The third stage of NR-717 typically moves from cause to measurement: showing that an outcome falls unequally across a population and writing that inequality in numbers a policy reader can act on. Two words carry the load. A disparity is a measured difference between groups. An inequity is a difference judged unjust because it traces to arrangements that could have been made otherwise. The measurement earns the first word; the argument earns the second, and doctoral papers are expected to do both and to keep them distinct. Your section may print this as NR 717 or NR717; 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.
What NR-717 Week 3 asks for
A hospital quality committee reviewed a dashboard every month showing a 30-day readmission rate for heart failure of 18.4 percent. The number was stable, slightly under the regional benchmark, and nobody had questioned it in two years. A doctoral student on the committee asked for the same measure broken out by preferred language, using the field the registration system had captured all along. Patients whose records listed a preferred language other than English readmitted at 26.1 percent. Patients recorded as English-preferring readmitted at 17.2 percent. The single aggregate number had been averaging a nine-point gap into invisibility every month, on a dashboard that the committee had been treating as evidence that the program was working.
That is why this stage exists, and why it belongs in a course that fuses population outcomes with policy. An aggregate rate is a management number. A stratified rate is a policy number, because it identifies who is bearing the burden and therefore who a decision would have to be aimed at. The written work at this stage is usually an analysis that presents an outcome broken out by one or more equity-relevant characteristics, quantifies the gap, and argues whether the gap constitutes an inequity rather than only a difference.
Doctoral expectations here are specific and often under-taught. You are expected to know that race and ethnicity fields in health records are frequently self-reported, frequently registrar-assigned, and frequently missing, and to say which applies to your data. You are expected to know that a difference between two rates can be expressed as an absolute gap or a ratio, that these tell different stories, and that reporting only the flattering one is a choice a reader may catch. You are expected to know that small subgroups produce unstable rates, and to say what your smallest cell size was. None of this is statistical showmanship. It is the minimum honesty that makes an equity claim usable by somebody who has to defend a decision publicly.
Keep the practice-doctorate frame. You are not conducting research into whether the disparity exists in general. You are documenting that a known pattern is present in a bounded population at a measurable size, so that an existing policy instrument can be aimed at it and its effect later checked. That is translation, and the measurement discipline in this stage is what makes the later evaluation stage possible at all.
The NR-717 Week 3 method, step by step
Six moves for writing a defensible equity analysis.
-
Fix the outcome and the denominator before you stratify anything
Write the measure specification in one paragraph: numerator definition, denominator definition, time window, exclusions, data source. Stratifying an underspecified measure produces gaps that dissolve when somebody asks how you counted.
-
Audit the stratifying variable itself
Find out who enters it, from what options, whether it is self-reported, and how often it is unknown or blank. A field that is missing for 22 percent of records cannot support a confident subgroup claim, and reporting the missingness is what separates a doctoral analysis from a dashboard screenshot.
-
Report both the absolute difference and the ratio
Nine percentage points and a rate ratio of 1.52 describe the same gap and land differently on a reader. Give both, and say which one you consider the decision-relevant framing for your audience and why. Absolute differences drive resource decisions; ratios drive equity framing.
-
State every cell size
A subgroup rate built on 19 patients moves several points when one patient changes category. Put the counts in the table beside the rates, and say plainly which subgroups are too small to carry a conclusion. Suppressing nothing and overclaiming nothing is the standard here.
-
Argue the inequity separately from the disparity
Once the difference is measured, write a distinct passage making the normative case: the gap traces to arrangements that could be otherwise, those arrangements were chosen or tolerated, and therefore the difference is unjust rather than merely unequal. Keep this argument out of the measurement section so a reader can accept your numbers even if they resist your framing.
-
Convert the gap into a target somebody could be accountable for
Close by translating the measured gap into a statement of what closing it would mean in counted terms: how many events per year sit in the gap, and what a partial closure would look like. A gap expressed only as a rate difference does not tell a decision maker what is at stake.
A layout and word budget for an equity gap analysis
Our frame for a stratified outcome analysis, sized for roughly 1,400 to 1,700 words plus a table. It is our own outline rather than anything the university issues, and your week's rubric outranks it wherever the two disagree.
| Section | What belongs in it | Word target |
|---|---|---|
| Measure specification | Numerator, denominator, window, exclusions and source, written so another analyst could reproduce your count. | 180 to 220 |
| The stratifying variable | How the characteristic is captured, by whom, with what categories, and how complete the field is. | 180 to 220 |
| Results with counts | Subgroup rates with numerators and denominators beside them, plus the absolute gap and the ratio. | 250 to 300 |
| Stability and limits | Smallest cell, missingness, and what those two facts prevent you from claiming. | 200 to 240 |
| Disparity to inequity | The normative argument, made explicitly and sourced, kept separate from the measurement. | 300 to 350 |
| The counted stake | How many events per year sit inside the gap and what partial closure would mean in whole numbers. | 180 to 220 |
Evidence craft for equity measurement
Write categories as the record wrote them, then say so. If your data source offers a fixed list of race, ethnicity or language options, those categories carry the assumptions of whoever built the list. Reproduce them accurately in your table, then add a sentence acknowledging that the categories are administrative constructs rather than natural kinds and that people who do not fit them are distributed somewhere. This is an expected doctoral courtesy and its absence is noticed.
Never report a subgroup rate without its counts. Twenty-six percent means one thing when it is 26 of 100 and something entirely different when it is 5 of 19. Counts beside rates is a hard rule at this level, and it is the single fastest way to make a table read as competent.
Do not let a difference become a deficit. A measured gap tells you where an outcome is worse, not why, and certainly not that the affected group is doing something wrong. Write the sentence so the arrangement is the subject: the referral pathway required a weekday appointment, rather than these patients did not attend. The grammar carries the analysis in equity writing more than students expect.
Match your comparison population honestly. If you benchmark your subgroup rate against a national figure, check that the national measure used the same numerator definition and window. Readmission measured at 30 days all-cause is not comparable to readmission measured at 30 days condition-specific, and presenting them side by side without saying so is an error a policy reader will find.
Five mistakes that cost points in this week's territory
- Using disparity and inequity interchangeably. The first is measured, the second is argued. Collapsing them means one of the two jobs went undone.
- Rates without denominators. A stratified table of percentages with no counts cannot be assessed for stability and is treated as unverified.
- Silence about missing data. If a quarter of records lack the stratifying field, the paper has to say so and say what it costs. Leaving it out is the omission graders check for first.
- Deficit grammar. Sentences that make the affected group the actor in their own worse outcome undo the determinants argument you built in the previous stage.
- A gap with no stake attached. Ending at a rate difference leaves the reader without the number that would justify spending anything to close it.
Before you submit
- The measure is specified tightly enough to be reproduced
- The stratifying field is described with its source, categories and completeness
- Every rate in the table is accompanied by its numerator and denominator
- Both the absolute gap and the ratio appear, with one named as decision-relevant
- The smallest cell size is stated and its instability acknowledged
- The inequity argument is made explicitly and sits apart from the measurement
- The closing paragraph converts the gap into counted events per year
Writing the equity analysis for NR-717?
Send the rubric and your measure definitions out of Canvas. A premium original draft comes back in 24 to 48 hours with the stratification specified, the counts reported and the inequity argued separately, and revisions run until the grade lands.