| Course | NR 714 Application of Analytic Methods II |
|---|---|
| Week | 4 |
| Paper type | Biostatistics write-up |
| Length | About 1,110 words, 5 pages |
| Format | APA 7 student paper |
| School | Chamberlain University |
| Program | Doctor of Nursing Practice (DNP) |
| Updated | October 2026 |
Free sample paper for NR 714 Week 4
Did Securement Practice Change After Training? A Paired Analysis of 28 Nurses' Bedside Audit Scores on an Adult Medicine Unit
Student Name
College of Nursing, Chamberlain University
NR 714: Application of Analytic Methods II
Instructor Name
Month Day, Year
Did Securement Practice Change After Training? A Paired Analysis of 28 Nurses' Bedside Audit Scores on an Adult Medicine Unit
The Question and the Data
Before the securement bundle could reduce catheter failure on 4 North, nurses had to apply it. This analysis asks whether bedside securement practice changed between August 2026, before bundle training, and October 2026, a month after it. In each month the project lead audited four peripheral catheters cared for by each registered nurse on the unit's roster, using an eight-item checklist: dressing intact with edges sealed, bordered dressing in use, securement device on the hub, insertion site visible, dressing dated, extension set looped and secured, no nonsterile tape over the site, and a site score documented that shift. Each catheter scored from 0 to 8, and each nurse's score for the month was the mean of their four catheters.
Thirty-one nurses were audited in August. Three left the unit before October, so 28 nurses have both scores and form the analysis set. The three who left had August scores of 4.50, 5.25 and 6.00, close to the group's median, so their loss is unlikely to have shifted the result in either direction.
Variables and Their Level
The variable is a nurse's mean audit score, which can take values from 0 to 8 in steps of 0.25. It is a bounded discrete score built from counts of items present, so it behaves as an interval measure for differences between months but has a hard ceiling at 8. The grouping variable is time, August or October, and it is not an independent grouping: the same 28 nurses appear in both months.
Paired, Because the Nurses Repeat
Because each nurse contributes a score in both months, the scores are paired. Treating August and October as two independent groups would ignore that a nurse who scored high before is likely to score high after, and would throw away the information that makes the comparison efficient. The analysis therefore works on the 28 within-nurse differences, October minus August (Polit & Beck, 2021).
Assumptions Reported, Not Assumed
A paired t-test assumes the differences are roughly normally distributed. They were not. The Shapiro-Wilk test on the 28 differences gave W = 0.91, p = .02, and the histogram showed a long left tail: most nurses improved by 1.5 to 2.5 points, while a few who were already near the ceiling in August could improve very little. Nine nurses scored 7.75 or 8.00 in October, which compresses the upper end. The primary test is therefore the Wilcoxon signed-rank test, with the Hodges-Lehmann estimate of the median difference and its 95% confidence interval. The paired t-test is reported alongside for comparison, because the two disagreeing would itself be worth knowing.
Results
Table 1
Nurses' Bedside Securement Audit Scores Before and After Training, 4 North
| Measure | August 2026 | October 2026 | Difference |
|---|---|---|---|
| Median score (interquartile range) | 5.25 (4.50 to 6.00) | 7.25 (6.75 to 7.75) | 2.00 (1.25 to 2.50) |
| Nurses improved, unchanged, lower | 25, 2, 1 | ||
| Hodges-Lehmann median difference (95% CI) | 1.88 (1.50 to 2.25) | ||
| Mean difference, paired t (95% CI) | 1.79 (1.45 to 2.13) |
Note. Note. n = 28 nurses with both audits. Scores are the mean of four catheters each, on an eight-item checklist.
Twenty-five of 28 nurses scored higher in October, two were unchanged and one scored a quarter point lower. The Wilcoxon signed-rank test, on the 26 nonzero differences, gave V = 350, z = 4.43, p < .001. The Hodges-Lehmann estimate put the median improvement at 1.88 points, and its 95% confidence interval ran from 1.50 up to 2.25. The paired t-test gave a mean difference of 1.79 points (95% CI 1.45 to 2.13), so the two approaches agree closely. The effect size, r = z divided by the square root of 28, was 0.84, which is large by conventional thresholds (Sullivan & Feinn, 2012).
The Interval as the Finding
The finding is the interval, not the p-value. On an eight-item checklist, the low end of 1.50 means that at the least, nurses were putting about one and a half more of the eight securement steps in place on each catheter; the high end, 2.25, means more than two. Both ends describe a change the unit would want: even the low end moves the typical catheter from five steps to more than six and a half. The item that moved most was the securement device on the hub, present on 19 of 112 audited catheters in August and 94 of 112 in October. The item that moved least was a documented site score, from 61 to 74 of 112, which suggests the shift check is the part of the bundle that most needs attention.
How the Scores Were Checked
Before any test was run, every audit sheet was entered twice, by the project lead and by the unit's data analyst, and the two files were compared line by line; four entry errors were found and corrected against the paper sheets. A second auditor, the unit educator, independently scored 20 of the October catheters on the same rounds, without seeing the project lead's sheets. The two auditors gave identical totals on 16 of the 20 and differed by one item on the other 4, always on whether the extension set counted as secured. That item's wording was tightened for later audits. Agreement at this level supports using the scores, but the single disputed item is a reminder that a checklist is only as firm as its vaguest line.
What This Analysis Could Not Detect
Three limits matter. First, the audits were announced and done by the project lead, who also ran the training, so nurses may have applied the bundle more carefully on audit days than on others, and the auditor knew which month each score belonged to. Second, the analysis shows practice changed; it says nothing about whether catheters failed less, because audit scores were not linked to catheter outcomes. That is the question for the outcome comparison planned for later in the course. Third, the ceiling means the analysis could not show improvement among the nurses who were already near 8 in August, so it understates change for strong performers. For the two nurses whose scores did not move, no change was detected, but with a ceiling in the way, the absence of a detected change is not evidence that their practice stayed the same (Altman & Bland, 1995).
References
Altman, D. G., & Bland, J. M. (1995). Absence of evidence is not evidence of absence. BMJ, 311(7003), 485. https://doi.org/10.1136/bmj.311.7003.485
Polit, D. F., & Beck, C. T. (2021). Nursing research: Generating and assessing evidence for nursing practice (11th ed.). Wolters Kluwer.
Sullivan, G. M., & Feinn, R. (2012). Using effect size, or why the P value is not enough. Journal of Graduate Medical Education, 4(3), 279-282. https://doi.org/10.4300/JGME-D-12-00156.1
Reading the NR 714 Week 4 instructions
Name each variable and its measurement level, and let the test follow from that and from the design. If the same people or patients were measured twice, treat the data as paired rather than as two independent groups. Check the assumptions your test needs and report them honestly, including the ones that hold poorly, and say what you did about it. Report the result with an estimate and its confidence interval first, and the test statistic and p-value after. Then read the interval in practice terms: what the low end would mean for the unit, what the high end would mean and whether either would change what you do. Say how the data were checked before analysis. Close with what this analysis was never able to detect, so the reader does not take more from it than it holds.
Inside the NR 714 Week 4 example
The sample explains the eight-item audit, how each nurse's score was formed from four catheters and why three nurses who left were not a threat. It names the score's level and its ceiling, then argues for pairing on the 28 nurses. The assumptions section reports a Shapiro-Wilk result of W = 0.91 and a left tail from nurses already near the top, which leads to the Wilcoxon signed-rank test. A results table gives medians, the count of nurses who improved, the Hodges-Lehmann difference and the paired t mean. The interpretation reads the interval as one and a half to more than two securement steps per catheter and names the hub device as the item that moved most. A double-entry and second-auditor check precede the limits.
Reading the NR 714 Week 4 grading rubric
The heaviest loss is a result reported only as significant or not, with no size and no interval. Next is a test that does not match the design, most often an independent groups comparison run on the same people measured twice. Marks are also lost when assumptions are skipped on visibly skewed data, when output is pasted in without the written reporting the assignment asks for and when a wide interval is described as if its midpoint were the finding. Treating statistical significance as clinical importance costs points, as does a conclusion about all patients drawn from one unit. Write-ups score well when they read both ends of the interval in practice terms and state plainly what the analysis could not detect.
NR 714 Week 4 help from the desk
Choosing the right test and then explaining the result in plain terms is where most students need help. If you share your variables, your design and your output, we can help you check whether the data are paired, which assumptions matter, how to report the estimate and interval and how to explain what the ends of the interval would mean on your unit. We can also help you set out the limits section. The data and the analysis choices remain yours, and the write-up reports them faithfully.
Get a NR 714 Week 4 example written to your instructions
Send the assignment brief, the rubric and whatever data or data structure you are working from, including the software your section requires. An original example is written around it and returned inside 24 to 48 hours. The first one is free, and the reporting conventions are matched to your classroom.
More NR 714 papers and DNP samples
- NR 714 Week 1: Practice Question Post
- NR 714 Week 2: Outcome Measure Write-Up
- NR 714 Week 3: Epidemiologic Rates Write-Up
- NR 714 Week 5: Data Analysis Plan
- NR 714 Week 6: Data Analysis Write-up
- NR 714 Week 7: Evaluation Plan
- NR 714 Week 8: Results Interpretation Presentation
- NR 709A Week 4: Conference Abstract Draft
- NR 707A Week 4: Midpoint Data Summary
- NR 705A Week 4: Baseline Description
- NR 702A Week 4: Site Agreement Summary
NR 714 Week 4 questions, answered
Which software should the example assume?
Tell us and we match it. Sections variously specify SPSS, Excel, R or a package bundled with the classroom, and the output conventions differ enough to matter for presentation marks. Where nothing is specified, the example reports results in a widely accepted written form and shows the output separately, so that half can be swapped for whatever your section actually requires.
Do I have to run the numbers myself?
Usually yes, and the example is meant as a model rather than a substitute. It shows what a defensible write-up around your own output looks like: the test justified from the design, assumptions reported instead of assumed, an interval read back into practice language. Send the dataset structure or a realistic description of it and the example is built to match what you have.
What if the result comes back non-significant?
That is a result, and at doctoral level it is often the more interesting one to write up. The strong version reports the estimate and the interval, states how large a difference the project could have detected given its numbers, and separates a finding of no difference from a failure to find one. Papers that quietly go hunting for a different test lose marks for exactly that.
Where can I find a free NR 714 Week 4 sample paper?
This page holds a complete paired analysis of nurses' audit scores before and after training, with assumptions, results table and interpretation.
Why does the NR 714 Week 4 sample use a Wilcoxon test instead of a t-test?
The paired differences failed a normality check and bunched near the score's ceiling, so a rank-based test was chosen and the t-test shown only for comparison.