23-03 Final Report: Racial Equity Effects of Restricting In-Person Education During the COVID-19 Pandemic

May 2023

Report Details

3. Assessment scores declined, and disparities increased

Student assessment scores declined during the 2021-22 school year. Existing racial disparities increased. School poverty level had the greatest association with assessment scores.

In Washington, an annual standardized test, called the Smarter Balanced Assessment, evaluates math and English language arts (ELA) skills for students in third through eighth and tenth grades. Smarter Balanced Assessment scores range from approximately 2,000 to 3,000. The state sets minimum passing scores for each grade level. These are called the state standards.

Students usually take the assessment each spring. Assessments in Spring 2020 were canceled due to the pandemic, and assessments in Spring 2021 were postponed to Fall 2021. The fall assessment tested students on material learned during the 2020-21 school year. Regular assessment scheduling returned in Spring 2022.

Exhibit 3.1: Students took the assessment in Spring 2019. The pandemic delayed the next assessment until Fall 2021.

Timeline2.svg

Source: JLARC staff analysis.

Math and ELA scores declined for all students after restrictions to in-person instruction

JLARC staff analyzed assessment score data for Spring 2019, Fall 2021, and Spring 2022 (the 2018-19 through 2021-22 school years). See Appendix A for more detail about student assessment scores.

Median assessment scores in math and ELA declined from Spring 2019 to Fall 2021. Scores increased in Spring 2022, but remained lower than Spring 2019 scores. The percentage of students in each grade that met the state standards for math and ELA scores also declined. To describe year-over-year changes, JLARC staff adjusted scores relative to the minimum possible scoreThe minimum score is approximately 2,000, with variation by grade level and subject..

  • Math scores decreased by 3.2% to 10.9%, depending on grade level, from Spring 2019 to Fall 2021. Scores changed by -0.5% to 6.7% in Spring 2022, and remained 3.6% to 8.2%% lower than Spring 2019 scores.
  • The percentage of studentsThese percentages may differ from those in OSPI's State Report Card due to the student records JLARC staff used in this analysis. meeting the state standard in math decreased from 47.7% in Spring 2019 to 32.8% in Fall 2021 and increased to 38.7% in Spring 2022.
  • ELA scores decreased by 2.8% to 4.1%, depending on grade level, from Spring 2019 to Fall 2021. Scores changed by -0.9% to 1.4% in Spring 2022, and remained 1.5% to 4.9% lower than Spring 2019 scores.
  • The percentage of studentsThese percentages may differ from those in OSPI's State Report Card due to the student records JLARC staff used in this analysis. meeting the state standard in ELA decreased from 59.9% in Spring 2019 to 52.7% in Fall 2021 and increased slightly to 52.9% in Spring 2022.
Exhibit 3.2: Median math and ELA scores declined from Spring 2019 to Spring 2022
Assessment%20Scores%20v3.png
Source: JLARC staff analysis of Office of Superintendent of Public Instruction (OSPI) data. JLARC staff estimates may not align with OSPI's due to the student populations analyzed in this report.

Assessment score declines varied by student race and ethnicity

Racial disparities in student assessment scores existed prior to the restrictions to in-person instruction (Exhibit 3.3). For example:

  • Asian students had the highest math and ELA scores.
  • White students and students of two or more races had the next highest scores.
  • American Indian/Alaskan Native and Native Hawaiian/Other Pacific Islander students had the lowest math and ELA scores.

These racial disparities continued after restrictions to in-person instruction. Assessment scores declined for all groups from Spring 2019 to Fall 2021 and began increasing for some in Spring 2022.

The rate of change in assessment scores varied for students of different races and ethnicities (Exhibit 3.3).

  • Asian students had the smallest decreases in Fall 2021.
  • Native Hawaiian/Other Pacific Islander students had the largest decreases in Fall 2021.
  • Math scores increased in Spring 2022 for all groups, but ELA scores continued to decline for students who were American Indian/Alaskan Native, two or more races, or White.

See here for an interactive dashboard of student assessment scores.

Exhibit 3.3: Math and ELA assessment scores declined for students of all races and ethnicities
Source: JLARC staff analysis of OSPI data. Student race and ethnicity categories are federal categories derived from OSPI data. To describe year-over-year changes, JLARC staff adjusted scores relative to the minimum possible score. Unadjusted scores are shown in the graph. Median scores include all grade levels.

Assessment scores varied by other school and student characteristics. School poverty level had the largest association with scores.

JLARC staff used linear modeling to assess how student and school variables affected variation in assessment scores. For a description of the methodology, see Appendix A. Key factors affecting assessment scores include school poverty level, whether a student was eligible for free and reduced-priced lunch (FRL)Free and reduced-priced lunch eligibility is a proxy indicator of household income, and instruction type.

  • School poverty level: School poverty level (measured by the percentage of students eligible for free and reduced-price lunch) had the largest association with assessment scores. Prior to the restrictions to in-person instruction, students at schools with the highest poverty level scored 13% to 15% lower than students at schools with the lowest poverty levels. This association increased during the Fall 2021 and Spring 2022 assessments. Students at high poverty schools scored 17% to 22% lower than students at low poverty schools in Fall 2021.

  • Free and reduced-priced lunch: Whether a student was eligible for free and reduced-priced lunch (FRL) also had a large association with assessment scores. Students eligible for FRL scored 10% to 12% lower than students who were not eligible for FRL. This association remained constant during the Fall 2021 and Spring 2022 assessments.
  • Instruction type: Instruction type was associated with math assessment scores. Students at schools offering 100% in-person instruction during the 2020-21 school year scored 3% higher than students at schools offering only remote or hybrid instruction in the Fall 2021 assessment. The association of instruction type on math scores was outweighed by that of student race and ethnicity, student FRL eligibility, and school poverty level. Instruction type was not associated with ELA scores.