit to show a reliable change over a few weeks. Measuring the fall rate four weeks after launch may give you only a handful of events.
Plan the primary outcome at eight to twelve weeks or more, and treat your process measure as the leading indicator during the pilot. Say plainly that a longer observation window would be needed to confirm a change in the outcome.
Sample wording. "Given the low baseline incidence of falls on this 32-bed unit, the 12-week pilot is unlikely to produce enough events for a robust statistical comparison. Assessment completion (target 90% or more) will therefore serve as the leading indicator of protocol fidelity. The fall rate per 1,000 patient-days will still be tracked as the primary outcome, with the recognition that six to twelve months of follow-up would be needed to confirm a sustained change."
This shows you understand the limitation and have designed around it honestly.
How You Will Interpret the Results
Say briefly how you will compare baseline and follow-up. Most BSN plans use descriptive statistics: rates, percentages and means shown in a table or run chart. A run chart that plots the measure weekly, with the launch date marked, is often the clearest way to show change over time.
Graduate plans may add a simple inferential test, such as a paired t-test for pre and post knowledge scores. State in advance what would count as success, so you are not deciding after you see the numbers.
Common Evaluation Plan Mistakes
- Outcome only, no process measure. You cannot tell whether a flat result means the protocol failed or was never used.
- Outcomes that do not match the PICOT. Measure what your question asks about.
- Unrealistic targets. Large reductions in a few weeks are not credible; ground targets in the evidence.
- No data source or collector. Every measure needs both.
- Changing definitions between baseline and follow-up. Keep tools and definitions fixed.
How NurseCapstone Helps with Evaluation Plans
Send your PICOT question, intervention, setting and rubric, plus any baseline data your site has given you. A writer matched to your level drafts a model evaluation plan with process, outcome and, if needed, impact measures, SMART targets, data sources, a logic model and a timeline. It links to your implementation plan so the two sections agree.
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 Evaluation Plan Done Properly?
Send your PICOT question, intervention and rubric. You get a model plan with process and outcome measures, data sources, targets and a timeline.
Start My Evaluation PlanFree revisions · Full refund if late · Written from scratch for your order
Frequently Asked Questions
Follow your rubric. If it says logic model, give a table or diagram; if not, a clear prose plan is enough, though a simple logic model table in an appendix rarely hurts.
Process evaluation checks whether the change was delivered as planned. Outcome evaluation checks whether the result your PICOT targets actually changed.
Usually not at BSN level, but MSN and DNP programs often expect it to show system-level effects such as cost, spread or culture change.
Use a process measure as your leading indicator, plan a longer measurement window, and state clearly that more follow-up would be needed to confirm a change.
Look at what similar studies in your evidence review reported and at your unit's baseline. Set a modest, defensible target and explain where it came from.
Measure process measures early and often. Measure clinical outcomes after the change has had time to work, typically eight to twelve weeks or longer.
Usually descriptive statistics are enough: rates, percentages and means, often shown in a table or run chart. Check your rubric for any required tests.