On May 8, TEA Deputy Commissioner of Analytics, Assessment and Reporting, Iris Tian, provided testimony and answered questions from members of the Artificial Intelligence Advisory Council about the agency’s new automated scoring engine (ASE) for STAAR RLA tests. The AI Advisory Council was established by HB 2060 (88R) to “study and monitor AI systems developed, employed, or procured by state agencies.” Rep. Giovanni Capriglione (R-Southlake) and Sen. Tan Parker (R-Flower Mound) authored the bill and co-chair the council.
Ms. Tian’s testimony and PPT presentation includes many of the same information that Commissioner Mike Morath has shared with superintendents and assessment directors in other meetings.
The committee’s co-chairs asked specific questions regarding training the ASE, margins of error, student data privacy, parent notification of the results, and what, if any, benefits are there to taxpayers.
The final question, asked by Chair Capriglione, was one that many ELAR teachers have asked. Will AI scoring of the STAAR test deter student creativity? Ms. Tian responded by saying: 1) a large reason why TEA added open-ended questions was to improve student learning; 2) part of the reason why TEA has hybrid scoring is to ensure that if there is a unique response, it can be routed to a human scorer; and 3) the agency is being “extremely clear” on how the rubric is applied.