brycewang-stanford/Auto-Empirical-Research-Skills / skills/17-DAAF-Contribution-Community-daaf/dot-claude/skills/education-data-source-meps
education-data-source-meps
VerifiedMEPS — Urban Institute modeled school-level poverty (% at 100% FPL), from CCD + SAIPE (public schools, 2009-2022, 2-3yr lag). Use when FRPL is unreliable due to CEP. Consistent cross-state measurement. Public schools only.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 13.8 KB |
| references/comparison-to-frpl.md | markdown | 8.0 KB |
| references/data-quality.md | markdown | 10.8 KB |
| references/data-sources.md | markdown | 8.3 KB |
| references/methodology.md | markdown | 7.2 KB |
| references/variable-definitions.md | markdown | 10.7 KB |
references/methodology.md
View on GitHub# MEPS Methodology
How Model Estimates of Poverty in Schools (MEPS) are calculated, validated, and should be interpreted for research purposes.
## Overview of the Modeling Approach
MEPS uses a **linear probability model** to estimate the share of students in each school from households with incomes at or below **100% of the Federal Poverty Level (FPL)**.
### Core Concept
The model leverages two key data sources:
1. **School-level data** from CCD (enrollment, FRPL counts where reliable)
2. **District-level poverty** from Census SAIPE (ground truth for calibration)
The model estimates school-level poverty while ensuring district totals align with Census estimates.
## The MEPS Model
### Model Specification
MEPS uses a linear probability model of the form:
```
Poverty_school = α + β₁(FRPL_school) + β₂(Demographics) + β₃(Geography) + ε
```
Where:
- **Poverty_school**: Estimated share of students in poverty (0-1)
- **FRPL_school**: School-level FRPL rate (where reliable)
- **Demographics**: School demographic characteristics from CCD
- **Geography**: District-level variables including SAIPE poverty rate
- **ε**: Error term
### Key Methodological Features
1. **District-level calibration**: School estimates are constrained to sum to the Census SAIPE district estimate when weighted by enrollment
2. **FRPL adjustment**: The model accounts for systematic differences between FRPL and true poverty rates, which vary by:
- State (due to policy differences)
- Year (due to CEP adoption)
- School type (charter vs traditional)
3. **Enrollment weighting**: Estimates are generated to be representative of student populations, not just school counts
## Model Training and Estimation
### Training Data
The model is trained on schools where reliable poverty information exists:
- Schools in states with consistent FRPL reporting
- Pre-CEP time periods where FRPL more closely tracked poverty
- Validation against independent poverty measures
### Estimation Process
1. **Estimate model coefficients** using training data
2. **Generate predicted values** for all schools
3. **Apply district constraints** so school estimates sum to SAIPE
4. **Calculate standard errors** for uncertainty quantification
## Modified MEPS (`meps_mod`)
### Why Modified MEPS Exists
The original MEPS model tends to **underestimate poverty** in districts with very high poverty rates. This occurs because:
- The linear model compresses extreme values
- Limited variation in predictors for very-high-poverty districts
### Modified MEPS Calculation
Modified MEPS adjusts estimates for schools in districts where the model systematically underestimates:
```
meps_mod = meps + adjustment_factor
```
The adjustment factor is larger for schools in districts where:
- District SAIPE poverty is very high (>30%)
- Original MEPS substantially underestimates district totals
### When to Use Modified MEPS
| Scenario | Recommendation |
|----------|---------------|
| General analysis | Use original `meps_poverty_pct` |
| Focus on high-poverty schools/districts | Consider `meps_mod_poverty_pct` |
| Comparing across poverty levels | Use original `meps_poverty_pct` for consistency |
| Policy analysis for Title I | Consider `meps_mod_poverty_pct` for high-poverty targeting |
## Validation Evidence
### Correlation with SAIPE
At the district level, MEPS aggregates correlate strongly with Census SAIPE:
- **Correlation**: 0.85-0.95 depending on year
- **Mean absolute error**: ~3-5 percentage points
### Comparison with Administrative Data
In states with reliable administrative data (e.g., tax records), MEPS shows:
- Strong correlation with actual poverty measures
- Better cross-state comparability than FRPL
### Face Validity
MEPS estimates are consistent with expected patterns:
- Higher in urban core and rural areas
- Lower in suburban areas
- Correlates with neighborhood characteristics
## Standard Errors (`meps_se`)
### Interpretation
The standard error (`meps_se`) quantifies estimation uncertainty:
- **68% confidence interval**: meps ± 1 × meps_se
- **95% confidence interval**: meps ± 1.96 × meps_se
### Factors Affecting Standard Errors
Standard errors tend to be larger for:
- Smaller schools (less data for estimation)
- Schools in districts with unusual characteristics
- Schools with missing predictor variables
### Using Standard Errors in Research
```python
import polars as pl
# Identify schools where estimates are precise (SE < 2 percentage points)
precise_schools = df.filter(pl.col("meps_poverty_se") < 2.0)
# Flag close comparisons between two schools
def compare_schools(pct_a: float, se_a: float, pct_b: float, se_b: float) -> str:
diff = abs(pct_a - pct_b)
se_combined = (se_a**2 + se_b**2)**0.5
if diff < 1.96 * se_combined:
return "Not statistically different"
return "Statistically different"
```
## MEPS 2.0 Updates
### Key Improvements in MEPS 2.0 (December 2025)
> **Portal Status (Feb 2026):** MEPS 2.0 has not yet been integrated into the Education Data Portal mirrors. Portal data still reflects MEPS 1.0 (2009-2022). The improvements below apply to the Urban Institute's direct release, not yet to mirror-accessible data.
1. **Extended time coverage**: Additional years of data
2. **ISP integration**: Incorporates Identified Student Percentage data
3. **Improved CEP handling**: Better accounts for universal meal schools
4. **Updated calibration**: Uses more recent SAIPE data
### ISP (Identified Student Percentage)
MEPS 2.0 uses ISP data, which identifies students directly certified for free meals through:
- SNAP (food stamps)
- TANF (cash assistance)
- Medicaid (in some states)
- Foster care, homeless, migrant, runaway status
ISP provides a more direct poverty signal than FRPL application forms.
## Technical Notes
### Sample Restrictions
MEPS estimates are generated for:
- **Public schools only** (no private schools)
- **Regular schools** (excludes special education-only, alternative schools vary)
- **Schools with enrollment data** in CCD
### Missing Data Handling
- Schools missing key predictors may have imputed values
- Standard errors account for imputation uncertainty
- Flagged variables indicate imputation status
### Replication
The MEPS methodology is documented in:
- Gutierrez, Blagg, & Chingos (2022). "Model Estimates of Poverty in Schools." Urban Institute.
- Gutierrez & Blagg (2025). "Model Estimates of Poverty in Schools 2.0." Urban Institute.
## Methodological Limitations
1. **Model assumptions**: Linear model may not capture all relationships
2. **Data quality**: Depends on accuracy of CCD and SAIPE inputs
3. **Temporal stability**: Model coefficients assumed stable over time
4. **Private schools**: Not covered by MEPS
5. **Within-district variation**: Model captures some but not all variation
## Summary
MEPS provides a **rigorous, validated, and comparable** school-level poverty measure by:
- Using a statistical model calibrated to Census data
- Accounting for FRPL inconsistencies across states and time
- Providing uncertainty estimates via standard errors
- Offering modified estimates for high-poverty contexts
For most research purposes, MEPS provides a superior measure of school poverty compared to raw FRPL data, especially for cross-state comparisons or analyses spanning the CEP era (2010+).
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |