NR-705C · Week 5 of 8 · Displaying data and describing variation

NR-705C Week 5 Display the Data and Describe Variation: How to Write It

The short answer

Midway through the operating period you have a series rather than a result, and the writing task is to display it correctly and describe what it does without overclaiming. Improvement work is read through time-ordered displays - a run chart with a median, annotated with the dates your changes took effect - and the prose that accompanies them describes variation, not victory. Doctoral readers score restraint here as heavily as they score technique. 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 5 grading scale at Chamberlain, the criterion levels this assessment is scored on, from Chamberlain Tutors
How Chamberlain grades NR 705C Week 5, visualized by Chamberlain Tutors.

What NR-705C Week 5 asks for

A family practice project has fourteen weekly points: eight before the reconciliation step went live and six after. The eye wants to see a rise. The discipline is to ask whether the pattern would be called a signal by any rule agreed in advance, whether the two sites behave the same way, whether a holiday week explains the dip at point eleven, and whether the denominator was stable throughout. Answering those four questions in prose, honestly, is the entire task, and a student who does it well produces a section that will survive into the final manuscript almost unchanged.

The display choice matters and is often made badly. A before-and-after bar chart with two columns throws away every piece of information the series contains and is the most common weak figure in DNP writing. A time-ordered chart with the median from the baseline period extended across the whole series shows stability, shift and the timing of your changes at once. Annotate it with the dates that matter - launch, protocol version two, the second-site briefing - so the reader can see what the project did and when, rather than being told.

Language must match the design. Improvement projects at one organization support shift, trend, run and special cause when a rule is met, and they support saw an increase from one period to another as a description. They do not support caused, produced, or demonstrated efficacy. A 256-hour block usually gives you enough points to say something meaningful about variation, which is exactly why the temptation to overclaim is stronger here than in a lighter term.

Where the boundary sits. Everything the practicum verifies belongs to you and stays with you. Your 256 clinical hours, the log that records them, encounter and activity counts, preceptor and mentor evaluations, site agreements and signatures are your own record and your own work - never drafted, reconstructed, back-filled or estimated with help, and never any part of what is offered here. What written support covers is the scholarly layer: how to structure a data display section, how to describe variation accurately, how to write results prose that does not outrun the design. Project figures should be aggregate in anything you write, handled under your organization's rules, with no service dates and no detail that could identify a family. The hours cannot be shortcut by anyone.

The NR-705C Week 5 method, step by step

Six moves for displaying an improvement series and writing about it honestly.

  1. Plot in time order before you calculate anything

    Every period on the horizontal axis, the measure on the vertical, no gaps and no reordering. Half the analytic questions answer themselves the moment the series is visible, and summary statistics computed before plotting hide exactly what you need to see.

  2. Set the median from the baseline period and extend it

    Calculate the centre line from the pre-implementation points only, then carry it across the whole chart. A median recalculated to include your post-implementation data will absorb the very change you are trying to detect.

  3. Annotate the chart with dated project events

    Launch, each protocol version, the second-site briefing, any external event that plausibly affected the measure. An unannotated chart makes the reader guess at causes; an annotated one lets them evaluate your reasoning.

  4. Apply a rule you can name before you claim a signal

    Standard run-chart rules for shifts, trends and runs about the median exist and are citable. State which rule you applied and whether it was met. A visual impression asserted as a shift is the commonest overreach at this stage.

  5. Chart each site separately, then the combined view

    Two sites with different volumes and different delivery pooled into one line produce a curve that describes neither. Show the components first and let the combined view be context rather than the analysis.

  6. Write the caption so the figure stands alone

    Measure, denominator rule, sites, period, source and what the annotations mark. A figure that needs the surrounding paragraph to be understood will be misread the moment it is pasted into a slide for a leadership meeting.

A layout and word budget for a data display section

Our frame for a mid-implementation data section, sized for roughly 1,400 to 1,800 words plus figures and tables. 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
What is displayed and whyThe measures shown, the display type chosen, and one sentence on why a time-ordered view suits this design.140 to 180
Data completenessPeriods available, any missing points, how missingness was handled, and any change in the denominator rule.200 to 250
Baseline period readThe pre-implementation points, the centre line, and whether the process looked stable before the change.200 to 250
Site-level seriesEach location's chart described in prose: level, variation, and anything the annotations explain.320 to 400
Signal assessmentThe rule applied, whether it was met, and what can and cannot be said on the evidence available so far.250 to 310
Alternative explanationsSeasonality, staffing changes, coding changes, attention effects, and anything else that could move the measure.230 to 290
Figure captionsOne per display, written so the figure is intelligible on its own, with source and denominator rule included.110 to 150

Evidence craft for improvement data

Name the charting method and cite it. Run charts and control charts have published rules and standard construction, and naming your source lets a reader check your reasoning against a known standard rather than against your judgment.

Report denominators alongside every point. A weekly proportion built on nine eligible visits and one built on eighty-one look identical on a chart and mean very different things. Put the denominators in the table beneath the figure so the reader can weigh each point.

Consider the attention effect out loud. A measure often moves simply because it is being watched and discussed. Naming that possibility does not weaken your project; failing to name it means a committee will raise it for you, and at a worse moment.

Resist inferential statistics on this design. A significance test on a two-period comparison at one organization adds no strength and invites a methodological argument. Improvement language describes what your design can actually support.

Keep displays free of anything identifying. Aggregate counts only, no cells small enough to identify an individual, no service dates in the axis labels beyond period ending, and no site names where organizational policy prefers them anonymized.

Five mistakes that cost points in this week's territory

  • Two bars instead of a series. Before and after columns discard the timing, the variation and the annotations, which is most of what your data can tell anyone.
  • A median calculated across the whole series. Including post-implementation points in the centre line absorbs the shift you are trying to detect and makes the chart mute.
  • Claiming a shift by eye. Without a named rule, a visual impression is an opinion, and an opinion about your own project's success is the least persuasive kind.
  • Pooling sites with different volumes. A combined line weights the busier location heavily and describes neither site accurately.
  • No alternative explanations offered. A section that considers only the intervention as a cause tells a reader the student has not thought about confounding at all.

Before you submit

  • Every measure is plotted in time order with no missing periods hidden
  • The centre line comes from baseline points only
  • Charts are annotated with dated project events
  • A named rule is applied before any signal is claimed
  • Each site has its own series before any combined view appears
  • Denominators are reported for every point
  • At least two alternative explanations are considered explicitly
  • Captions make each figure intelligible on its own

Writing the NR-705C data section?

Send the rubric and your own aggregate series out of Canvas. A premium original draft comes back in 24 to 48 hours with the display described in improvement language and the signal assessment written to what the design supports, revised free until it lands. Hours, logs and site evaluations are never touched.

Questions students ask about this stage

My chart shows no change at all. What do I write?
You write what the process measures say, which is why you built them. A flat outcome series with strong delivery - the step completed in most eligible visits, fidelity observations showing it done as designed - is a genuinely informative finding: the change was delivered and did not move the measure at this site in this window, which points at the mechanism, the measure, or the timeframe. A flat series with weak delivery is a different finding entirely and says nothing about the intervention, because it was never really implemented. Write which situation you are in, evidenced, and then reason about why. Common candidates include a measurement window too short for the effect to appear, a measure too distal from the change, a ceiling effect where baseline was already high, and a denominator that includes visits the intervention could never affect. Doctoral committees are not looking for a positive result; they are looking for a student who can explain a null one correctly, and many of the strongest projects report exactly this.
How many data points do I need before a run chart means anything?
More than most students have, which is why the honest sentence about limitation matters. Standard run-chart rules assume a reasonable number of points on each side of your change, and a series of six weekly points before and six after is thin by any account. Two responses help. First, choose the period length that gives you enough points without making each one absurdly small: weekly may be right at a high-volume location and monthly better at a smaller one, and different sites can legitimately use different intervals if you say so. Second, extend backwards where retrospective data exists, since a baseline pulled from an existing report costs hours rather than weeks and this hour load gives you those hours. Where the series remains short, say plainly that the display is descriptive, that no rule was met or that the rule applied is underpowered on this many points, and let your process measures and your qualitative account carry more of the argument.
Should I put charts in the weekly submission or save them for the final paper?
Build them now and use them now. Charts drafted mid-implementation do two things a final-stage chart cannot. They surface data problems while there is still time to correct them - a denominator that shifted, a period with a missing report, a site whose figures were never split out - and they give your chair something concrete to react to, which produces far more useful feedback than prose about data you have not shown. Formatting will change for the final manuscript and that is a small cost. What matters is that the analysis exists early enough to change your decisions. Keep each version dated in your project file, keep the underlying figures in your collection folder rather than only in the chart, and write the caption fully every time so that six weeks from now you can tell which denominator rule produced which line. Students who leave all display work to the end routinely discover they cannot reproduce their own earlier numbers.

Keep going

Online now