NR-503 Week 3 works the three questions descriptive epidemiology exists to answer: who, where and when. Person, place and time, drawn from surveillance systems, vital records, national surveys and registries, produce a profile of a health problem before anybody argues about its causes. The stage also teaches what a surveillance system is and how it differs from a study: it runs continuously, it is designed for detection and monitoring rather than for testing a hypothesis, and its limits are built into how the data is collected. Your section may print this as NR 503 or NR503; 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-503 Week 3 asks for
A descriptive profile is not a list of numbers. It is an argument about pattern. Person means the characteristics that distribute the problem unevenly: age band, sex, race and ethnicity as reported by the source, occupation, insurance status, education. Place means where the pattern concentrates, at whatever geography the data supports, from region down to census tract. Time means trend, seasonality and any abrupt change worth asking about.
What earns the analysis row is noticing. Two age bands with a threefold difference, a rate that dropped in one year and stayed down, a concentration in three zip codes, a seasonal peak that does not match the national pattern. Each of those is an observation you can state and then generate a question from, and generating the question is the bridge to the analytic work that comes later in the session.
The stage also asks you to know your sources. Active and passive surveillance, notifiable disease reporting, vital statistics, national health surveys and disease registries all produce different data with different completeness. Deliverables at this point in an eight-week session commonly run three to five pages, sometimes with a table or a simple figure. If your section runs a discussion this week, draft the post outside the reply box, since posts do not reopen once submitted in Canvas.
The NR-503 Week 3 method, step by step
Six moves that turn a pile of published figures into a profile with a point.
-
Fix the case definition before you gather anything
Decide what counts as a case: diagnosed, self-reported, meeting a clinical threshold, appearing in a registry. Every figure you collect has to match that definition or be labelled as using a different one, and mixed definitions are the most common reason a profile stops adding up.
-
Choose the geography your data actually supports
Ambition here is expensive. If the surveillance system reports at county level, build the profile at county level rather than estimating a neighbourhood figure from it. Where you do estimate, say that you are estimating and from what.
-
Build person, place and time as three separate passes
Gather all the person data, then all the place data, then all the time data, rather than collecting whatever appears about your topic. Working in passes exposes the gaps, and a gap you can name is a limitation rather than a hole.
-
Look for the biggest contrast in each pass
One striking difference per dimension is worth more than eight modest ones. Find the age band, the place and the period where your problem behaves least like the average, and build the section around it.
-
Turn each contrast into a question, not a cause
Descriptive work generates hypotheses; it does not test them. Write the observation, then the question it raises, and stop there. The section that quietly starts explaining why the difference exists has left the territory the row is grading.
-
Say what the surveillance system could not see
Undiagnosed cases, people outside the reporting network, categories the system does not collect, lag between event and publication. A profile that names its own blind spots reads as competent rather than as incomplete.
A layout and word budget for a descriptive profile
Sized for a profile of roughly 1,200 to 1,500 words plus any table your section requires. It is our own drafting frame rather than a form the university issues, and your week's rubric outranks it wherever the two disagree.
| Section | What belongs in it | Word target |
|---|---|---|
| Problem and case definition | The condition, exactly what counts as a case, and the population the profile describes. | 130 to 170 |
| Data sources | Which systems supplied the figures, whether collection is active or passive, and what each system was built to do. | 180 to 220 |
| Person | Distribution across the characteristics the source reports, with the largest contrast identified and quantified. | 250 to 300 |
| Place | Geographic pattern at the level the data supports, with a comparison to a wider area. | 200 to 250 |
| Time | Trend across years, any seasonality, and any abrupt change, each with the period stated. | 200 to 250 |
| Questions raised and limits | Two or three hypotheses the pattern suggests, then what the data could not see. | 220 to 280 |
Evidence craft for surveillance data
Say how the data was collected, not only where it came from. Passive reporting depends on clinicians submitting cases and undercounts by design. Survey data depends on who answered and how the question was worded. One clause naming the collection method changes how every figure in the paragraph should be read.
Report categories exactly as the source defined them. Race, ethnicity, sex and age bands are constructed differently across systems, and quietly merging or renaming them makes your figures untraceable. Use the source's categories and note in a sentence where two sources are not comparable.
Give small numbers their counts. A rate calculated on a handful of cases moves wildly from year to year. Where the count behind a rate is small, write the count out and say that the estimate is unstable, because presenting it as a clean rate implies a precision the data does not have.
Attach the publication lag to every trend claim. If the most recent complete year is two years back, a sentence about the current situation is an extrapolation. Say the years the trend covers, then say what may have changed since and in which direction.
Five mistakes that cost points in this week's territory
- Numbers with no pattern drawn from them. Three paragraphs of figures and no statement of what varies most is a data dump rather than a profile.
- Causal explanation smuggled into the description. Saying a group has higher rates because of poor health behaviour crosses into a claim descriptive data cannot support.
- Sources mixed without checking definitions. Two systems counting different things produce a contrast that belongs to the definitions rather than to the population.
- Geography borrowed upward. Using a national figure to describe a county, without labelling it as a comparison, replaces your population with somebody else's.
- The system's blind spots unmentioned. Undiagnosed cases and people outside the reporting network are part of the profile, and leaving them out overstates what the data showed.
Before you submit
- One case definition governs every figure, and exceptions are labelled
- Each data source is named with its collection method
- Person, place and time each carry at least one quantified contrast
- Every observation is followed by a question rather than by an explanation
- Unstable estimates built on small counts are flagged as unstable
- A limitations passage names what the surveillance system could not capture
Building the profile this week?
Send the prompt, the rubric and your population. A premium original profile comes back in 24 to 48 hours with person, place and time built from named systems and the limits written in, revisions included.