NR-670A Week 4 Evidence Synthesis: How to Write It

What a synthesis has to settle before any plan is written

A nurse educator in an oncology day hospital reached this stage with fourteen articles in a shared folder and a draft organized exactly as the folder was, one heading per paper. Her mentor read a page and a half and asked a question the draft could not answer: in which of these did the intervention work in a setting without a dedicated coordinator, because that is the setting you have. The articles contained the answer. The document did not, because nothing in it had been compared. Reorganizing the draft around three findings instead of fourteen authors turned a book report into an argument in a single afternoon.

That is the whole distinction this stage is scored on. A summary tells a reader what each source said. A synthesis tells a reader what the body of evidence supports, where it disagrees with itself, how strong it is, and under what conditions it held. Only the second gives your following stage anything to build on, because an implementation plan has to inherit not just an intervention but the circumstances the intervention needed in order to work. Students who skip the conditions produce plans that propose a staffing-dependent solution in an environment with no spare staff, and the closing paper then has to explain why nothing moved.

The searchable question comes directly out of the problem statement you wrote in the previous stage, and it should be narrow enough to fail. If your question is what improves communication in healthcare, everything is relevant and nothing is decisive. If it is which documentation or handoff interventions have improved completeness of transferred information in comparable outpatient settings, the search has edges, the retrieved evidence can genuinely disappoint you, and a synthesis that reports honest disappointment is worth more in a capstone than one that finds convenient support for a decision already made.

The practicum boundary applies as it does throughout. Your 144 hours, the hour log, activity records, site paperwork and the mentor evaluation are your own record of work you personally performed, never drafted, reconstructed or estimated with help. This stage is academic writing about published literature, so the sensitive material is thinner here than elsewhere, but wherever your setting appears in the applicability section, it appears de-identified: role and volume rather than names, and no incident described specifically enough to be recognized. What the written layer supports is the reasoning: how to build a matrix, how to weight a body of evidence honestly, and how to write conditions that your implementation stage can actually use.

The NR-670A Week 4 method, step by step

Six moves that turn a folder of retrieved articles into an argument with conditions attached.

1. Derive the question from the problem statement, not from the search results

   Name the population, the intervention type, the comparison if one exists and the outcome you defined in the previous stage. Writing the question before searching is what stops the retrieved literature from quietly redefining your project into something with more convenient evidence behind it.

2. Record the search well enough that someone could repeat it

   Databases used, the terms and combinations, the date limits, and the inclusion and exclusion rules you actually applied. A recorded search is scored as method. An unrecorded one leaves a reader unable to distinguish a systematic effort from whatever surfaced first, and rubrics at this level notice.

3. Build the extraction matrix before writing any prose

   One row per source, columns for design, setting, sample, intervention, outcome measured, direction and size of effect, and limitation. The matrix is a working document, usually not the submission, and its function is to make comparison possible. Prose written without one almost always reverts to a source-by-source march.

4. Write your headings as findings, never as author names

   Each section of the synthesis should be a claim the evidence supports, with the sources gathered underneath it as support, including the ones that disagree. If you can rearrange your sections without changing the meaning, they are not findings, they are containers.

5. Weigh the body of evidence explicitly and out loud

   Say how many studies, of what designs, in what settings, and how consistent the direction of effect was. Two small single-site studies pointing the same way is a legitimate basis for a local pilot and an illegitimate basis for a confident claim, and stating which one you have is a scoring row in most capstone rubrics.

6. End on the conditions the findings depended on

   What resources, roles, technology or authority the successful studies had in place. This is the paragraph your implementation stage will use most, and it is the one students most often omit, which is why so many capstone plans propose interventions their settings cannot support.

A layout and word budget for a leadership synthesis

Our tutors' frame for a synthesis submission of roughly 1,400 to 1,800 words. It is our own outline rather than anything the university issues, and your section's rubric outranks it wherever the two disagree. Scale proportionally if your assigned length differs.

Evidence craft for synthesis writing

A citation stack is not synthesis. A sentence ending with three sources tells a reader that three papers exist. Saying that two found an effect in staffed settings while the third found none where the role was absorbed into existing duties tells a reader something, and the second version costs one extra clause.

Report disagreement as a finding in its own right. Where studies diverge, the interesting question is what differed between them: the setting, the population, the intensity, the measure. That analysis is the highest-value paragraph you can write here, and it is the one that most clearly separates graduate synthesis from undergraduate literature review.

Match the strength of your verbs to the strength of your evidence. Demonstrates, suggests, is associated with and has been reported in are different claims. A body of small observational studies supports the weaker verbs only, and a capstone that reaches for the strongest verb available loses credibility precisely where it was trying to gain it.

Say what the evidence did not address. The gap between the published question and your question is real and naming it protects you. It also does useful work, because a plan built to address something the literature never tested has to be justified locally, and acknowledging that up front is stronger than being caught out by it in the closing stage.

Keep internal data out of the literature section. Your audit findings are evidence about your setting and belong in the transfer paragraph, clearly labelled as internal. Mixing them into the synthesis of published work blurs the line a grader is specifically checking.

Five mistakes that cost points in this week's territory

• One paragraph per article. The structural signature of a summary, and it is visible from the headings before a grader reads a line.
• Sources that only agree with you. A synthesis with no dissenting finding anywhere reads as a search conducted to confirm a decision already made.
• No stated appraisal of quality. Treating a small pilot and a multi-site trial as equivalent forfeits the rubric row that asks you to weigh evidence.
• Applicability asserted rather than argued. Saying the findings apply to your setting without comparing populations and resources skips the analytic step the stage exists for.
• No conditions paragraph. Without it the next stage has an intervention and no idea what the intervention requires, and the plan built on top of it will not survive contact with the setting.

Before you submit

• The question is traceable to the problem statement you wrote earlier
• The search is recorded in enough detail to be repeated
• Every heading in the findings section states a claim, not an author
• Designs, settings and consistency are appraised explicitly
• At least one disagreement between sources is examined rather than smoothed over
• Applicability is argued by comparing population and resources in both directions
• The document ends with conditions the implementation stage can design around

Questions students ask about this stage

Keep going