Texas kids are getting state-mandated education standardized testing This week measures their proficiency in reading, writing, science, and social studies. But these tests are no longer necessarily graded by human teachers. In fact, the Texas Education Agency will deploy a new “automated scoring engine” for open-ended questions on the exam. The state hopes to save millions of dollars through the new program.
The technology, which the Texas Education Agency calls the Automated Scoring Engine (ASE), uses natural language processing to grade student essays, according to the Texas Education Agency. texas tribune. After initial scoring based on the AI model, approximately 25% of test answers will be sent back to human raters for review. san antonio report news media.
The state of Texas expects to save $15-20 million using the new AI tool, primarily because it will need to hire fewer human graders through third-party contracting agencies. The Texas Tribune reported that about 6,000 graders were previously needed, but that has now been reduced to about 2,000.
Speech given by the Texas Education Agency website It appears that tests of the new system showed that humans and automated systems gave most children comparable scores. But many questions remain about how exactly the technology works and which company might help the state develop the software. The Texas Education Agency’s website names Cambium and Pearson as contractors, but the agency did not respond to questions emailed Tuesday.
The Texas Assessment of Academic Readiness (STAAR) was first introduced in 2011 but was redesigned in 2023 to include more open-ended essay-style questions. Previously, the test included more multiple-choice questions that were, of course, also scored by computer tools. The biggest difference is that scoring bubble sheets is different from scoring written responses, and this is harder for computers to understand.
The Texas Education Agency was apparently quick to dismiss the idea of using generative AI chatbots like Chat GPTAccording to the Texas Tribune. A comparison of a PowerPoint presentation on the Texas Education Agency’s website with software like ChatGPT seems to confirm this unease.
“This technology differs from artificial intelligence, which is a computer that adapts using progressive learning algorithms, allowing data to be programmed and essentially teach itself,” the presentation explains. “Instead, the automated scoring engine is a closed database and student response data is only accessible by TEA and its assessment contractors, Cambium and Pearson, and has strict contractual privacy controls.”
According to the agency, any family dissatisfied with their child’s results can request a human review of the test results. san antonio report. But it will cost you $50.