Assessing Need & Recognizing Peer Progress: An Additional Way to Understand Maryland School Ratings
Every year, the Maryland State Department of Education (MSDE) gives public schools a rating from 1 to 5 stars against fixed statewide benchmarks. That official rating is an essential tool for assessing where students and schools need the most support and resources. This project offers a complementary Peer Comparison Star Rating—not a replacement—that shows how schools perform compared to peers serving similar levels of student poverty, helping communities see both where resources are needed and where schools are excelling given their student demographics.
Answering Two Different Questions: Where Support Is Needed Most & How Schools Compare to Similar Peers
In Maryland's official Report Card system, every school is evaluated against the same fixed statewide thresholds: 75% of points for 5 stars, 60% for 4 stars, 45% for 3 stars, and 30% for 2 stars. Holding all schools to a common statewide benchmark is valuable for identifying where achievement gaps exist and directing resources and interventions to communities with the greatest need.
At the same time, students in high-poverty neighborhoods face real-life challenges outside of school—such as food insecurity, unstable housing, and fewer out-of-school academic resources. Because out-of-school poverty strongly affects test scores and attendance, schools serving high-poverty communities can achieve strong student growth while still falling short of a fixed statewide cutoff. In 2025–26, 60.1% of the difference in Elementary school scores (and 50.7% in High schools) was associated with a single factor: the percentage of students facing Economic Disadvantage (ED%).
Because fixed cutoffs reflect both school performance and neighborhood poverty, adding a complementary Peer Comparison Star Rating lets us answer a second vital question: How well is a school serving its students compared to other Maryland schools with similar demographics?
Interactive Simulator: Compare Standard & Peer Comparison Star Ratings
Live CalculatorMove the sliders below to see how a school's Economic Disadvantage (ED%) and Points Earned (%) affect its official state star and its peer comparison star:
How Adding a Peer Comparison Star Rating Supports Students, Families, and Educators
While MSDE's standard star rating helps state and district leaders target funding and academic support to high-need schools, providing a complementary Peer Comparison Star Rating alongside it offers vital perspective for students, families, and staff:
Celebrating Hard Work & Building School Pride
When students in a high-poverty school work hard every day, improve their reading and math skills, and beat the odds, they deserve to see their growth recognized alongside standard state benchmarks. Seeing their school's peer comparison star rating alongside state benchmarks builds student pride and motivation.
Seeing Both Resource Needs & "Hidden Gem" Schools
Families benefit from seeing both ratings side by side: the official MSDE rating shows where a school stands relative to statewide goals and where extra community investment is needed, while the Peer Comparison Star Rating highlights neighborhood schools where children are making exceptional progress compared to similar peers.
Supporting Educator Retention & Peer Learning
Teachers, principals, and support staff in high-need schools pour their energy into supporting students. Pairing MSDE's official rating with a Peer Comparison Star Rating recognizes effective instruction in high-need classrooms, reduces educator burnout, and helps districts identify peer exemplar schools to learn from.
What Happens When We Adjust Maryland's Star Ratings Across 1,292 Schools?
High-poverty schools gain nearly a full star on average (+0.74★) across 316 schools, while the statewide average remains virtually identical.
Among Maryland's 316 high-poverty schools, 5-star ratings jump from just 5 schools to 46 schools for beating expectations.
More than half of high-poverty schools (56.0%) earn 4 or 5 stars (up from 18.0%) when compared against similar peer schools.
Three-quarters of 1★ and 2★ high-poverty schools move up, with 69.6% of high-poverty schools gaining at least 1 star.
Before vs. After: Star Ratings for High-Poverty Schools (≥60% Low-Income, n=316 Schools)
Compare how many schools earn 1★, 2★, 3★, 4★, and 5★ under both ratings (Use buttons above to switch cohorts):
What This Means for School Districts Across Maryland (Average Star Shift)
Districts serving higher-poverty communities see how their schools perform compared to similar schools:
Spotlight Schools: Outperforming Similar Schools Across Maryland
When we compare schools against peers serving similar student populations, communities across Maryland—such as SEED School (+1.50★ average gain), Kent County (+0.80★), Baltimore City (+0.68★), Somerset County (+0.62★), Allegany County (+0.55★), Wicomico County (+0.50★), Baltimore County (+0.48★), and Worcester County (+0.46★)—see dozens of hard-working schools recognized as 4-star and 5-star peer leaders. Click any school below to inspect both ratings on its full diagnostic card:
Explore the Interactive Maryland Report Card Dashboard
Choose any section below to open interactive charts, district summaries, school lookups, and guides:
Official State Star Ratings (Resource Need Perspective)
Fixed thresholds at 30%, 45%, 60%, and 75% of possible points (Click any dot to inspect school)
Peer Comparison Star Ratings (Compared to Similar Schools)
Color-coded by peer comparison star rating with peer cutoff lines (Click any dot to inspect school)
Star Rating Changes (State Star vs. Peer Comparison Star)
Rows show Official State Stars; columns show Peer Comparison Stars under the selected model:
| State ↓ \ Peer Comparison → | 1★ | 2★ | 3★ | 4★ | 5★ | Total Orig |
|---|
Highlighted High-Poverty Exemplar Schools
Click any exemplar school below to open its full Accountability Diagnostic Card:
Accountability Indicator Poverty Gradients (BERC 2019 Framework)
Testing which components of Maryland's ESSA model measure out-of-school poverty versus in-school educator impact
A foundational finding of Wohn & Boberiene (2019) is that Maryland's accountability indicators do not relate equally to student poverty. Standardized test proficiency (Academic Achievement) tracks neighborhood socioeconomic status with extraordinary severity (R² ≈ 70% in Elementary), reflecting out-of-school family wealth. Conversely, annual student gains (Academic Progress / SGPs) have a much flatter gradient (R² ≈ 27%), proving that high growth is achieved in high-poverty schools every year. In addition, student attendance (Chronic Absenteeism) heavily tracks poverty (R² ≈ 55%), penalizing schools for community healthcare, housing, and transportation barriers.
Complete Indicator-by-Indicator Poverty Gradient Reference Table
Linear regression of each accountability indicator on Economic Disadvantage (%) for the currently filtered cohort (Click any school in the plots above to inspect)
| Report Card Indicator / Metric | Bands | Reporting (n) | Mean Score | Pearson r | R² (ED%) | Slope (Δ per 10% ED) | Fitted OLS Equation | Accountability / Equity Takeaway |
|---|
District Star Rating Changes (Peer Comparison Star minus State Star)
Comparing district-level impact across Maryland school districts (24 LEAs; SEED School included when specialized programs are toggled on)
Maryland School Districts Summary (Click any district to view its schools)
Click any column header or sort menu to reorder| LEA Name | Schools (n) | Mean ED% | Mean Pts% | Avg State Star | Avg Peer Comparison Star | 2026 Shift | 5★ (State → Peer) | 1–2★ (State → Peer) |
|---|
Maryland School Performance Directory
Showing 1–50 of 1,292 schools (Click any column header to sort • Click any school row for full diagnostic card)
| District | School Name (ID) | Level | State Score (%) | Peer Comparison Score (%) | Poverty (ED%) | Special Ed (SWD%) | English Learner (ML%) | State Star | Peer Comparison Star | Gain | Vs. Peers (+/-) |
|---|
User Guide: Exploring Maryland's Equitable Report Card System
Step-by-step instructions for navigating the public-facing Data Storyboard, comparing adjustment models, filtering districts, grade bands, and poverty ranges, and inspecting individual school diagnostic cards.
Opening Tab: Data Storyboard (Public-Facing Landing Page & Launchpad)
When opened, the application loads directly to the full-width ★ Data Storyboard—designed as an accessible, public-facing title and overview page (written at a middle-school reading level without the technical sidebar or KPI header cards) that positions Peer Comparison Star Ratings as a complementary perspective alongside MSDE's official star ratings:
Choose a Comparison Model (Left Sidebar)
On any analytical tab (Tabs 1–5), use the Choose Comparison Model dropdown in the left sidebar to switch between six accountability models in real time. Every model dynamically updates both the Peer Comparison Star Rating (1★–5★) and the Peer Comparison Score Percentage:
- 1. Peer Comparison Model (2019 BERC • Recommended): Compares each school against peers with similar student poverty rates, setting peer comparison star cutoffs that preserve exact statewide star totals.
- 2A. Straight-Line Peer Comparison (OLS Residuals): Draws one statewide trend line and ranks schools by how many points they score above or below similar schools. (93% match with Model 1).
- 2B. Standardized Bell-Curve Score: Rescales scores to prevent bunching in the middle, then applies standard 30/45/60/75% cutoffs.
- 2C. Curved Trend Line (Monotone Spline): Uses a smooth curve that bends to account for extreme poverty challenges (above 80% ED) without ever curving upward.
- 3. Full Student Background: Simultaneously accounts for economic disadvantage, students with disabilities, and multilingual learners.
- 4. Interactive Custom Weight Simulator: Unlocks three sliders so you can test custom policy weights on economic disadvantage, students with disabilities, and multilingual learners.
Filter by School Level, District, & Poverty Range
Every chart, KPI card, transition matrix, and lookup table updates immediately when you adjust the cohort filters in the left sidebar:
- Exclude Special Ed, Alternative & SEED Schools (Default On): Compares traditional public schools (1,292 schools). Uncheck at any time to include all specialized centers, alternative programs, and the SEED School (1,311 schools).
- School Level Buttons: Toggle between All Schools (1,292), Elementary (760), Middle (216), High (193), or Combined (123). Regressions are performed within each grade band to maintain statistical accuracy, and multi-band schools (E/M, M/H, E/M/H) are combined proportionally based on total possible points so each school appears once.
- District (LEA) & Quick Spotlights: Filter to any of Maryland's 24 school districts (or the SEED School when specialized schools are toggled on), or use the quick buttons for ★ Spotlight: Baltimore City, High Poverty (≥60% ED), or Low Poverty (≤20% ED).
- Poverty Range Sliders: Drag the Min ED% and Max ED% sliders to isolate any specific student poverty band.
- School Search: Type any school name or 4-digit MSDE ID to instantly find specific schools across all views.
Explore the Three Analytical Views (Tabs 1–3)
- Tab 1 — Interactive Chart (Two Modes):
- Mode 1 — Star Ratings (State vs. Peer Comparison): Side-by-side comparison of official state horizontal thresholds on the left versus peer comparison thresholds on the right, plus the Star Rating Changes Matrix and High-Poverty Exemplar Schools.
- Mode 2 — What Drives Scores? (Test Scores vs. Student Growth): Visually compare how test scores (R² ≈ 70%) track out-of-school family poverty, while annual academic progress (R² ≈ 27%) reflects classroom teaching.
- Tab 2 — School Districts: Bar chart and complete table comparing district-level star ratings, 5★ counts, and shifts across Maryland's 24 school districts. Tip: Click any district row to view its schools in Tab 3!
- Tab 3 — School Directory: Searchable, sortable directory of all Maryland schools featuring State Score (%), Peer Comparison Score (%), State Star, Peer Comparison Star, and Vs. Peers (+/-). Click ↓ Export Filtered CSV to download your filtered view to Excel.
Inspect Individual School Diagnostic Cards
Click any school dot on either scatter plot, any exemplar school card on the Storyboard or Tab 1, or any row in the Tab 3 School Lookup table to open that school's pop-up Accountability Diagnostic Card:
- School Demographics & Student Context: Displays school type and student subgroup context including ED%, SWD%, and ML%.
- Side-by-Side Model Comparison (Stars + Est. % Pts): Displays both the Star Rating (1★–5★) and the Model-Specific Estimated % Points across all six specifications simultaneously (Original MSDE, BERC QR, OLS Rank, OLS Standardized, Monotone I-Spline, Multivariable, and Custom Weight).
- "Nearest Demographic Peers" Comparison Table: Automatically identifies the 6 Maryland schools serving the most similar student poverty rate (ED%, breaking ties by SWD% and ML%) and ranks them by Peer Comparison Points % alongside the selected school. Click any peer row to jump directly to its diagnostic card.
- Combined Multi-Band School Rollup (ΣEarned / ΣPossible): For any of Maryland's 124 traditional multi-band schools (136 when non-traditional schools are included) (E/M, M/H, E/M/H), automatically displays a complete grade-band breakdown table showing how each individual grade level's points sum together to determine the overall school-wide percentage and star rating.
- School-Specific Quantile Cutoffs, Full Indicator Breakdown & Printable / Shareable Reports: Displays the exact points percentage required for 2★, 3★, 4★, and 5★ at that school's ED% alongside full, unabbreviated MSDE accountability indicators and every individual metric, plus header buttons to Copy Direct Link (#school=LEA_SCH_SPAN) or Print / Save PDF as a clean report.
How the Math Works: Adjusting Maryland School Ratings
You don't need a graduate degree in statistics to understand how we adjust school ratings. If you've taken high school Algebra or Intro to Statistics, you already know the core concepts: scatter plots, lines of best fit, and residuals. Here is how the entire system works from start to finish.
The 6 Core Statistics Concepts
Concept 1: The Trend Line (y = mx + b)
Just like in high school Algebra, we draw a line of best fit across all Maryland schools:
- x-axis: Percentage of low-income students (0% to 100%).
- y-axis: Points earned on the state report card (0% to 100%).
- Slope (m = −0.39): For every 10% increase in student poverty, a school's score drops by about 4 points on average across Maryland.
Concept 2: What is R²? ("The Wealth Effect")
In statistics, R² (R-squared) tells us how much of the difference in test scores is tied directly to family income:
What this means: Over 60% of an elementary school's score is driven by neighborhood wealth, food security, and home stability—not what teachers do in the classroom.
Concept 3: What is a "Residual"?
A residual is simply the vertical distance between where a school scored and what the trend line predicted:
- Dot above the line (+): The school is beating expectations compared to similar peers!
- Dot below the line (−): The school has room to improve compared to similar peers.
Concept 4: No Grade Inflation!
We don't hand out extra stars. We set adjusted cutoffs so that the statewide number of stars stays 100% identical:
Maryland gives 5★ to the top ~10% of schools and 4★ to the next ~40%. By matching those exact percentiles, stars simply shift to schools that outperform peers.
Concept 5: The 0–100% Score (Mapping)
To make numbers easy to read, we convert each school's peer standing onto Maryland's familiar 15-point star scale:
Just like converting Celsius to Fahrenheit, this doesn't change a school's rank—it translates their standing into intuitive, familiar percentages.
Concept 6: Combined Schools (K–8 / 6–12)
What about schools that serve both Elementary and Middle grades?
Just like calculating your GPA (where a 4-credit course counts more than a 2-credit course), each grade band is evaluated against its own peers first, then combined proportionally based on total points.
Step-by-Step Walkthrough: Harlem Park Elementary/Middle
Let's see how the math works for a real Maryland school: the Elementary span of Harlem Park Elementary/Middle in Baltimore City:
The 6 Comparison Models (At a Glance)
In the sidebar dropdown on the analytical tabs, you can test six different statistical specifications. Here is what each one does:
How it works: Draws separate, custom slope lines for each star level. Allows higher and lower poverty thresholds to curve naturally. This is the primary model used in the 2019 BERC study.
How it works: Draws one single straight trend line through all schools, calculates each school's distance from the line, and ranks schools by how far they score above or below similar schools. (93.8% identical to Model 1).
How it works: Prevents scores from bunching in the middle by standardizing the spread of scores before applying standard 30/45/60/75% cutoffs.
How it works: Uses a flexible curved line that bends to account for the unique challenges of extreme concentrated poverty (over 80% low-income), while ensuring the line never goes up.
How it works: Adjusts for three student challenges at once: Economic Disadvantage (ED%), Students with Disabilities (SWD%), and Multilingual Learners (ML%).
How it works: Unlocks three interactive sliders in the sidebar so you can choose your own adjustment strength for economic disadvantage, students with disabilities, and multilingual learners.
📐
4. Advanced Mathematical Specifications & Technical Deep Dives
▼ Click to Expand
For researchers, econometricians, and policymakers requiring exact mathematical formulations, piecewise-linear mapping equations, and multi-band rollup proofs, the complete technical documentation is provided below:
A. Bivariate Ordinary Least Squares (OLS) Formulation
- Yi : Observed unadjusted Total Points % (Earned / Possible × 100).
- Ŷi : Expected Points % for school i given student poverty EDi%.
- ei = Yi − Ŷi : Peer comparison residual measuring performance above/below statewide expectation for similar schools.
B. Quantile Regression Check-Loss Formulation (Model 1)
Evaluated at empirical cumulative proportions τ ∈ {τ1,b, τ2,b, τ3,b, τ4,b} matching MSDE's 1★, 2★, 3★, 4★, and 5★ statewide distribution.
C. Monotone Decreasing I-Spline (Model 2C) & Multivariable OLS (Model 3)
D. Piecewise-Linear Mapping Formula onto MSDE's 0–100% Scale
Under MSDE's unadjusted system, star cutoffs are spaced exactly 15 percentage points apart (30%, 45%, 60%, 75%). To translate each model's metric into an intuitive Estimated Adjusted Points Percentage (Adj_Pct) where 30%, 45%, 60%, and 75% retain identical meaning across all models, we apply monotonic piecewise-linear interpolation:
E. Combined Grade Band Rollup (COMB, N = 1,292 Traditional / 1,311 Full): Mathematical Formulation
In Maryland's traditional cohort, 123 multi-band schools serve 2 or 3 grade levels inside the same facility (104 E/M, 14 M/H, 5 E/M/H). Because Elementary, Middle, and High schools have different indicator mixes, points possible, and baseline distributions, regressions are always estimated within Elementary, Middle, and High cohorts first, and then combined into whole-school ratings weighted by MSDE Possible Points (Possibleb):
1. Unadjusted Combined Score
Sums earned and possible points directly, then compares to standard 30/45/60/75% cutoffs.
2. Threshold Point Summation
Sums raw required cutoff points across bands first, then maps once to MSDE scale.
3. Possible-Points-Weighted Adj %
Maps each band to MSDE scale first, then computes weighted average of adjusted points.
- Elementary: 42.0/90 (46.6%, 3★ unadj), ED = 93.2% → Adj 4★ (66.6%)
- Middle: 51.9/90 (57.7%, 3★ unadj), ED = 90.4% → Adj 5★ (75.0%)
- Elementary (52.6% wt): 38.9/100 (38.9%, 2★ unadj), ED = 83.0% → Adj 3★ (51.0%)
- Middle (47.4% wt): 46.1/90 (51.2%, 3★ unadj), ED = 86.3% → Adj 4★ (66.9%)
5. Complete 2025–26 Grade-Band Parameter & Quantile Cutoff Reference Table (Default Traditional Cohort, N = 1,420 spans across 1,292 schools)
Parameters below reflect the default cohort excluding Special Education Centers, Alternative Education Schools, and the SEED School (N = 1,420 spans across 1,292 schools). When the sidebar toggle is unchecked (N = 1,457 spans across 1,311 schools), the application automatically switches to the full-cohort parameter table (BAND_META_ALL).
| Parameter / Threshold Equation | Elementary Band (E, n = 869) | Middle Band (M, n = 339) | High School Band (H, n = 212) |
|---|---|---|---|
| Unadjusted Star Proportions (1★ / 2★ / 3★ / 4★ / 5★) | 0.1% / 7.6% / 33.3% / 46.6% / 12.4% | 1.2% / 19.1% / 48.5% / 28.2% / 2.9% | 0.0% / 17.8% / 41.8% / 30.0% / 10.3% |
| Cumulative Quantile Cutoffs (τ1, τ2, τ3, τ4) | 0.0012, 0.0771, 0.4097, 0.8757 | 0.0118, 0.2029, 0.6882, 0.9706 | 0.0000*, 0.1784, 0.5962, 0.8967 |
| Bivariate OLS Equation (Points% ~ ED%) | 79.99 − 0.3921×ED% (R²=60.1%) | 70.35 − 0.3412×ED% (R²=44.7%) | 76.73 − 0.4408×ED% (R²=50.7%) |
| QR 4★/5★ Cutoff Line (Q̂τ4) | 85.30 − 0.3310×ED% | 82.76 − 0.2774×ED% | 84.69 − 0.4085×ED% |
| QR 3★/4★ Cutoff Line (Q̂τ3) | 79.54 − 0.4196×ED% | 73.46 − 0.3319×ED% | 77.86 − 0.4395×ED% |
| QR 2★/3★ Cutoff Line (Q̂τ2) | 73.46 − 0.4628×ED% | 64.52 − 0.3603×ED% | 74.01 − 0.5473×ED% |
| QR 1★/2★ Cutoff Line (Q̂τ1) | 55.04 − 0.3270×ED% | 55.81 − 0.3859×ED% | 51.03 − 0.4408×ED%* |
| Monotone I-Spline Knot Ŷ (0/20/40/60/80/100% ED) | 82.3 / 72.1 / 63.0 / 57.1 / 50.4 / 28.2 (R²=61.3%) | 76.5 / 64.4 / 55.1 / 48.6 / 44.5 / 44.5 (R²=47.1%) | 81.2 / 69.6 / 56.4 / 50.0 / 44.1 / 44.1 (R²=53.5%) |
| Monotone I-Spline Slopes (per 20% ED bracket) | −0.51, −0.46, −0.29, −0.33, −1.11 | −0.60, −0.46, −0.32, −0.21, 0.00 | −0.58, −0.66, −0.32, −0.30, 0.00 |
| Multivariable OLS (ED%, SWD%, ML%) | 80.12 − 0.387·ED − 0.000·SWD − 0.022·ML | 74.57 − 0.279·ED − 0.341·SWD − 0.211·ML | 82.58 − 0.286·ED − 0.753·SWD − 0.349·ML |
6. Citations, Public Data Sources, & Author Credits
-
•
Foundational BERC Research Brief (June 2019):
Wohn, C., & Boberiene, L. (2019). Looking at School Performance From a Different Perspective: A Maryland Report Card Reflecting Students Served. Baltimore Education Research Consortium (BERC). Available at baltimore-berc.org (Full PDF Brief). -
•
Official Maryland Report Card Public Website (MSDE):
Maryland State Department of Education (2026). 2025–26 Maryland Report Card Public Accountability Dataset & School Performance Reports. Available at reportcard.msde.maryland.gov. -
•
Supporting Literature Referenced in the BERC Framework:
Berner, A., & Steiner, D. (2019). Quality and schools: Managing school closings to optimize student outcomes (BERC); Hasan, S., & Kumar, A. (2018). Digitization and divergence: Online school ratings and segregation in America; Reardon, S. F. (2011). The widening academic achievement gap between the rich and poor. -
•
AI Coding & Web Implementation Assistance:
Google. (2026). Gemini [Large language model]. https://gemini.google.com/
Conceptual framework, statistical modeling, and analysis by Christopher Wohn, building on the June 2019 Baltimore Education Research Consortium (BERC) study co-authored with Liepa Boberiene, Ph.D. Front-end HTML/JavaScript coding, interactive visualization engineering, and data pipeline implementation assisted by Google Gemini (Google, 2026).