DNP 965A Topic 1 prespecified analysis report example

DPI Project: Part III Grand Canyon University Free custom sample in 24 to 48h

This DNP 965A Topic 1 sample paper reports the prespecified analysis of a statin outreach project exactly as the plan filed in July 2029 required, with exploratory comparisons kept apart. It was written for GCU DNP 965A, DPI Project: Part III, catalog code DNP-965A, a doctoral seminar in Grand Canyon University's DNP. The plan's central passage is quoted. The cohort of 971 is reduced to 912 paired patients with every exclusion counted. McNemar's test on the paired yes-or-no outcome shows a 9.6-point rise, above the failure line. One test change for the balancing measure is disclosed. Two exploratory comparisons are labeled, with Wasserstein and Lazar and Sullivan and Feinn.

CourseDNP 965A DPI Project: Part III
TopicTopic 1
Paper typePrespecified analysis report
LengthAbout 970 words, 3 pages
FormatAPA 7 student paper
SchoolGrand Canyon University (GCU)
ProgramDNP
UpdatedOctober 2026

Free sample paper for DNP 965A Topic 1

1

The Plan Said McNemar, So the Report Says McNemar: The Prespecified Analysis of a Statin Outreach Project, With Exploratory Comparisons Kept Apart

[Author Name]

College of Nursing and Health Care Professions, Grand Canyon University

DNP-965A: DPI Project: Part III

Topic 1 Assignment

[Instructor Name]

March 18, 2030

The organization, setting and figures below are a composite written as a model document. No real employer, client, colleague or patient is described.

What this page is doingThe title makes the prespecified plan the authority for the analysis, which is the discipline this paper demonstrates. The network and data are composites; the methods cited are real.
2

The Plan Said McNemar, So the Report Says McNemar: The Prespecified Analysis of a Statin Outreach Project, With Exploratory Comparisons Kept Apart

The Plan, Quoted

The analysis plan was written and filed with the network's quality director in July 2029, before any implementation data existed. Its central passage reads: "The primary measure is a yes-or-no outcome recorded for the same patients at two points, August 31, 2029, and February 28, 2030, so the observations are paired, and McNemar's test for paired proportions is used rather than a chi-square test for independent groups. Alongside its p value, the plan reports the difference in proportions with a 95 percent confidence interval." It also specified that the run chart would be judged by the standard shift and trend rules, that the project would count as failing its primary aim if the proportion rose by less than 5 percentage points, and that no subgroup analyses would be run within the primary measure. This report follows that plan, reports the primary result whichever way it points and places everything the plan did not specify under a separate heading.

The Primary Outcome, as Specified

The cohort fixed on August 31, 2029, numbered 971. After removing the 50 patients transferred with Clinic E, the 2 who developed a contraindication and 7 who left the network individually, 912 patients were analyzed at both time points. On August 31, 523 of them, 57.3 percent, counted as on a statin under the network's one-year look-back definition. On February 28, 611, 67.0 percent, did. The difference is 9.6 percentage points, with a 95 percent confidence interval of 6.9 to 12.4 points.

The assumption checks for McNemar's test are straightforward and were done before the test. The outcome is binary and measured on the same patients twice, which is the pairing the test requires. The test's chi-square approximation is reliable when the discordant pairs are numerous, and here they were: 130 patients moved from no statin to a statin, and 42 moved the other way, most because a statin last filled more than a year earlier aged out of the look-back window. With 172 discordant pairs, the approximation holds, and the continuity-corrected statistic was 44.0 with one degree of freedom, p less than 0.001. One assumption is only approximately met: patients are clustered within clinics and prescribers, and the test treats them as independent. The plan chose not to model clustering, since six clinics are too few to estimate it well, and this report says so rather than adjusting after the fact.

The lower end of the confidence interval, 6.9 points, is above the 5-point failure threshold the plan set, so the project meets its primary aim by its own prespecified standard. The run chart agrees: the trend rule was met with the January value and the shift rule with the February value (Perla et al., 2011).

The monthly new-starts measure, the plan's secondary outcome, is reported descriptively as the plan required: 138 new starts across the six implementation months, against about 54 in the six baseline months.

What this page is doingThe primary result is reported exactly as the plan specified, with the test named, assumptions checked and the prespecified threshold applied, before any other comparison appears.
3

A Change of Test, Disclosed

One prespecified test was changed, and the change is disclosed here. The balancing measure, muscle symptom contacts within 90 days of a new statin, was to be compared between new starts in the implementation period and new starts in the baseline months using a chi-square test. The baseline had only 54 new starts, and the expected count of muscle symptom contacts in that group was below five, which makes the chi-square approximation unreliable. Fisher's exact test was used instead, a decision made when the counts were first tabulated on March 4 and recorded in the analysis log that day. The result: 19 of 138 implementation-period starters, 13.8 percent, had a muscle symptom contact, against 4 of 54 baseline starters, 7.4 percent; the two-sided Fisher's exact p value was 0.32. The difference is not statistically detectable, but its direction is against the intervention, and it is reported with the same prominence as the primary result rather than dismissed because the test was not significant, since a p value above 0.05 is not evidence of no difference (Sullivan & Feinn, 2012; Wasserstein & Lazar, 2016).

Exploratory Comparisons

Two comparisons were not in the plan, and they appear here under an explicit exploratory heading. Neither should be read as a test of the project's aim.

The first compares the cohort's change with the change among patients aged 55 to 75 with diabetes in the same network, who received no outreach. It was decided on January 7, 2030, in the interim review, because the implementation months fell in autumn and winter, when visits rise. Over the same six months, statin use in the older group rose 1.4 points, from 74.9 to 76.3 percent. The comparison suggests that seasonal visit patterns account for a small part of the cohort's 9.6-point rise, not most of it, but the older group differs in age, insurance and baseline rate, so it is a rough check rather than a control.

The second compares the proportion who started a statin among patients reached in the daytime and in the evening. It was decided on December 3, 2029, when the shortened evening call was introduced, to see whether the lighter format changed decisions. Among daytime patients reached, 66 percent started; among evening patients, 49 percent. The difference cannot be attributed to the call format, since evening patients also differ in age, work and prior contact with the clinic.

What This Report Does Not Claim

The report claims that the prespecified primary outcome improved by 9.6 points, above the failure threshold, in a before-and-after comparison of the same patients. It does not yet say why. Attribution to the outreach, and the other explanations for the change, belong in the interpretation section, which will treat the exploratory results as context and not as confirmation.

References

Perla, R. J., Provost, L. P., & Murray, S. K. (2011). The run chart: A simple analytical tool for learning from variation in healthcare processes. BMJ Quality & Safety, 20(1), 46-51. https://doi.org/10.1136/bmjqs.2009.037895

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

Wasserstein, R. L., & Lazar, N. A. (2016). The ASA statement on p-values: Context, process, and purpose. The American Statistician, 70(2), 129-133. https://doi.org/10.1080/00031305.2016.1154108

The DNP-965A Topic 1 assignment instructions

Topic 1 in DNP-965A asks students to report the analysis of their DPI Project data. Strong reports quote the analysis plan written before data existed and follow it, reporting the primary outcome with its denominator and every exclusion counted. Students should disclose any change to a prespecified test with the reason, such as a distribution that made the planned test unsuitable, and place comparisons that were not in the plan under an explicit exploratory heading so they are not read as tests of the project's aim. The report should state what it claims and what it does not yet claim, leaving interpretation and explanation for the discussion that follows. Figures must match the plan.

Inside the DNP 965A Topic 1 example

The finished example begins by quoting the analysis plan from the approved proposal, which fixes the comparison before anyone can be tempted to improve it. The primary outcome is then reported exactly as that plan specified, with the test named, the assumption checks shown, and the result given whichever direction it points. Two further comparisons appear afterward under an explicit exploratory heading, and the paper says why each was run and when the decision to run it was taken. A change of test is disclosed openly, along with the distributional problem that prompted it. The closing paragraph states what the primary comparison alone supports, which is narrower than the writer had hoped at proposal.

DNP 965A Topic 1 rubric: what Excellent looks like

Faculty reward an analysis that follows its own plan. An Excellent report quotes the July 2029 plan and reduces 971 patients to 912 with each exclusion counted, including 50 transferred with Clinic E. Credit goes to McNemar's test on paired outcomes showing a 9.6-point rise past the failure line, with effect size after Sullivan and Feinn. The rubric values disclosing the changed test for muscle symptom contacts. Strong reports label two comparisons exploratory, with Wasserstein and Lazar's statement as support. Top marks go to stating what the report does not yet claim. Exclusions are counted.

Common DNP 965A Topic 1 mistakes, and how to avoid them

The weakest submissions report a comparison the proposal never named, which is visible the moment a reader sets the two documents beside each other. A second failure is running several tests and reporting the one that returned the most agreeable figure, since selection of that kind is invisible in the result and obvious in the method. Marks also go for assumption checks left undone, because a violated condition makes a reported interval meaningless rather than merely imprecise. Exploratory analysis placed alongside the primary comparison, with nothing marking the difference, invites a conclusion the design cannot carry. Departures concealed rather than disclosed are the error a committee remembers longest. Outcomes reported only in the favorable direction misstate what was found.

Get a DNP-965A Topic 1 example written to your instructions

Send the DNP-965A Topic 1 assignment instructions and the rubric your classroom posts, with the analysis section of your approved proposal. We write a custom example to those criteria, with the prespecified comparison run first and exploratory work kept separate, in 24 to 48 hours. Your project, your data and your review submission are yours and are never drafted here. The first one is free.

More DNP 965A papers and DNP samples

DNP 965A Topic 1 questions, answered

Why report a result that went the wrong way?

Because the analysis plan committed you to reporting it, and a committee can read the proposal. A comparison specified in advance and then omitted is the clearest signal available that selection has occurred. An unfavorable primary outcome costs a doctoral paper very little on its own, and it protects everything else in the document from the same doubt.

What separates prespecified from exploratory analysis?

The date of the decision. Anything named in the approved plan before collection began carries the weight the design was built to give it. Anything decided once the figures were visible is exploratory, however sensible it looks, and can raise a question rather than settle one. Recording when each choice was made is what keeps the distinction checkable by somebody else.

Can I change the test after seeing the distribution?

Sometimes, and the change has to travel with its reason. A skewed outcome or a sparse cell can make the planned test indefensible, and substituting a more appropriate one is better practice than proceeding regardless. What loses marks is the silent swap, presented as though it had been the intention throughout, which any reader finds by comparing two documents.

Where can I find a free DNP 965A Topic 1 sample paper?

The analysis report for DNP 965A Topic 1 covers the plan quoted, the primary outcome as specified, a disclosed change of test, exploratory comparisons and what the report does not claim. Send your DNP 965A Topic 1 prompt and a report follows free.

Why use McNemar's test in DNP 965A Topic 1?

Because the plan specified it for a yes-or-no outcome measured twice in the same patients. Following the plan keeps the result from being chosen after the data were seen.