Results say what the data showed. Discussion says what you make of it. Keeping those two jobs separate, and being honest about small samples and missed targets, is what makes a scholarly project believable. This guide shows how to present, interpret and qualify.
The results section reports findings without opinion: numbers, themes, and what they show, in the order of your aims. The discussion interprets those findings: what they mean, how they compare with the literature, what they suggest for practice, and where they fall short. Some programs combine the two. Even then, keep reporting and interpreting in separate paragraphs so the reader always knows which is which.
Begin with a brief description of the sample or participants: how many took part, how many were lost and why. Then follow your aims one by one, using the same labels as your methods. For each, state the result first, then point to the table or figure, then add only a short comment. Do not narrate every cell of a table; highlight what matters.
Report more than a p-value. Give the size of the change (a difference in proportions or means), a measure of spread or a confidence interval, and the test used. A small improvement with a wide interval is a different message from a large one with a narrow interval. In improvement projects, describe run-chart patterns such as shifts and trends, not just the before-and-after means. For qualitative work, present themes with a short supporting extract each. Report unexpected results, negative findings and missing data with the same care as the successes.
Use a table when readers need exact values and a figure when they need to see a pattern, such as change over time. Each should be understandable without the text: a clear title, labelled axes or columns, units and sample sizes. Do not repeat the same data in text, a table and a figure.
Formatting follows your style guide. For APA 7 tables and figures, see the APA tables and figures page and the official APA Style site. Keep patient and organizational identifiers out of every table and every figure, including small cell counts that could identify individuals.
Share your findings and rubric. The ProNursingScholars team can help you organize the results and draft the discussion. Details stay confidential.
Open with a one- or two-sentence summary of the main finding. Then take each key result in turn, relating it to the literature you reviewed: consistent, different, or new. Offer explanations, and mark them clearly as possible rather than proven. If a framework guided the project, say what the results suggest about it. Follow with implications for practice, policy, education or further work, and note how the change might be sustained.
Limitations belong here, stated plainly: small or convenience sample, single site, no control group, short follow-up, self-reported measures, and any events during the project that could have affected results. Say what each limitation means for how far the findings can be trusted, not only that it exists. Finish with a conclusion that answers your aim directly. The DNP outcomes evaluation help page explains how evaluation is usually framed.
Fictional project: a delirium screening prompt on an inpatient unit. All numbers are invented.
| Weak | Strong |
|---|---|
| The project was a success. Screening improved a lot and everyone agreed it was better. This proves the prompt works and should be used in all hospitals. | Screening completion rose from 52% of eligible shifts in the baseline period to 74% in the implementation period (a difference of 22 percentage points), with a run-chart shift beginning in week 3. The target of 85% was not met. Completion was lowest on night shifts, possibly because of staffing, though this was not tested. With one unit and a 10-week period, the findings cannot show that screening prompted better outcomes or that the result would hold elsewhere. |
The strong version reports a number, a comparison, a missed target and a limit on what can be claimed. The weak one claims proof from an opinion.
Report them plainly and discuss possible reasons, including sample size and implementation problems. A well-analyzed null result is legitimate scholarship.
If your template allows it, yes. Keep findings and interpretation in different paragraphs.
Long enough to cover the real threats to validity, usually a paragraph or two, each limitation tied to its consequence.
It depends on the design. SQUIRE suits many improvement projects, and the EQUATOR Network lists the others.