NR-585NP · Week 2 of 8

NR-585NP Week 2 Literature Search and Levels of Evidence: How to Write It

The short answer

NR-585NP Week 2 usually turns the question you wrote into a search a stranger could run again, then asks you to rank what came back. Two skills sit underneath: building a string out of subject headings and Boolean operators, and placing each result on a hierarchy that puts filtered synthesis above single studies and single studies above opinion. Your section may print this as NR 585NP or NR585NP; it is the same course. A search you cannot describe is a search nobody can score.

Chamberlain publishes no syllabi outside Canvas. The placement here is our teaching judgment from the course's catalog arc; your section's rubric decides what your week actually asks.

NR 585NP Week 2 grading scale at Chamberlain, the criterion levels this assessment is scored on, from Chamberlain Tutors
How Chamberlain grades NR 585NP Week 2, visualized by Chamberlain Tutors.

What NR-585NP Week 2 asks for

The search half is mechanical and it rewards patience. Split the question back into concepts, gather the words the literature uses for each concept, look up the controlled vocabulary that indexes them, then join synonyms with OR inside a concept and concepts with AND across them. Record the number of hits after each move. That running count is the evidence that the search happened rather than a claim that it did.

The ranking half is conceptual. Systematic reviews and meta-analyses of trials sit at the top of an intervention hierarchy because somebody has already appraised and pooled the studies underneath. Single randomized trials follow, then quasi-experimental work, then cohort and case-control designs, then descriptive and qualitative studies, then expert opinion. Level answers how much a design controls for error, not how good the individual paper is, and saying so is worth a sentence in your draft.

Deliverables at this point in an 8-week session tend to run 800 to 1,200 words and often carry a small table of terms and hit counts. If your section runs a discussion this week, expect to post the strategy for peers to pull apart, so build the table before you write the post rather than after.

The NR-585NP Week 2 method, step by step

Six moves produce a search you can defend line by line.

  1. Break the question back into concepts

    Most clinical questions carry two or three searchable concepts, usually the population and the intervention, sometimes the outcome. Comparison and timeframe rarely belong in the string. Fewer concepts means more results, which is the correct problem to have at the start.

  2. Collect the words the literature actually uses

    Every concept has several names. Older adult, elderly, geriatric and aged are different doorways into the same idea, and a database indexes them differently. Skim three relevant abstracts and harvest their vocabulary before you write anything.

  3. Look up the controlled vocabulary

    PubMed indexes with medical subject headings and CINAHL with its own subject headings. Searching a heading catches articles whose authors chose different words from yours. Run headings and free text together so nothing indexed slowly gets missed.

  4. Assemble the string one concept at a time

    Join synonyms with OR, then combine the concept sets with AND, noting the count at each stage. A count that collapses from thousands to nine at one step tells you exactly which line to loosen.

  5. Set limits you can justify

    Years, peer reviewed, language, age group and publication type each remove evidence from view. Apply them deliberately and say why in one clause. An unexplained limit reads to a grader like a search that was trimmed until it was convenient.

  6. Screen in three passes and record the losses

    Titles first, then abstracts, then full text, against inclusion and exclusion criteria you wrote before screening started. Report how many were removed at each pass. The arithmetic is the transparency the scoring rows are looking for.

What a search summary contains, and how long each part runs

Sizing for a 1,000 word summary from our own drafting practice. Your week's rubric sets the real limit; these proportions keep the mechanical parts from eating the reasoning.

PartWhat belongs in itWord target
Question restatedThe question in one sentence with its elements visible, so the search has something to answer.80
Databases and rationaleWhich databases and why each suits this question, including any specialist source.120
Terms and headingsProse around the table: synonym sets, headings used, truncation and phrase choices.150
Limits and justificationEvery limit applied and one clause per limit explaining what it protects against.150
Screening and yieldCounts at each pass, inclusion and exclusion criteria, and the number retained.200
Levels of the retained evidenceEach retained source placed on the hierarchy with the design that puts it there.200
What the search failed to findGaps, thin areas and any question the literature has not answered yet.100

Search craft that survives a grader

Date your search. Databases change weekly. One sentence giving the date you ran the final string makes every count in your table verifiable, and it is the first thing an experienced reader looks for.

Headings and keywords do different work. A subject heading finds articles indexed under a concept. A keyword finds articles using your exact word. Running only one of them leaves a predictable hole in the yield.

Truncate carefully. Nurs catches nurse, nurses and nursing, and it also catches nursery. Say what you truncated so a reader can judge the noise it let in.

Guidelines and reports are evidence with a different address. Professional body guidelines and federal reports are not indexed like journal articles and often need a direct site search. Their absence from a nursing search is a hole, not a virtue.

Five years is a default, not a law. Follow the window your week's rubric names, and when you keep an older paper, say in the sentence whether it is a seminal source or the only measurement that exists.

Report the counts you actually got. Rounded or tidied numbers invite the one question you cannot answer. An awkward yield of seven articles, honestly reported, scores better than a suspiciously round thirty.

Five ways a search section loses marks

  • A general web search dressed up as a database search. If the strategy cannot be re-run in a named database with named terms, it is browsing, and the reproducibility rows will say so.
  • One database, no rationale. Nursing questions usually need CINAHL plus at least one biomedical index. Using a single source is defensible only when you write the defense.
  • Keywords with no controlled vocabulary anywhere. This is the most common cause of a thin yield, and it is invisible to the student because the search still returns something.
  • Limits that quietly removed the best evidence. A five year filter can delete the landmark trial your whole argument rests on. Check what a limit costs before you keep it.
  • Level assigned by journal reputation. The hierarchy ranks designs, not titles. A cohort study in a famous journal is still a cohort study, and saying otherwise is scored as a reasoning error.

Before you submit the search

  • The final string is printed in full and could be pasted into the database as it stands
  • Subject headings and free text terms both appear in the strategy
  • Every limit has a clause explaining why it is there
  • Hit counts are recorded at each stage and the screening arithmetic adds up
  • Each retained source carries a level with the design that justifies it
  • The date of the search appears somewhere a reader will find it

Search returning nothing usable?

Send the question and the rubric from Canvas. A premium original draft comes back inside 24 to 48 hours with the string built, the counts recorded, the retained sources levelled, and revisions free until every reproducibility row is satisfied.

What students ask about the search week

How many articles do I actually need?
Let the rubric set the number and the question set the mix. Where no count is given, a defensible small set beats a padded large one: one recent synthesis, two or three primary studies close to your population, and a guideline if one exists. What earns marks is the fit between each source and the question, plus an honest statement of what you left out and why. Ten loosely related articles score worse than four that argue with each other.
My search returns nothing. What do I do?
Loosen one concept at a time and watch the count. Drop the outcome from the string first, since outcomes are described in a hundred ways and rarely indexed the way you phrased them. Then broaden the population, then swap keywords for subject headings. If the yield stays empty after all three, the literature may genuinely be thin, which is a finding worth reporting rather than a failure to hide. Say what you searched, what you found, and what nobody has studied yet.
Can a clinical practice guideline count as a source?
Yes, and at graduate level it usually should appear when one exists for your question. Treat it as filtered evidence: say who produced it, what year it was issued, what evidence grading system it uses, and whether it has been updated. The mistake is quoting a recommendation with no attention to the evidence underneath it. A guideline recommendation supported by expert consensus is not the same claim as one supported by pooled trials, and your sentence should show that you looked.

Keep going

Online now