NR-599 Week 2 Information Literacy and Search: How to Write It

What NR-599 Week 2 asks for

Picture the question that starts a search in a pediatric clinic. A four-year-old with intermittent wheeze has been on a rescue inhaler for eight months and the parent wants to know whether a daily controller is worth it. That curiosity is not yet a searchable question. It becomes one when you name the population, the intervention, the comparison and the outcome you would accept as an answer, at which point the search stops being a hunt for articles about asthma and becomes a request for a specific body of evidence. Half the points in an information literacy stage sit in that conversion, and students who skip it end up writing about the six thousand results they could not narrow.

The second half of the territory is the mechanics: knowing that a database has a controlled vocabulary and that mapping your natural language to it changes both the size and the relevance of the set, knowing what an explode does to a subject heading, knowing why a filter for the last five years is a defensible choice in one topic and an indefensible one in another. Advanced practice makes this practical rather than academic. You will be answering questions between visits for the rest of your career, and the difference between a clinician who can build a search in four minutes and one who cannot is measured in decisions, not in grades.

Deliverables at this depth commonly ask for a documented search with a written account of how it was built, sometimes with a table of terms and yields, and sometimes with a short appraisal of what you found. If your section runs a discussion here, expect it to ask for something checkable, since a claim about a database is easy to verify and posts do not reopen once submitted in Canvas. Write the numbers down as you go. Nobody reconstructs a search accurately from memory the next day, and a fabricated yield count is the fastest way to lose a support row you had already earned.

The NR-599 Week 2 method, step by step

Six moves for building a search a stranger could repeat.

1. Write the clinical question in a structured format first

   Population, intervention, comparison, outcome and, where it matters, the time window. Do this before you open a database. A question written after the searching has begun is always shaped by what you happened to find.

2. Split the question into concept blocks, not keywords

   Three or four blocks, each holding every way that concept gets said. Preschool wheeze, recurrent wheezing and early childhood asthma belong in one block joined by OR, and the blocks join each other with AND.

3. Map each block to the database's controlled vocabulary

   Find the subject heading that already exists for your concept, then combine it with free text so you capture recent records that have not been indexed yet. Say in your write-up which terms were headings and which were keywords.

4. Record the yield after every single step

   Keep a running line for each search: the string, the database, the limiters and the number of results. That log is the raw material of your paper and it cannot be rebuilt afterward with any honesty.

5. Justify each limiter instead of applying it silently

   A date filter, a language filter and a peer-reviewed toggle each remove records, and each removal needs a reason. Say why five years is the right horizon for your topic, or say that you widened it because the foundational trial sits outside the window.

6. Screen in two passes and report both

   Titles and abstracts first, full text second, with the count that survived each pass and the reasons the rest were excluded. Two or three exclusion reasons stated plainly beat a vague sentence about relevance.

A layout and word budget for a documented search

Our frame for a search documentation paper, sized for roughly 1,000 to 1,300 words plus the table. It is our own outline rather than anything the university issues, and your week's rubric outranks it wherever the two disagree. If your section asks for a specific template, follow the template and use this only to decide what belongs in each part.

Evidence craft for information literacy writing

Name the database and the platform, not just the name. The same index behaves differently through different interfaces, and a grader who searches your topic needs to know where you were standing. Write that you searched a named database through your university library platform on a stated date, because search results move and a date makes the record honest.

Report yields as counts at each step, not as an ending total. The teaching value of a search log is in the drop between lines: 3,410 records for the population block alone, 612 once the intervention block is added, 47 after the design limiter. That progression demonstrates control. A single final number demonstrates nothing.

Prefer a named source type over a vague one. A systematic review, a clinical practice guideline from an issuing body, and a single cohort study are three different kinds of evidence, and calling all of them articles flattens a distinction the course is teaching. Say what each item is when you introduce it.

Do not let a search engine result stand in for a database record. Open web material can be legitimate when it is a guideline or an agency publication, in which case name the issuing organization and the publication year. What cannot appear in a graduate information literacy paper is a summary site or a content mill standing where the primary source should be.

De-identify the clinical trigger completely. The scene that produced your question should carry an age range, a setting description and nothing else. No date, no practice name, no detail combination that would let a colleague recognize the family. Specificity in a paper comes from the reasoning, never from the identifiers.

Five mistakes that cost points in this week's territory

• A question written after the search. It always shows, because the outcome named in the question matches exactly what the first useful article happened to measure.
• Keyword-only searching. Skipping controlled vocabulary produces sets that are simultaneously too large and missing the best records, and the write-up cannot hide it.
• Yields invented after the fact. Numbers that do not descend plausibly from step to step are the single most detectable form of guessing in this assignment.
• Limiters applied without reasons. Every filter is an argument about relevance, and an unexplained five-year window is an argument you declined to make.
• No verdict on the evidence base. A paper that lists what it found without saying whether the literature actually answers the question has stopped one section early.

Before you submit

• The structured question appears before any search step is described
• Each database is named with the platform and the date you searched it
• Controlled vocabulary terms are distinguished from free text terms
• Every search line carries its own yield count
• Both screening passes are reported with counts and exclusion reasons
• The clinical trigger contains no identifying detail of any kind

Questions students ask about this stage

Keep going