Analysis / Achievement context

Achievement in context

Student needs, not neighborhood income, explain school score differences

MCPS schools in wealthier attendance areas post higher MCAP scores. Four measures of enrolled students — economic disadvantage, multilingual learners, students with disabilities, and chronic absenteeism — explain 90% of the school-to-school variation in ELA proficiency. Adding neighborhood income to those measures raises that figure by less than a tenth of a percentage point.

Neighborhood income and test scores

Each dot is one of the 172 MCPS schools with complete data. Schools in higher-income attendance areas score higher on MCAP ELA (r=+0.66). The fitted line rises about 39 points of proficiency from the lowest- to the highest-income areas.

0 25 50 75 100 0 20 40 60 80 100 ELA proficient % Neighborhood income percentile

What high- and low-income areas differ on

Attendance-area income moves together with the student measures MCPS reports separately. In the wealthiest fifth of attendance areas, economic disadvantage, multilingual learner share, and chronic absenteeism are a fraction of their levels in the least wealthy fifth. A raw income comparison mixes these differences together.

Students with disabilities are the exception: across the all-schools quintiles, SWD share ranges from 13.9% to 17.8% and does not follow the income gradient.

Income quintile Median income ELA Math Econ. disadv. ML SWD Chronic absent
Q1 $99k 36.9% 23.3% 55.4% 37.3% 14.8% 24.9%
Q2 $113k 48.6% 30.7% 44.0% 25.5% 14.9% 23.5%
Q3 $129k 53.0% 32.3% 40.7% 23.4% 16.5% 22.7%
Q4 $163k 62.8% 45.5% 27.6% 16.1% 17.8% 16.9%
Q5 $207k 77.5% 59.8% 16.0% 11.1% 13.9% 12.9%

Quintiles are computed within this view's own income distribution: 195 schools, about 39 per quintile. High-school math figures reflect MCAP Algebra 1, taken only by students who reach the course in high school; most students test in middle school, so the high-school figures describe a smaller, non-representative group.

How much each factor explains

Five models were fit to the same 172 schools, each predicting ELA proficiency from a different set of measures. The four student measures account for nearly all of the variation a model can capture here; adding neighborhood income to them raises explained variation by less than a tenth of a percentage point.

School level only 3% Elementary, middle, and high school baseline differences.
Neighborhood income + level 53% Attendance-area income, before enrolled-student context measures are added.
Economic disadvantage + level 85% Enrolled-student economic disadvantage, before other context measures are added.
Enrolled-student context + level 90% Economic disadvantage, ML, SWD, and chronic absenteeism, without neighborhood income.
Context + neighborhood income + level 90% Full context check: does neighborhood income add signal after enrolled-student context is included?

Math shows the same pattern: income and school level explain 66% of the variation, the four student measures explain 90%, and adding income brings the figure to 90% (163 schools).

Each measure's estimated effect

Each row shows how many points of ELA proficiency separate two schools that are one standard deviation apart on that measure but match on the others; the whisker is a 95% confidence interval. Measured with nothing held constant (top row), income is associated with +12.2 points. Once the four student measures are accounted for, the income estimate falls to +0.3 points, with an interval that spans zero. The student measures retain large effects under the same comparison.

-15 -10 -5 0 +5 +10 +15 Income, measured alone +12.2 Income, in the full model +0.3 Economically disadvantaged -9.6 Multilingual learners -5.5 Students with disabilities -2.5 Chronic absenteeism -2.4 ELA points per 1 SD of each measure

Negative values mean schools with more of that measure score lower (172 schools). Grey means the interval includes zero. Economic disadvantage accounts for most of the change in the income estimate: without it, the other three measures leave income at +3.4 points [+1.9, +5.0]. Interval construction is described in the technical details.

Accounting for the income gap in scores

Schools in the top income quintile (median $190k) average 69.5% proficient in ELA; schools in the bottom quintile (median $99k) average 36.6%. The student-measure model accounts for essentially all of that 32.9-point gap, most of it through economic disadvantage and multilingual learner share.

Top-vs-bottom quintile ELA gap +32.9 pts
Economically disadvantaged +21.0 pts 55.1% in bottom quintile vs 19.9% in top
Multilingual learners +10.0 pts 37.6% in bottom quintile vs 12.6% in top
Students with disabilities -1.2 pts 14.6% in bottom quintile vs 17.4% in top
Chronic absenteeism +3.4 pts 25.2% in bottom quintile vs 13.7% in top
School-level mix -0.2 pts
Unexplained -0.2 pts

Each component is the no-income model coefficient times how far apart the top and bottom income quintiles sit on that measure. Accounting, not causation: it shows what the income gap is made of, not what changing any one measure would do.

Scores against expectations, across incomes

Each school is plotted by how far its ELA score lands above or below what the four student measures predict, against the same income axis as the first chart. The fitted line rises +0.4 points across the full income range (r=+0.02) — the remaining differences between schools show little relationship to income.

0 25 50 75 100 -20 -10 0 +10 +20 ELA points above/below student-context expectation Neighborhood income percentile

Same 172 schools and income axis as the first chart. The dark horizontal line marks zero — a school scoring exactly what its context predicts; the teal line is the fitted trend. Expectations come from the student-measure model (R²=90%), which does not use income.

Rise across income range +0.4 pts versus +39 points in the first chart; r=+0.02
Typical deviation ±5.3 pts how far schools typically land from the context expectation
Added-variable slope +0.3 pts/SD income after residualizing on the student measures; r=+0.04. See technical details.

A school-by-school check

As a check on the model results, each school is compared with the same-level schools most similar to it across the four student measures, and its ELA score is set against the average of those peers. If wealth mattered beyond student context, schools scoring above their peers would cluster in wealthier areas. Peer-matching details are in the technical notes.

Default view Elementary has enough schools (118) for a steadier 10-peer comparison.

Actual ELA vs the peer expectation

For each school, expected ELA is the average actual ELA of the 10 same-level schools most similar to it across the four student measures. The diagonal means actual ELA equals that peer expectation; dots outside the typical band land notably above or below their peers.

0 0 20 20 40 40 60 60 80 80 100 100 Actual ELA proficient % Expected from similar-context peers %

Each level uses one fixed set of schools with neighborhood income, all four student-context measures, and ELA present.

Peer differences by neighborhood income

If wealth explained the remaining differences, this fitted line would rise with income the way the raw pattern does (about 39 points across the range). Here it rises +1.9 points. Schools in similar-income neighborhoods can land 13 points apart against their peers.

0 25 50 75 100 -25 -13 0 +13 +25 ELA points above/below similar-context peers Neighborhood income percentile

118 elementary schools, each compared with the 10 same-level schools most similar to it across the four student measures. The dark horizontal line marks zero — a school scoring exactly what its peers score; the teal line is the fitted trend, and the shaded stripe is the typical ±7-point band.

Rise across income range +1.9 pts fit-line rise from lowest- to highest-income areas; r=+0.09
Typical gap from peers +/-7 pts about two in three schools land this close to their similar-context peers
Peer-group match ~10 pts median peer-group span in economic disadvantage, the most heavily weighted measure

School explorer

Search, filter by level, or sort any column. “Expected” is the similar-context peer expectation used in the charts above; positive differences mean a school scores above the schools most like it across the four student measures.

172 schools

A. Mario Loiederman MiddleMS39.6%39.1%+0.545.0%33.6%12.9%20.4%$124k
Albert Einstein HighHS68.4%63.5%+4.937.6%21.8%15.2%41.4%$124k
Arcola ElementaryES29.3%32.9%-3.658.5%49.4%12.1%30.8%$98k
Argyle MiddleMS42.3%44.9%-2.648.5%28.9%10.3%19.8%$89k
Ashburton ElementaryES64.3%67.8%-3.519.9%15.1%10.8%17.1%$117k
Bayard Rustin ElementaryES57.3%58.2%-0.932.0%28.3%15.6%18.1%$126k
Beall ElementaryES66.8%65.6%+1.225.1%14.9%12.8%12.4%$111k
Bells Mill ElementaryES83.0%78.7%+4.314.2%8.9%10.2%8.9%$226k
Belmont ElementaryES73.7%79.1%-5.415.5%9.6%14.0%8.3%$201k
Benjamin Banneker MiddleMS34.7%47.9%-13.252.3%15.5%16.0%19.8%$112k
Bethesda ElementaryES67.4%67.1%+0.324.0%24.5%11.1%16.6%$140k
Bethesda-Chevy Chase HighHS78.5%76.6%+1.921.6%11.2%8.4%24.8%$171k
Beverly Farms ElementaryES85.8%79.7%+6.18.2%9.3%11.4%11.6%$226k
Briggs Chaney MiddleMS42.9%46.3%-3.451.5%18.2%14.0%11.8%$118k
Brooke Grove ElementaryES53.0%56.8%-3.834.1%13.6%31.9%16.8%$157k
Brookhaven ElementaryES40.7%33.9%+6.858.4%37.9%24.8%20.5%$127k
Brown Station ElementaryES34.4%32.1%+2.356.8%47.6%20.0%27.8%$96k
Burning Tree ElementaryES71.4%78.7%-7.38.3%12.2%18.6%9.7%$253k
Burnt Mills ElementaryES34.6%36.7%-2.162.3%33.1%12.0%27.8%$105k
Burtonsville ElementaryES43.5%52.4%-8.949.4%15.9%14.0%20.2%$137k
Cabin Branch ElementaryES68.9%66.5%+2.424.6%12.7%14.4%14.7%$190k
Candlewood ElementaryES62.0%61.4%+0.627.5%15.7%19.9%15.0%$141k
Cannon Road ElementaryES35.8%39.8%-4.056.5%24.6%26.4%21.1%$96k
Captain James E. Daly ElementaryES37.4%32.7%+4.759.8%43.1%12.3%31.7%$121k
Cashell ElementaryES64.0%61.1%+2.927.3%11.6%26.8%12.4%$196k
Cedar Grove ElementaryES73.7%62.3%+11.425.3%9.5%22.6%20.8%$160k
Clarksburg ElementaryES60.0%64.2%-4.228.1%17.8%13.4%10.2%$163k
Clarksburg HighHS63.6%67.7%-4.131.6%10.4%11.3%30.0%$139k
Clearspring ElementaryES64.2%55.8%+8.438.1%13.0%17.7%20.5%$140k
Clopper Mill ElementaryES33.3%34.6%-1.460.9%28.3%21.7%21.8%$104k
Cloverly ElementaryES42.9%53.8%-10.934.7%22.9%24.1%17.5%$169k
Col. Zadok Magruder HighHS61.6%62.6%-1.041.3%20.0%13.7%37.9%$147k
College Gardens ElementaryES59.1%54.3%+4.837.0%12.4%18.4%23.6%$111k
Damascus ElementaryES51.6%50.8%+0.840.1%23.1%24.6%26.1%$129k
Damascus HighHS68.1%69.7%-1.628.6%7.5%14.1%34.4%$156k
Darnestown ElementaryES63.5%66.8%-3.316.4%10.0%32.2%17.3%$224k
Diamond ElementaryES73.7%78.3%-4.612.5%18.1%13.9%12.8%$165k
Dr. Charles R. Drew ElementaryES64.6%43.2%+21.451.6%16.4%20.8%19.3%$111k
Dr. Martin Luther King Jr. MiddleMS46.4%42.7%+3.751.3%17.3%13.9%25.3%$105k
Dr. Sally K. Ride ElementaryES33.2%40.0%-6.855.1%23.9%27.7%19.2%$150k
Dufief ElementaryES57.8%64.7%-6.917.3%14.4%34.7%12.8%$194k
Earle B. Wood MiddleMS54.1%53.3%+0.837.0%20.4%14.4%21.5%$113k
East Silver Spring ElementaryES44.1%40.5%+3.653.4%22.7%16.7%17.4%$115k
Eastern MiddleMS53.9%48.3%+5.644.0%23.9%14.1%12.0%$106k
Fairland ElementaryES34.0%35.3%-1.360.1%19.5%20.7%29.1%$140k
Fallsmead ElementaryES73.9%74.3%-0.418.1%9.1%11.2%9.8%$154k
Farmland ElementaryES67.2%67.3%-0.118.8%29.7%13.7%15.9%$111k
Fields Road ElementaryES49.5%46.0%+3.543.7%26.8%14.5%29.8%$108k
Flora M. Singer ElementaryES50.2%55.8%-5.630.8%28.4%26.4%10.6%$152k
Flower Hill ElementaryES31.8%37.4%-5.655.5%37.0%15.3%24.5%$105k
Flower Valley ElementaryES52.5%67.9%-15.423.2%19.2%15.7%16.0%$94k
Forest Knolls ElementaryES53.9%55.1%-1.235.9%20.3%23.5%13.6%$166k
Forest Oak MiddleMS32.9%32.6%+0.356.9%30.9%14.8%24.7%$108k
Fox Chapel ElementaryES52.0%38.8%+13.250.5%33.0%9.4%22.4%$100k
Francis Scott Key MiddleMS35.1%33.8%+1.361.8%35.3%9.6%16.1%$103k
Gaithersburg ElementaryES17.1%27.1%-10.058.4%64.7%14.6%29.7%$89k
Gaithersburg HighHS42.8%44.5%-1.747.8%32.4%15.9%51.0%$113k
Gaithersburg MiddleMS34.1%36.8%-2.747.9%29.9%16.3%31.0%$119k
Galway ElementaryES39.3%34.9%+4.458.9%34.0%16.0%20.5%$109k
Garrett Park ElementaryES62.7%67.1%-4.421.5%24.3%8.5%16.8%$129k
Georgian Forest ElementaryES26.9%33.5%-6.657.8%47.1%9.4%31.8%$82k
Germantown ElementaryES46.0%54.0%-8.044.9%15.8%25.9%22.9%$124k
Glen Haven ElementaryES32.6%38.0%-5.448.2%35.0%21.5%24.8%$117k
Glenallan ElementaryES37.3%43.1%-5.947.3%30.8%21.1%23.7%$137k
Goshen ElementaryES49.6%48.5%+1.145.5%26.9%13.8%24.3%$124k
Great Seneca Creek ElementaryES44.6%44.2%+0.447.1%25.2%12.9%25.8%$120k
Greencastle ElementaryES32.4%35.6%-3.263.1%21.3%20.0%28.4%$82k
Greenwood ElementaryES75.7%80.3%-4.610.3%5.3%12.5%10.9%$199k
Harmony Hills ElementaryES21.8%25.6%-3.863.3%66.6%9.9%21.7%$89k
Highland ElementaryES41.8%29.9%+11.956.5%50.6%12.6%19.9%$99k
Highland View ElementaryES54.5%49.5%+5.044.8%29.8%13.9%14.7%$126k
Jackson Road ElementaryES31.9%35.6%-3.758.0%34.4%16.5%20.7%$108k
James Hubert Blake HighHS55.1%55.4%-0.349.7%19.7%10.5%38.7%$112k
JoAnn Leleck Elementary at Broad AcresES27.9%25.0%+2.968.6%69.3%9.8%20.9%$64k
John F. Kennedy HighHS36.9%46.4%-9.554.7%37.2%14.4%48.9%$99k
John T. Baker MiddleMS58.5%61.2%-2.731.2%11.5%19.1%22.0%$153k
Jones Lane ElementaryES64.7%63.1%+1.625.3%22.3%14.8%19.0%$195k
Judith A. Resnik ElementaryES39.8%37.1%+2.756.1%24.8%11.9%29.5%$140k
Julius West MiddleMS60.3%57.7%+2.631.6%16.8%13.4%19.2%$126k
Kemp Mill ElementaryES25.4%24.0%+1.469.6%60.7%7.3%24.9%$148k
Kensington Parkwood ElementaryES78.9%78.0%+0.910.5%10.9%14.8%10.4%$161k
Kingsview MiddleMS62.6%67.2%-4.630.6%10.4%9.4%13.7%$147k
Lake Seneca ElementaryES24.8%36.3%-11.561.7%22.4%26.4%26.3%$90k
Lakelands Park MiddleMS62.4%66.1%-3.726.7%15.6%11.3%16.2%$163k
Lakewood ElementaryES73.6%67.9%+5.719.5%14.7%16.9%15.7%$190k
Laytonsville ElementaryES54.0%61.7%-7.730.1%17.5%17.8%15.0%$168k
Little Bennett ElementaryES57.9%62.3%-4.424.9%14.7%22.6%15.7%$157k
Lois P. Rockwell ElementaryES58.7%62.5%-3.831.8%10.5%29.9%17.2%$167k
Lucy V. Barnsley ElementaryES67.6%56.0%+11.636.6%24.1%14.4%12.7%$133k
Luxmanor ElementaryES59.2%63.4%-4.223.8%21.7%21.7%15.5%$121k
Maryvale ElementaryES63.4%56.3%+7.143.5%17.4%12.6%15.8%$114k
Meadow Hall ElementaryES35.6%37.5%-1.945.7%45.7%21.7%15.7%$158k
Mill Creek Towne ElementaryES55.6%49.5%+6.140.9%24.4%20.1%22.1%$131k
Monocacy ElementaryES64.6%69.3%-4.823.3%11.7%13.5%11.0%$207k
Montgomery Blair HighHS65.0%64.4%+0.637.4%20.4%8.1%33.7%$101k
Montgomery Village MiddleMS23.3%36.2%-12.959.6%39.5%16.4%21.2%$95k
Neelsville MiddleMS39.7%43.2%-3.547.3%26.5%9.8%25.0%$119k
Newport Mill MiddleMS50.9%48.0%+2.944.5%27.0%21.3%14.0%$129k
North Bethesda MiddleMS78.2%74.1%+4.114.2%7.1%11.5%11.4%$157k
Northwest HighHS71.6%65.0%+6.630.3%10.3%9.7%29.7%$152k
Northwood HighHS52.9%51.5%+1.446.3%29.4%9.3%46.4%$125k
Oakland Terrace ElementaryES59.6%63.5%-3.925.6%14.2%22.6%14.7%$171k
Odessa Shannon MiddleMS29.5%34.4%-4.953.3%35.4%15.8%24.1%$120k
Olney ElementaryES59.7%70.8%-11.118.7%10.0%13.3%15.7%$170k
Paint Branch HighHS54.6%56.8%-2.251.0%12.9%11.7%25.9%$109k
Parkland MiddleMS50.0%45.1%+4.946.2%28.6%8.8%18.1%$105k
Poolesville ElementaryES69.7%72.0%-2.314.2%8.0%12.5%18.2%$186k
Potomac ElementaryES79.8%80.3%-0.57.1%6.7%6.0%11.0%$237k
Quince Orchard HighHS63.3%66.7%-3.431.3%17.2%10.2%35.9%$131k
Rachel Carson ElementaryES75.9%65.8%+10.123.1%17.4%14.4%15.3%$175k
Redland MiddleMS49.0%51.2%-2.244.0%20.8%18.0%22.2%$168k
Richard Montgomery HighHS73.8%71.2%+2.625.5%13.5%8.6%26.6%$126k
Ridgeview MiddleMS52.1%58.6%-6.532.6%17.0%16.7%20.8%$123k
Ritchie Park ElementaryES77.6%78.8%-1.212.6%7.3%11.1%11.4%$182k
Robert Frost MiddleMS82.2%73.3%+8.915.6%6.0%11.4%9.6%$176k
Roberto W Clemente MiddleMS47.4%46.1%+1.344.9%16.4%18.0%17.9%$107k
Rock Creek Forest ElementaryES66.7%63.6%+3.127.0%20.1%16.1%12.0%$101k
Rock Creek Valley ElementaryES49.0%56.2%-7.227.4%23.2%29.4%15.9%$178k
Rock View ElementaryES36.7%43.6%-6.942.2%37.3%30.1%21.3%$129k
Rockville HighHS63.8%63.1%+0.740.0%19.4%18.2%38.7%$113k
Rocky Hill MiddleMS49.9%54.2%-4.338.7%16.6%12.5%21.4%$127k
Rolling Terrace ElementaryES13.8%26.4%-12.666.2%63.4%11.0%21.1%$63k
Ronald McNair ElementaryES58.6%62.4%-3.830.9%13.4%11.5%17.0%$155k
Rosemont ElementaryES46.1%31.8%+14.360.3%38.6%15.3%29.4%$104k
S. Christa McAuliffe ElementaryES37.7%37.6%+0.156.3%26.3%14.5%22.5%$103k
Sargent Shriver ElementaryES26.6%26.8%-0.256.8%58.4%11.5%23.1%$113k
Seneca Valley HighHS59.3%57.1%+2.245.9%15.3%15.0%30.7%$115k
Sequoyah ElementaryES48.0%45.9%+2.139.5%28.9%24.7%21.1%$183k
Shady Grove MiddleMS50.0%50.4%-0.439.5%22.1%10.7%21.2%$127k
Sherwood ElementaryES57.0%65.5%-8.524.7%9.3%22.5%21.2%$169k
Silver Creek MiddleMS69.5%65.8%+3.727.5%13.2%14.2%11.7%$206k
Silver Spring International MiddleMS54.3%50.8%+3.537.5%23.8%10.1%24.3%$110k
Sligo Creek ElementaryES64.2%69.4%-5.221.9%11.1%11.1%17.6%$111k
Sligo MiddleMS50.1%51.6%-1.538.3%24.9%16.0%12.0%$120k
Snowden Farm ElementaryES69.4%69.1%+0.320.0%11.6%11.2%14.0%$164k
Somerset ElementaryES78.7%66.3%+12.419.9%15.9%10.7%20.6%$153k
South Lake ElementaryES21.0%25.5%-4.567.0%60.3%11.4%35.3%$68k
Spark M. Matsunaga ElementaryES71.9%64.6%+7.325.0%14.5%15.8%17.9%$206k
Springbrook HighHS56.0%54.5%+1.547.8%27.6%9.3%37.2%$118k
Stedwick ElementaryES32.3%36.4%-4.160.6%33.4%19.2%25.3%$131k
Stone Mill ElementaryES73.6%75.9%-2.314.4%14.4%16.4%12.2%$170k
Stonegate ElementaryES54.7%58.8%-4.130.1%15.1%22.2%15.0%$172k
Strawberry Knoll ElementaryES47.6%37.9%+9.755.1%21.2%26.7%22.0%$124k
Summit Hall ElementaryES20.7%35.2%-14.569.5%41.5%19.0%25.9%$105k
Takoma Park MiddleMS70.0%65.7%+4.329.6%14.3%12.4%13.0%$115k
Thurgood Marshall ElementaryES44.7%54.1%-9.437.4%18.2%21.8%22.3%$109k
Tilden MiddleMS68.3%70.4%-2.121.6%17.3%13.5%16.5%$122k
Travilah ElementaryES82.2%78.7%+3.514.5%8.9%9.7%7.8%$194k
Tubman ElementaryES28.2%34.3%-6.162.3%46.6%19.1%24.1%$102k
Twinbrook ElementaryES28.5%35.0%-6.553.8%43.2%14.0%25.0%$129k
Viers Mill ElementaryES28.4%36.7%-8.349.2%37.7%27.8%23.0%$120k
Walter Johnson HighHS84.4%73.5%+10.918.0%8.0%11.9%22.4%$135k
Washington Grove ElementaryES35.6%36.2%-0.654.6%32.4%17.5%22.1%$123k
Waters Landing ElementaryES36.0%35.2%+0.856.0%29.2%18.1%25.6%$103k
Watkins Mill ElementaryES24.3%27.4%-3.161.2%58.8%14.0%34.3%$86k
Watkins Mill HighHS43.6%44.2%-0.651.3%32.8%14.0%45.1%$94k
Wayside ElementaryES87.3%78.1%+9.29.5%8.9%16.0%10.2%$222k
Weller Road ElementaryES27.3%32.6%-5.360.6%49.1%15.6%20.0%$139k
Westbrook ElementaryES78.0%79.0%-1.013.6%14.0%10.1%8.9%$175k
Westland MiddleMS76.6%73.0%+3.615.4%10.5%13.3%13.8%$156k
Westover ElementaryES68.7%61.4%+7.428.4%10.4%24.1%13.9%$124k
Wheaton HighHS55.6%56.8%-1.244.5%26.2%11.0%35.0%$120k
Wheaton Woods ElementaryES31.1%27.5%+3.661.8%53.8%18.9%21.8%$107k
Whetstone ElementaryES34.6%34.0%+0.658.3%41.2%17.7%23.9%$96k
White Oak MiddleMS41.1%38.0%+3.155.8%29.4%9.1%16.0%$110k
William B. Gibbs Jr. ElementaryES61.1%53.0%+8.137.6%13.2%23.5%20.2%$149k
William H. Farquhar MiddleMS73.7%66.6%+7.124.4%7.5%13.0%12.3%$163k
William Tyler Page ElementaryES60.1%58.6%+1.537.4%11.4%11.4%17.2%$101k
Wilson Wims ElementaryES81.1%72.3%+8.819.5%8.5%18.3%9.9%$206k
Wood Acres ElementaryES83.0%80.0%+3.06.7%6.9%11.6%6.7%$211k
Woodfield ElementaryES70.2%59.6%+10.629.2%10.0%35.3%16.9%$177k
Woodlin ElementaryES58.6%57.3%+1.331.3%22.8%16.6%19.3%$112k

Source: MSDE MCAP and Special Services; MSDE Attendance; Census ACS attendance-area income · Updated: 2025 / ACS 2024

Technical model details

One-at-a-time relationships

These show each factor's raw relationship with ELA before accounting for overlap.

Economically disadvantaged very strong lower-score pattern r=-0.92 · n=192
Multilingual learners very strong lower-score pattern r=-0.83 · n=182
Adults with bachelor's degree or higher very strong higher-score pattern r=+0.81 · n=189
Neighborhood median income strong higher-score pattern r=+0.75 · n=189
Neighborhood poverty strong lower-score pattern r=-0.62 · n=189
Chronic absenteeism moderate lower-score pattern r=-0.49 · n=195
Students with disabilities weaker lower-score pattern r=-0.27 · n=196

Model comparison

Rows use the same 172 schools. Larger values mean the prediction fits more of the school-to-school ELA variation.

School level only 3% Elementary, middle, and high school baseline differences.
Neighborhood income + level 53% Attendance-area income, before enrolled-student context measures are added.
Economic disadvantage + level 85% Enrolled-student economic disadvantage, before other context measures are added.
Enrolled-student context + level 90% Economic disadvantage, ML, SWD, and chronic absenteeism, without neighborhood income.
Context + neighborhood income + level 90% Full context check: does neighborhood income add signal after enrolled-student context is included?

The income estimate as controls are added one at a time

Every dot below is the income coefficient; the row labels say which enrolled-student measures the model also includes, added cumulatively.

-5 0 +5 +10 +15 Income alone +12.2 + economic disadvantage +0.7 + multilingual learners -0.2 + students with disabilities +0.6 + chronic absenteeism +0.3 ELA points per 1 SD of neighborhood income

ELA proficiency points per one standard deviation of log neighborhood income; whiskers are 95% school-resample bootstrap interval (2,000 draws). Because the five measures are correlated, how fast the estimate collapses depends on the order controls are added — this ordering (largest first) is one choice, not a decomposition. The order-independent facts: measured alone, income is worth +12.2 points; with all four student measures it is +0.3 [-1.1, +1.7]; and without economic disadvantage the other three leave it at +3.4 [+1.9, +5.0] — so economic disadvantage specifically, not the sheer number of controls, is what absorbs it.

Conditional coefficients

Neighborhood median income +0.3 [-1.1, +1.7]
Economically disadvantaged -9.6 [-11.3, -7.7]
Multilingual learners -5.5 [-7.2, -3.9]
Students with disabilities -2.5 [-3.5, -1.4]
Chronic absenteeism -2.4 [-4.0, -0.9]

Continuous coefficients are ELA proficiency-point changes for a one-standard-deviation increase, comparing schools that match on the other listed measures; brackets are 95% school-resample bootstrap interval (2,000 draws). They should not be read as separate causal effects — neighborhood income and enrolled economic disadvantage correlate at r=-0.78, so the matched comparison is statistical, not a claim about mechanism.

Level terms are relative to elementary schools: Middle school -3.0 pts; High school +9.7 pts.

Formal added-variable check: ELA and log income each residualized on the four student measures and school level. Slope +0.31 points per SD, r=+0.04 — the formal version of the scores-against-expectations chart. Confidence intervals throughout are 95% school-resample bootstrap interval (2,000 draws); intervals widen where measures overlap and the data cannot fully separate them.

Peer-matching notes

Average ELA proficiency among the nearest same-level schools by four-measure student context (economic disadvantage, multilingual learners, students with disabilities, and chronic absenteeism — each standardized within level and weighted by how strongly it predicts ELA in the student-context model), excluding the school itself.

Distance weights: economic disadvantage 47%, multilingual learners 27%, chronic absenteeism 14%, students with disabilities 12%.

  • This is a descriptive like-to-like comparison, not a causal estimate or school-quality score.
  • Middle and high school peer charts are exploratory because far fewer schools report all four measures.
  • Peer groups match on the four reported measures; differences they do not capture (programs, mobility, staffing) can still matter.
  • Peer similarity weights each measure by its modeled importance; an unweighted match loosens the economic-disadvantage match and leaves disadvantage-related structure in the differences.
  • Distance weights are estimated once across all levels, so they mostly reflect elementary schools. Per-level fits hint that economic disadvantage and absenteeism matter relatively more at secondary levels, but 34 middle and 20 high schools are too few to estimate level-specific weights reliably — one more reason the secondary panels are checks, not findings.

Limitations

  • This is a cross-school context model, not a causal estimate or a school-quality score.
  • Neighborhood income describes attendance areas, not necessarily enrolled students.
  • Neighborhood income and enrolled economic disadvantage are strongly related; coefficients should not be read as separate causal effects.
  • Residuals are sensitive to model choices and omitted factors; they should be read as context flags, not rankings.
  • Suppressed subgroup values are excluded where needed rather than treated as zero.
  • The gap decomposition is accounting, not causation: it shows what the income gap is made of, not what changing any one measure would do.