Oregon School Assessment

Context + education + attendance joint model (scored-student-weighted)

Published analysis report from Evidence Lab artifacts.

Context + education + attendance joint model (scored-student-weighted)
Metric: Percent Proficient
Attendance field: Attendance_Percent_Regular_Attenders
Regenerated: 2026-08-06

Context specs included:
- Income: ACS_Median_Household_Income
- Poverty: Students Experiencing Poverty

Note on betas: standardized regression coefficients from 'Percent Proficient ~ context + education + attendance'. Variables are z-scored using scored-student weights. A beta indicates the change (in SD units) in the metric for a 1-SD increase in the predictor, holding the others constant.

Filters: Student Group = Total Population; All Grades dropped if grade-specific rows exist.
Percent Proficient is reconstructed as Number Proficient / Scored Performance Denominator.
Censored AAG context values use documented interval midpoints; raw tokens and bounds remain in the processed data.

Context spec: Income (ACS_Median_Household_Income)

Dataset: SchoolDataWithAddressesAndCensusSES.csv
  Rows used: 3,365 (from 3,505)
  Students with reported Level 1-4 results: 253,682
  Weighted means (student-centered): Income=$86,969, education=34.3%, attendance=69.1%, metric=42.9%
  Missing Income rows: 30
  Missing education rows: 0
  Missing attendance rows: 109
  Missing metric rows: 0

  Weighted correlations
    r(Income, metric): 0.449
    r(education, metric): 0.576
    r(attendance, metric): 0.461
    r(Income, education): 0.690
    r(Income, attendance): 0.351
    r(education, attendance): 0.325

  Standardized regression (metric ~ Income + education + attendance)
    Beta Income: 0.025
    Beta education: 0.460
    Beta attendance: 0.303

  Partial correlations (controlling for the other two)
    Income | Education, Attendance: 0.023
    Education | Income, Attendance: 0.397
    Attendance | Income, Education: 0.345

  Weighted R^2 (model variants)
    Income only: 0.201
    Education only: 0.332
    Attendance only: 0.212
    Income + Education: 0.337
    Income + Attendance: 0.306
    Education + Attendance: 0.415
    All three: 0.416

  Unique variance (approx.)
    Unique Income: 0.000
    Unique education: 0.109
    Unique attendance: 0.079
    Shared (overlap): 0.227

  Added value of attendance beyond Income+education: 0.079
  Added value of education beyond Income+attendance: 0.109
  Added value of Income beyond education+attendance: 0.000

Dataset: SchoolDataMathWithAddressesAndCensusSES.csv
  Rows used: 3,278 (from 3,419)
  Students with reported Level 1-4 results: 247,243
  Weighted means (student-centered): Income=$87,333, education=34.5%, attendance=69.4%, metric=31.9%
  Missing Income rows: 28
  Missing education rows: 0
  Missing attendance rows: 107
  Missing metric rows: 0

  Weighted correlations
    r(Income, metric): 0.485
    r(education, metric): 0.597
    r(attendance, metric): 0.570
    r(Income, education): 0.689
    r(Income, attendance): 0.346
    r(education, attendance): 0.322

  Standardized regression (metric ~ Income + education + attendance)
    Beta Income: 0.043
    Beta education: 0.433
    Beta attendance: 0.416

  Partial correlations (controlling for the other two)
    Income | Education, Attendance: 0.044
    Education | Income, Attendance: 0.409
    Attendance | Income, Education: 0.487

  Weighted R^2 (model variants)
    Income only: 0.235
    Education only: 0.356
    Attendance only: 0.325
    Income + Education: 0.367
    Income + Attendance: 0.419
    Education + Attendance: 0.516
    All three: 0.517

  Unique variance (approx.)
    Unique Income: 0.001
    Unique education: 0.097
    Unique attendance: 0.150
    Shared (overlap): 0.268

  Added value of attendance beyond Income+education: 0.150
  Added value of education beyond Income+attendance: 0.097
  Added value of Income beyond education+attendance: 0.001

Dataset: SchoolDataScienceWithAddressesAndCensusSES.csv
  Rows used: 1,276 (from 1,330)
  Students with reported Level 1-4 results: 107,068
  Weighted means (student-centered): Income=$86,086, education=34.3%, attendance=65.4%, metric=30.1%
  Missing Income rows: 14
  Missing education rows: 0
  Missing attendance rows: 40
  Missing metric rows: 0

  Weighted correlations
    r(Income, metric): 0.407
    r(education, metric): 0.539
    r(attendance, metric): 0.395
    r(Income, education): 0.683
    r(Income, attendance): 0.348
    r(education, attendance): 0.301

  Standardized regression (metric ~ Income + education + attendance)
    Beta Income: 0.004
    Beta education: 0.459
    Beta attendance: 0.255

  Partial correlations (controlling for the other two)
    Income | Education, Attendance: 0.004
    Education | Income, Attendance: 0.382
    Attendance | Income, Education: 0.284

  Weighted R^2 (model variants)
    Income only: 0.165
    Education only: 0.290
    Attendance only: 0.156
    Income + Education: 0.293
    Income + Attendance: 0.239
    Education + Attendance: 0.350
    All three: 0.350

  Unique variance (approx.)
    Unique Income: 0.000
    Unique education: 0.111
    Unique attendance: 0.057
    Shared (overlap): 0.182

  Added value of attendance beyond Income+education: 0.057
  Added value of education beyond Income+attendance: 0.111
  Added value of Income beyond education+attendance: 0.000


Context spec: Poverty (Students Experiencing Poverty)

Dataset: SchoolDataWithAddressesAndCensusSES.csv
  Rows used: 3,360 (from 3,505)
  Students with reported Level 1-4 results: 253,602
  Weighted means (student-centered): Poverty=34.2%, education=34.3%, attendance=69.1%, metric=42.9%
  Missing Poverty rows: 13
  Missing education rows: 30
  Missing attendance rows: 101
  Missing metric rows: 0

  Weighted correlations
    r(Poverty, metric): -0.696
    r(education, metric): 0.576
    r(attendance, metric): 0.461
    r(Poverty, education): -0.635
    r(Poverty, attendance): -0.399
    r(education, attendance): 0.324

  Standardized regression (metric ~ Poverty + education + attendance)
    Beta Poverty: -0.488
    Beta education: 0.201
    Beta attendance: 0.201

  Partial correlations (controlling for the other two)
    Poverty | Education, Attendance: -0.476
    Education | Poverty, Attendance: 0.223
    Attendance | Poverty, Education: 0.263

  Weighted R^2 (model variants)
    Poverty only: 0.484
    Education only: 0.332
    Attendance only: 0.212
    Poverty + Education: 0.514
    Poverty + Attendance: 0.524
    Education + Attendance: 0.415
    All three: 0.548

  Unique variance (approx.)
    Unique Poverty: 0.132
    Unique education: 0.024
    Unique attendance: 0.034
    Shared (overlap): 0.358

  Added value of attendance beyond Poverty+education: 0.034
  Added value of education beyond Poverty+attendance: 0.024
  Added value of Poverty beyond education+attendance: 0.132

Dataset: SchoolDataMathWithAddressesAndCensusSES.csv
  Rows used: 3,274 (from 3,419)
  Students with reported Level 1-4 results: 247,170
  Weighted means (student-centered): Poverty=33.9%, education=34.5%, attendance=69.4%, metric=31.9%
  Missing Poverty rows: 11
  Missing education rows: 28
  Missing attendance rows: 100
  Missing metric rows: 0

  Weighted correlations
    r(Poverty, metric): -0.617
    r(education, metric): 0.597
    r(attendance, metric): 0.570
    r(Poverty, education): -0.636
    r(Poverty, attendance): -0.394
    r(education, attendance): 0.322

  Standardized regression (metric ~ Poverty + education + attendance)
    Beta Poverty: -0.284
    Beta education: 0.300
    Beta attendance: 0.362

  Partial correlations (controlling for the other two)
    Poverty | Education, Attendance: -0.304
    Education | Poverty, Attendance: 0.328
    Attendance | Poverty, Education: 0.447

  Weighted R^2 (model variants)
    Poverty only: 0.381
    Education only: 0.356
    Attendance only: 0.325
    Poverty + Education: 0.451
    Poverty + Attendance: 0.508
    Education + Attendance: 0.516
    All three: 0.561

  Unique variance (approx.)
    Unique Poverty: 0.045
    Unique education: 0.053
    Unique attendance: 0.110
    Shared (overlap): 0.353

  Added value of attendance beyond Poverty+education: 0.110
  Added value of education beyond Poverty+attendance: 0.053
  Added value of Poverty beyond education+attendance: 0.045

Dataset: SchoolDataScienceWithAddressesAndCensusSES.csv
  Rows used: 1,273 (from 1,330)
  Students with reported Level 1-4 results: 107,015
  Weighted means (student-centered): Poverty=32.7%, education=34.3%, attendance=65.4%, metric=30.1%
  Missing Poverty rows: 8
  Missing education rows: 14
  Missing attendance rows: 35
  Missing metric rows: 0

  Weighted correlations
    r(Poverty, metric): -0.656
    r(education, metric): 0.539
    r(attendance, metric): 0.394
    r(Poverty, education): -0.630
    r(Poverty, attendance): -0.347
    r(education, attendance): 0.302

  Standardized regression (metric ~ Poverty + education + attendance)
    Beta Poverty: -0.479
    Beta education: 0.185
    Beta attendance: 0.172

  Partial correlations (controlling for the other two)
    Poverty | Education, Attendance: -0.451
    Education | Poverty, Attendance: 0.195
    Attendance | Poverty, Education: 0.217

  Weighted R^2 (model variants)
    Poverty only: 0.430
    Education only: 0.291
    Attendance only: 0.155
    Poverty + Education: 0.456
    Poverty + Attendance: 0.461
    Education + Attendance: 0.350
    All three: 0.482

  Unique variance (approx.)
    Unique Poverty: 0.132
    Unique education: 0.020
    Unique attendance: 0.026
    Shared (overlap): 0.303

  Added value of attendance beyond Poverty+education: 0.026
  Added value of education beyond Poverty+attendance: 0.020
  Added value of Poverty beyond education+attendance: 0.132


Launch dashboard