MPH-507 Week 2 Needs Assessment and Data Sources: How to Write It

What a defensible needs assessment documents

Community health needs assessments published by hospital systems and health departments are public documents, and their methods appendix is more instructive than their findings. It lists the data sources with vintages, the geographic units they were available at, the number of key informant interviews and how those informants were selected, the survey instrument and its response rate, the focus groups and their recruitment route, and usually a short section on limitations that names which populations were underrepresented. That appendix is a quality audit of the assessment, and writing one is exactly the skill this stage teaches.

The conceptual work is distinguishing kinds of need. Normative need is what professionals or standards say a population should have, and it comes from benchmarks and guidelines. Expressed need is what people actually seek out, visible in service demand, waiting lists and utilization data. Felt need is what people say they need when asked, and it comes from surveys, interviews and focus groups. Comparative need is what a similar population has that this one does not. These four rarely point the same direction, and an assessment that reports only one of them has described a slice and called it a picture.

The second layer is capacity. A need is a gap between what exists and what is required, so an assessment that never inventories the services, workforce, funding, facilities and community assets already present cannot demonstrate a gap. Asset mapping matters here as more than a courtesy: a plan built as though a community has nothing will duplicate what is working, and it will fail politically before it fails technically because the organizations already doing the work will not support it.

Deliverables at this stage are usually a written assessment section, sometimes a data source table or a matrix of sources against the questions they answer, and often a posted response comparing methods. Posts in Canvas do not reopen after submission, so verify what a source actually measures before describing it publicly.

A method for assembling and reporting the evidence

Six moves that produce an assessment a reviewer can audit.

1. Write your assessment questions before touching any data

   Three to five questions, each answerable. How common is the condition in this population, who carries most of it, what services exist and what is their reach, what do residents say gets in the way. Data collected without a question first becomes a section nobody can use.

2. Match each question to the source type that can answer it

   Surveillance and vital statistics for magnitude, claims and program records for utilization, surveys and interviews for barriers and preferences, capacity inventories for supply. Build the matrix on paper; it exposes questions with no source behind them while there is still time.

3. Record vintage, geography and definition for every source

   Year of collection, the smallest geography available, and how the measure is defined. Two sources reporting the same indicator often define it differently, and comparing them without noticing produces a finding that is really a definitional artifact.

4. Collect the community's account with a described method

   Say how you would recruit, how many people, in what settings, in which languages, and what you would ask. A plan for primary collection is graded on design rather than on results, and vague plans to survey the community lose the row.

5. Inventory what already exists before naming a gap

   Programs, providers, hours, eligibility rules, capacity, waiting times, funding streams and informal community assets. The gap is the difference between this inventory and the need, and it cannot be asserted without the inventory in place.

6. Write the limitations as a method section, not an apology

   Which subgroups your sources undercount, which geographies were unavailable, which data are too old, and what direction each limitation biases the picture. This paragraph reads as competence and it protects your later objectives from being written against a distorted baseline.

A layout and word budget for a needs assessment section

Our frame for an assessment write-up, sized for roughly 1,200 to 1,500 words. It is our own outline rather than anything the university issues, and your section's scoring guide outranks it wherever the two disagree. A source matrix with columns for question, source, vintage, geography and limitation is the highest-yield table in this stage and will absorb around 200 words of the budget.

Evidence craft across mixed data sources

Describe what a source measures before you use its number. A survey of households and a registry of diagnosed cases produce different counts of the same condition for structural reasons, and neither is wrong. Write one clause explaining what the instrument captures, then give the figure. Readers who understand the instrument can weigh the finding; readers who do not will assume the number is more precise than it is.

Report response rates, sample sizes and confidence where the source provides them. Small-area estimates from national surveys are often modeled rather than directly measured, and they carry intervals that are wide enough to change conclusions. Quoting a point estimate from a modeled small-area product without its interval overstates what you know, and it is a routine finding on returned drafts.

Attribute qualitative material honestly and keep it de-identified. Interview and focus group content is evidence, and it should be reported with the recruitment method, the number of participants and the settings, without identifying anyone. Present it as what it is: an account of experience and barriers that quantitative sources cannot produce, not a measurement of prevalence.

Set a recency rule and apply it visibly. Where your guide is silent, treat baseline and utilization data older than five years as needing justification in the sentence, because a program will be evaluated against the baseline you chose. Structural and demographic context ages more slowly and can carry an older source, provided you say why it is still valid.

Five mistakes that cost points in this week's territory

• Data dumped without questions. Pages of tables with no stated assessment question read as effort rather than method, and no scoring row rewards effort alone.
• Only one kind of need reported. An assessment built entirely on surveillance figures has never asked the population anything, and one built only on interviews has no magnitude.
• No capacity inventory. Without knowing what exists, a gap cannot be demonstrated, only claimed, and the recommendation risks duplicating a working program.
• Sources compared across different definitions. Two figures for the same indicator that were built differently produce a false trend or a false disparity.
• Limitations reduced to a sentence at the end. If the limitation does not say who is undercounted and which way it skews the picture, it has done nothing for the reader.

Before you submit

• Three to five assessment questions appear before any data
• Every source carries its system name, vintage and geography
• At least two different kinds of need are documented
• Primary collection is described as a design rather than an intention
• Existing services and assets are inventoried before the gap is claimed
• Limitations name the undercounted groups and the direction of the bias

Questions students ask about this stage

Keep going