Texas is implementing artificial intelligence (AI) to grade the written portion of the State of Texas Assessments of Academic Readiness (STAAR) exam, aiming to streamline the scoring process and potentially save the state millions of dollars annually. The revised exam, introduced in 2023, emphasizes open-ended questions over multiple-choice ones, resulting in significantly more constructed response items.
The transition to AI scoring is motivated by cost-saving measures, with the Texas Education Agency (TEA) estimating potential savings of up to $20 million per year compared to hiring human scorers from third-party contractors. While machines are not replacing human scorers entirely, the number of temporary scorers needed has drastically reduced, from around 6,000 to fewer than 2,000.
“Artificial Intelligence & AI & Machine Learning” by mikemacmarketing is licensed under CC BY 2.0.
The AI scoring engine utilizes natural language processing to evaluate responses, having been trained on a dataset of 3,000 previously scored responses. However, a portion of responses (about a quarter) will still be reevaluated by human scorers, particularly those deemed challenging for the AI, such as non-English responses or those containing slang.
Chris Rozunick, division director for assessment development at the TEA, emphasized that the AI system does not possess autonomous capabilities and does not “learn” from each response. Instead, it relies on its initial training to provide consistent scoring similar to humans. Despite the integration of AI, rigorous quality control processes remain in place to ensure accuracy and fairness in scoring.
The shift towards AI scoring signals a notable change in how assessments are conducted, underscoring the advantages of technology in education evaluation while still recognizing the vital role of human supervision and expertise in the process.
Filed in . Read more about AI (Artificial Intelligence).