Methodology
This page explains every calculation in NourishScope, version 1.0.0, Methodology 1.0. It follows the NourishScope Calculation Specification written by the founder, Yetunde Khadijat Olaogun. All calculations run in your browser.
1. Data the tool combines
- Kentucky public data (default). All 120 counties, compiled from the U.S. Census Bureau, CDC PLACES, the USDA Economic Research Service Food Environment Atlas and the Appalachian Regional Commission. See Data Sources.
- Sample data. Six fictional communities, always labelled “Sample data”.
- Your upload. A CSV in the template format. Each row is matched to a Kentucky county by FIPS code, or by county name. Where you fill in a value, it replaces the public value for that field only; every other field keeps the public data.
2. Inputs and defaults
- Number of households in need
- For Kentucky county data this starts as the number of households receiving SNAP (Census ACS 5-year, 2020-2024). You may replace it. Required, a whole number of 0 or more.
- Number of households served
- No public default; blank unless you enter it. A whole number of 0 or more. If it is more than households in need, the tool warns you and sets households still needing help to 0.
- Cost per household
- Required. Pre-filled with $100, labelled “Example only. Replace it with your program’s cost.” One value for all communities, unless your upload gives a cost for a community. The planning brief cannot be saved until you confirm or change the cost.
- Budget (B)
- Required, more than $0.
- Minimum for every community
- Used by “Share by need, with a minimum for all”. Default 5%.
- Health measures
- Diabetes, obesity and high blood pressure for each county (CDC PLACES, age-adjusted adult prevalence). Optional in uploads.
Values that are negative or not numbers are rejected with a message naming the row and field. “-9999”, empty cells and “Not reported” are treated as missing, never as zero.
3. Core measures for each community
- Households still needing help (U) = the larger of (households in need − households served) and 0. If households served is blank, U equals households in need and the tool runs in “need-only” mode.
- Coverage is shown only when households served is entered for every selected community: total served ÷ total in need × 100. In need-only mode the tool shows “Not entered” and asks you to enter households served.
- Health burden score (H) = the average of (county rate ÷ Kentucky median) across the available health measures, or 1.0 if none are available. The Kentucky medians are fixed from all 120 counties: diabetes 12.3%, obesity 40.3%, high blood pressure 36.7%.
4. The five strategies
For n selected communities and budget B. If the total of U (or W) is 0, the tool falls back to an equal split. Each allocation is rounded to the cent.
Equal split
Every community gets the same amount. Allocation = B ÷ n.
Share by unmet need
Communities with more households still needing help get more. Allocation = B × U ÷ (total U).
Share by need, with a minimum for all
Every community gets a small guaranteed amount, and the rest follows unmet need. Floor F = floor share × B ÷ n. Allocation = F + (B − n × F) × U ÷ (total U).
Share by need and health burden
Like sharing by need, but communities with higher diabetes, obesity and blood-pressure rates get more. Weight W = U × H. Allocation = B × W ÷ (total W).
Lowest cost first
Funds the communities where each household costs least, until the money runs out. Most useful when costs differ by community. Communities are sorted by cost per household, lowest first; ties go to the higher health burden score, then to more households still needing help. Each gets the smaller of the remaining budget and U × cost, until the budget is used.
5. Results for each community and in total
- Households reached = the smaller of (allocation ÷ cost, rounded down) and U. A partly funded household is not counted.
- Spend = households reached × cost. Unspent = B − total spend. Version 1.0 reports unspent money; it does not move it to other communities.
- Coverage after (when households served is entered for every community) = (total served + total reached) ÷ total in need × 100.
A worked example with real Kentucky data
Suppose a program has $50,000.00 to serve three eastern Kentucky counties, all rated Distressed by the Appalachian Regional Commission. Need is the number of households receiving SNAP, and the cost per household is an example value of $100.00.
| County | Households in needSource: Census ACS 5-year, 2020-2024 | Health burden scoreSource: CDC PLACES, 2021-2022 | Allocation | Households reached |
|---|---|---|---|---|
| Magoffin County | 1,759 | 1.15 | $26,427.28 | 264 |
| Owsley County | 593 | 1.07 | $8,909.25 | 89 |
| Wolfe County | 976 | 1.19 | $14,663.46 | 146 |
The same budget, divided the five ways the tool compares:
| Strategy | Households reached | Unspent |
|---|---|---|
| Equal split | 498 | $200.00 |
| Share by unmet need | 499 | $100.00 |
| Share by need, with a minimum for all | 498 | $200.00 |
| Share by need and health burden | 498 | $200.00 |
| Lowest cost first | 500 | $0.00 |
Health burden score: diabetes, obesity and high blood pressure compared with the Kentucky middle.Source: CDC PLACES, 2021-2022. Calculated with NourishScope Methodology 1.0.
6. Cost sensitivity
The tool repeats the allocation and results with every cost multiplied by 0.8 and by 1.2, keeping the budget the same, and shows households reached and unspent for −20%, as entered, and +20%.
7. Equity flags
A community is flagged as rural or distressed if its county’s Appalachian Regional Commission status for fiscal year 2027 is Distressed or At-Risk, or its county is nonmetro (USDA ERS, 2023). The tool reports the share of the budget, and of households reached, that goes to flagged communities.
8. Evaluation module
- Change = post % − pre %, in percentage points.
- Participants improving = participants × change ÷ 100.
- Cost per participant improving = program cost ÷ participants improving. Not shown when no participants improved.
9. Where the tool differs from the specification
- Unmatched upload rows are flagged, not added. NourishScope covers Kentucky only, so a row that does not match a Kentucky county is shown for you to fix (choose the county, or correct the file) rather than added as a new community without public data.
- County names are matched with leeway. Case, spacing, punctuation and a trailing “County” or “Co.” are ignored, so “Adair”, “adair county” and “ADAIR CO.” all match Adair County. A FIPS code that disagrees with the county name is flagged.
- Upload column names. The template uses
number_of_households_in_needandnumber_of_households_served;households_in_needandhouseholds_servedare also accepted. - A blank need in an upload is flagged. The SNAP-households default applies only to Kentucky county data, so two uploaded communities in one county never both inherit the county’s full count.
10. Assumptions and limitations
- SNAP households estimate need. They are not a count of households eligible for, or interested in, your program.
- Health and food-access figures describe the whole county, not your participants.
- Public figures come from different years (2018 to 2024). Each figure in the tool shows its source and year.
- CDC PLACES 2024 release is used because the 2025 release does not include these measures for Kentucky.
- USDA does not report grocery stores per 1,000 residents for 42 small counties; the tool shows “Not reported (suppressed by source)” for those, never zero.
- Results support a decision; they do not replace local knowledge of your communities.