How Kids Are Performing

Notes

1. The data supporting this display came from Star Early Literacy, Star Reading, and Star Math adaptive assessments taken in English in the United States. In each school year, the sample includes about 3 million records for Star Early Literacy, 18 million records for Star Reading, and 15 million records for Star Math. For more information about Star Assessments, start here.

2. Performance results for spring 2020 are not available due to the pandemic, when very few tests were completed. Likewise, SGP results for fall to spring of the 2019–2020 school year are not reported for the same reason. For more on the impacts of the pandemic on learning, refer to our series of How Kids Are Performing results in the US and UK.

3. Some degree of caution should be exercised when interpreting these data or any similar data set. Of particular note:

   • This is a convenience sample of Star Assessments users and has been drawn from all 50 states and DC, but it has not been weighted to reflect characteristics of US schools as a whole. Also, while the cohort of schools using Star is both large and generally consistent year over year, there is some degree of change. Therefore, a small portion of any increases or decreases in performance and growth could be a reflection of the changing user base.
   • Data from 2020–2021 is comprised of a combination of in-school and remote testing results. In previous How Kids Are Performing reports, we examined results for 2020–2021 as a function of whether tests were completed in school buildings or remotely. In most grades, the results were comparable, but students in early grades who tested remotely performed substantially higher than in-school testers, raising questions about the validity of those scores. That said, we’ve elected to retain the remote records in this summary, because in 2020–2021, remote learning and testing were a reality for approximately one-third of our sample, and excluding those records would restrict our ability to present a comprehensive picture of achievement and growth during the pandemic.
   • Sample sizes and test-taking practices are not uniform across grades. In particular, Star is used less often in grades 9–12 than earlier grades, so that student population may be less representative of the overall population. Additionally, note that not all Kindergarten and first grade students take the same tests. Independent readers take Star Reading, while pre-readers take Star Early Literacy and move to Star Reading only when they are ready.

4. The data set was cleaned to remove tests that were interrupted and resumed, or that had other implementation issues such as extreme duration values. Related, if a student took more than one test per season, we used the first test in the fall (August through November), the test closest to January 15 in the winter (December through March), and the last test in spring. To be included in the sample, a student did not have to test in every seasonal window or in every school year.

5. Student Growth Percentiles are a norm-referenced score, and the norms are updated periodically with more recent data. The data included in this display came from three different norming samples, and changes in the norms could account for some of the year-to-year differences in median SGPs. In 2017–2018 and 2018–2019, SGPs were based on growth data from the 2016–2017 school year and two prior school years. SGPs for 2019–2020 are not reported in this display, as noted above. In 2020–2021 and 2021–2022, SGPs were based on data from 2017–2018 school year and two prior school years. In 2022–2023, SGPs are based on data from the 2018–2019 school year and two prior school years. Learn more about SGPs.