TL;DR
Turn the school run and nap time into listening time
- Thousands of audiobooks, podcasts and originals
- Listen on your phone, tablet or Echo — also offline
- Cancel anytime
Texas schools and districts submitted 64,489 requests this spring to rescore open-ended responses on STAAR reading exams, nearly three times last year’s total. Human reviews led the state to raise more than 13,300 scores; the Texas Education Agency says fewer than 0.5% of all open-ended reading responses changed.
Texas schools and districts requested 64,489 rescoring reviews of open-ended responses on spring STAAR reading exams, nearly three times the 21,620 requests last year, according to figures reported by The 74. The Texas Education Agency said human reviews led to more than 13,300 score increases, renewing questions about the state’s use of automated scoring on written answers.
The state received 2,758 requests in spring 2024, followed by 21,620 the next year and 64,489 this spring. The number of corrections also doubled from the previous year, The 74 reported. Of the scores that were raised after review, 42% had initially been assigned by human scorers and 58% by computers. Humans conducted all rescoring reviews.
More than 3.2 million students took the reading STAAR this spring, including pupils in grades three through eight and students taking the English I and English II end-of-course exams. The agency reported that fewer than 0.5% of open-ended reading responses had their scores changed. More than 250 school districts and public charter networks requested reviews, and nearly 90% had at least one submitted reading exam score increase, according to agency data cited by The 74.
The scale of requests reflects schools seeking a second look, but does not by itself establish why the requests increased or that automated scores were systematically wrong. Texas Education Agency spokesperson Jake Kobersky said the state uses a hybrid scoring model, with human experts involved, and that the automated system is not generative AI. He said grading every response by hand would require four to five times as many human scorers and cost an additional $15 million to $20 million annually, while delaying results.
Why Schools Seek Human Reviews
STAAR results have consequences beyond a student’s test record. Schools use scores to monitor performance, and campus and district accountability ratings are tied to the share of students reaching designated achievement levels. State law allows the education commissioner to intervene in districts with campuses that underperform for multiple years. A revised score can also affect whether a student reaches a performance threshold, potentially changing whether the student must complete remediation or qualifies to graduate.
The jump in requests also puts attention on the reliability and public confidence of the scoring process. Allison Matney, director of the Texas Center for School Accountability, said district leaders may request reviews when they believe a score could affect a school or student. University of Texas at Austin professor Rachel White said the rising volume indicates that educators who read students’ work are increasingly disagreeing with assigned scores, describing it as a trust problem. Those are interpretations of the trend, not proof that the automated system caused every disputed result.
For districts, requesting reviews takes staff time and can carry costs, while an increase in corrected scores may affect local planning and accountability. The agency’s automatic rescoring of extended responses one point short of the next performance level offers some relief: districts can use their requests on students farther from a cutoff. The figures do not show how many requests were submitted for each reason.
As an affiliate, we earn on qualifying purchases.
How Texas Added Automated Scoring
The TEA introduced an automated scoring engine in December 2023 to grade open-ended STAAR responses, including work on reading and English exams. The change followed a redesign of the assessment that reduced the number of multiple-choice questions and added more written, or constructed-response, items. The agency used thousands of responses previously scored by people to train the system to assign scores aligned with human scoring.
The approach drew criticism after the first computer-graded exams saw a rise in students receiving zeros on writing prompts. State officials cited other possible factors in the low scores in 2023, including the number of students retaking the test. The agency has said its scoring engine does not learn from one response to another like generative AI tools such as ChatGPT or Google Gemini. Under the system described by the TEA, humans review a portion of exam scores, and all requested rescoring is done by people.
Rescoring requests have risen each year since 2024. Matney told The 74 that requests were historically rare and attributed the later increase in part to greater awareness among districts about the review process. Districts can request reviews, while the state automatically rescores certain extended responses that are one point from the next performance level: did not meet, approaches, meets or masters.
“That traction of superintendents talking to each other — (and) media continuing to talk about it — made more districts aware that not only can they do this process, but actually they should, and they need to.”
— Allison Matney, director of the Texas Center for School Accountability
As an affiliate, we earn on qualifying purchases.
What the Request Totals Cannot Show
The published totals do not explain how many requests challenged computer-assigned scores versus human-assigned scores, or why each district requested a review. Although the agency raised more than 13,300 scores, the available figures do not state how many reviews produced no change or how often a score was lowered. The reported correction rate also does not establish whether changes were concentrated in particular grades, districts or types of responses.
The count of rescoring requests is not the same as the count of students or individual answers: a request may cover exams or responses, and the source report does not provide a full breakdown of the units behind every total. The overall figure that fewer than 0.5% of open-ended reading responses changed gives scale, but does not resolve concerns about cases where a revised score could affect a student or a school’s accountability status.
The source report also includes a Dallas ISD statement that the district sought reviews because it believes students benefit when human scorers assess their answers, but the supplied material cuts off before giving the district’s total number of requests. It is not clear from the published figures whether the agency plans to release a more detailed breakdown or change its scoring process.
As an affiliate, we earn on qualifying purchases.
District Reviews and State Reporting
The immediate next step is for districts and families to rely on the finalized scores and review outcomes made available by the TEA. Districts may continue to request rescoring under the agency’s process, while the state’s automatic review for responses one point short of the next performance level remains part of the system described in the report. No new scoring policy or deadline for a policy change was announced in the supplied material.
Further reporting from the agency could clarify the number of reviews that resulted in no change, the distribution of corrections by grade and district, and whether the initial score came from a person or the automated engine. Those details would help explain what drove the increase and how the revised results affect students and school accountability. The TEA has defended the hybrid system on grounds of speed and cost; the available source material does not indicate that it plans to replace or expand the process.
As an affiliate, we earn on qualifying purchases.
Key Questions
How many STAAR rescoring requests did Texas receive?
The state received 64,489 requests this spring to rescore open-ended responses on reading exams, compared with 21,620 the previous year and 2,758 in spring 2024, according to figures reported by The 74.
Did the reviews change students’ scores?
Yes. The state raised more than 13,300 reading scores after reviews. The report says fewer than 0.5% of open-ended responses on the reading exams had their scores changed; it does not provide a complete breakdown of reviews that left scores unchanged or lowered them.
Were all initial STAAR responses graded by computers?
No. The TEA uses a hybrid model. Of the scores that were raised after rescoring, 58% had initially been graded by computers and 42% by humans. Human scorers performed all requested rescoring reviews.
Why did rescoring requests increase?
Matney attributed some of the increase to greater awareness among school districts about the review process. The totals do not identify the reason for each request, so they do not establish that all schools were challenging computer-scored answers or that automated scoring alone caused the rise.
Can a revised score affect a student or school?
It can. STAAR scores contribute to school accountability ratings, and reaching a different performance level may affect a student’s next steps. The report does not say how many of the score increases changed an individual student’s placement, graduation status or a school rating.
Source: rss
Cold & flu season Picks
humidifiers
As an affiliate, we earn on qualifying purchases.
