NR-705C · Week 3 of 8 · The data collection procedure and data governance

NR-705C Week 3 Write the Data Collection Procedure: How to Write It

The short answer

A measurement plan says what will be counted. The collection procedure says how the counting physically happens: who runs which report on which day, where the file lands, who checks it, what happens when a figure looks wrong, and how the whole thing continues if you are on leave for a week. This is the document that makes your project reproducible, and reproducibility is what lets a site keep the measure after you finish. Your section may print this as NR 705C or NR705C; it is the same course. 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 705C Week 3 grading scale at Chamberlain, the criterion levels this assessment is scored on, from Chamberlain Tutors
How Chamberlain grades NR 705C Week 3, visualized by Chamberlain Tutors.

What NR-705C Week 3 asks for

A pediatric group running a reconciliation project across two locations discovers in the third week that the weekly figure has been produced twice by different people using slightly different date filters, and that neither version was saved with a label. The numbers differ by eleven. Nobody did anything wrong; the procedure simply never existed. That is the failure this stage is designed to prevent, and it is much more common than a failure of the intervention itself, because clinicians are trained to deliver care and are rarely trained to run a data process.

Write the procedure as a sequence with roles and days attached, in the same executable register a clinical protocol demands. The analytics lead runs the standing report on the first working day of each week for the preceding week and saves it to the project folder with the week ending date in the file name. That is a step. Data will be collected weekly is a statement of intent that leaves every practical question open, and every open question becomes an inconsistency in your dataset.

Governance belongs here too, and a 256-hour block gives you time to get it right rather than retrofitting it. Where does the data live, who can open it, what is the minimum necessary set of fields, how is it de-identified before it goes anywhere near a course submission, and what happens to it at the end of the project. Doctoral readers look for this section specifically, and its absence reads as inexperience with organizational data even when the project itself is sound.

Where the boundary sits. Everything the practicum verifies is yours. The 256 clinical hours, the log recording them, activity and encounter counts, evaluations completed by preceptors or site mentors, agreements and signatures are your own record and your own work, and they are never drafted, reconstructed, back-filled or estimated with any help, from anyone, at any stage. Written support covers the written layer only: structuring a collection procedure, writing operational language precisely, organizing a governance section. Handle project data under your organization's rules, aggregate it in anything you write, and remove any detail that could identify a family or a staff member. The clinical hours cannot be shortcut and no honest service offers to shorten them; what is on offer is clearer writing about work you genuinely did.

The NR-705C Week 3 method, step by step

Six moves for a collection procedure another person could run.

  1. Assign each collection task to a role and a day

    Not a frequency. A named role and a specific day of the week or month. Weekly is a wish; the analytics lead on the first working day is a schedule, and schedules survive your absence in a way that intentions do not.

  2. Specify the file naming and storage convention in the document

    Measure, site, period ending date, version. It sounds bureaucratic until the week you have four files called reconciliation data and no way to tell which produced the figure in your last submission.

  3. Write one verification step per collection cycle

    Denominator within an expected range, no impossible values, site totals reconciling to the combined figure. Verification catches the broken report in the week it breaks rather than in the week you write your results.

  4. Define what happens when a figure looks wrong

    Who is contacted, what is rechecked, and how a corrected figure is recorded without overwriting the original. Corrections are normal; silent corrections are a data integrity problem.

  5. Write the minimum necessary field list and defend it

    List every field you are extracting and one clause on why each is needed. Fields nobody can justify should not be in the extract, and a governance reviewer will ask exactly this question.

  6. State the de-identification point in the flow

    At which step identifiers are dropped, who holds any linking key if one exists, and what form the data takes by the time it reaches your writing. This sentence is the one doctoral readers scan for.

A layout and word budget for a collection procedure

Our frame for a data collection and governance section, sized for roughly 1,500 to 1,900 words plus the schedule table. It is our own outline rather than anything the university issues, and your chair's direction and your week's rubric outrank it wherever they disagree.

SectionWhat belongs in itWord target
Scope of the procedureWhich measures it governs, which sites, over what period, and what it deliberately does not cover.120 to 160
Collection schedule tableMeasure, source, role, day, output file, storage location, verification step.Table plus 150
Step-by-step procedureThe sequence written executably, including how a report is run and what parameters are set each time.320 to 400
Verification and correctionThe checks applied each cycle, the tolerance that triggers a query, and how corrections are recorded without overwriting.250 to 310
Field list and minimum necessaryEvery field extracted with a one-clause justification, and the fields deliberately excluded.200 to 260
Storage, access and de-identificationWhere data lives, who has access, at which step identifiers are removed, and how the set is disposed of at the end.250 to 310
Continuity planWhat happens to collection when you are unavailable, and who can run the procedure from the document alone.150 to 200

Evidence craft for the data section

Write in the imperative and name the actor. A procedure is an operating document, not a description of one. The analytics lead runs, the project lead verifies, the practice manager is notified. Passive constructions hide the owner and owners are the whole point.

Cite your organization's own policy where it governs you. Data handling rules, minimum necessary standards and retention requirements are institutional, and referencing them by name shows you have read them rather than improvised. Do not quote them at length; name and apply.

Quantify your verification tolerances. A denominator more than fifteen percent outside the historical range triggers a query is a check. Figures will be reviewed for reasonableness is not, and it will not catch anything because nobody knows what would fail it.

Distinguish project data from the clinical record in every sentence. You are extracting from a system of record into a project dataset. The clinical record is not yours to alter, annotate or supplement for project purposes, and saying so plainly demonstrates that you know the boundary.

De-identify before anything leaves the organizational environment. Course submissions, drafts shared with a chair and anything stored outside approved systems should contain aggregates only. Write the rule into the procedure so it is a step rather than a habit you might forget under deadline.

Five mistakes that cost points in this week's territory

  • Frequency without a day and a role. Weekly collection with no owner produces gaps in exactly the weeks that are busiest, which are the weeks most likely to matter.
  • No verification step. A report that silently breaks mid-term will not announce itself, and the gap is usually discovered when the numbers stop making sense.
  • Corrections that overwrite. If the original figure disappears, you cannot show what changed or why, and a committee reading a corrected dataset with no audit trail will ask.
  • Extracting more than you need. Every unnecessary field is a governance question you will have to answer and a risk you did not have to take.
  • No continuity plan. A procedure only you can run is a procedure that stops the week you have influenza, and the site cannot sustain the measure after you leave.

Before you submit

  • Every collection task names a role and a specific day
  • File naming and storage conventions are written down
  • Each cycle carries at least one quantified verification check
  • The correction process preserves the original figure
  • Every extracted field has a one-clause justification
  • The de-identification point is identified explicitly in the flow
  • Someone else could run the procedure from the document alone

Writing the NR-705C data procedure?

Send the rubric and your measurement plan out of Canvas. A premium original draft comes back in 24 to 48 hours with the procedure written executably and the governance section built to what doctoral readers look for, revised free until it lands. Practicum hours, logs and evaluations stay entirely yours.

Questions students ask about this stage

I am collecting the data myself. Do I still need all this procedure?
More than if someone else were, because a single collector is the least reproducible arrangement there is. When you run every extract yourself, your parameters live in your head and drift without anyone noticing - a date filter set inclusively one week and exclusively the next, a site filter forgotten during a busy period. Writing the procedure forces those decisions to be fixed and makes them checkable against your own outputs. It also does two things for your project beyond consistency. It gives the site something they can continue after you leave, which is a genuine sustainability contribution and reads as one in your final chapters. And it protects you if a figure is ever questioned, because you can show precisely how each number was produced rather than reconstructing it from memory months later. At this hour load there is time to write it properly, and it is one of the highest-return hours in the block.
What do I do about missing data?
Decide the rule before you see the data, write it down, and apply it consistently. The decision is usually between treating a blank as not done and treating it as unknown and excluding it, and the two produce different numbers with different meanings. Treating blanks as not done is conservative and usually defensible for a process measure where the record is the evidence of the act. Excluding them can be right where you know a field is unreliable, but it shrinks your denominator and can flatter your result, so it needs a stated justification. Either way, report how much data was missing alongside your figures - forty-one visits with no recorded value out of 387 eligible - because a reader cannot judge a result without knowing how much of it was inferred. What is not acceptable is choosing the rule after looking at both versions of the number, and a committee that suspects it will ask when the rule was written.
How much detail about data security belongs in a course submission?
Enough to show you understand the obligations, and not so much that the submission itself becomes a security document. Two or three sentences covering where data resides, that it stays inside approved organizational systems, who has access, and at what point identifiers are removed will satisfy most doctoral readers. Do not name specific file paths, system credentials or server details in a document that will be uploaded to a course platform. If your organization requires a formal data management or security plan, that is a separate artifact submitted through their process, and your course writing should reference its existence rather than reproduce it. The general principle is worth carrying through the whole program: write enough that a reader can see the safeguards are real, and never include operational detail whose disclosure would itself be a problem.

Keep going

Online now