AI and IEPs: Can Technology Improve Quality and Reduce Special Educators’ Workload?
Danielle Waterfield knows first-hand the unique weight of the responsibilities held by special education teachers.
Before earning her doctorate in special education at the UVA School of Education and Human Development in May, Waterfield taught special education for four years and then served as an administrator, overseeing the special education identification processes for another two years.
Nationwide, special education teachers are on the front line of serving the approximately 15% of all public school students who receive special education services.
“There are demands on all types of teachers,” Waterfield said. “Just scratching the surface—general education teachers are responsible for things like classroom management, communicating with families, preparing students for assessments, etc. It’s a lot.
Special education teachers are responsible for all those things and for upholding all the legal compliance obligations that exist to protect students with disabilities.DWDanielle Waterfield
One of those legal requirements is the individualized education program, or IEP. A yearlong legal contract between each special education student, their family, and the school, the IEP maps out three to five learning goals for the year, how progress toward those goals is measured, and what services and accommodations will be provided to the student.
“On average, it took me between three to five hours to write and finalize each IEP with the families and partnering service providers at my school,” Waterfield said. “And my caseload, which was typically small to average-sized, ranged from ten to nineteen students, depending on the year.”
Finalizing IEPs is just the beginning. Special education teachers are then responsible for teaching and measuring progress toward the identified goals, ensuring their students receive the services identified in the IEP.
Burnout is common. Special education teachers are more likely to leave their jobs than general education teachers, often citing overwhelming paperwork and compliance responsibilities.
Waterfield, now a research scientist at the UVA School of Education, has been asking—and examining—can artificial intelligence effectively and responsibly be leveraged to support special educators and their students?
Waterfield and former UVA School of Education researcher Olivia Coleman first wanted to know if AI could help special education teachers increase the quality of their students’ IEP goals, a critical starting point of the IEP process.
Inspired by an earlier study that found ChatGPT increased the quality of IEP goals written by preservice and novice teachers, Waterfield and Coleman wanted to know if the same was true for more experienced teachers.
With data collected in 2023, the researchers found no statistically significant difference in the quality of goals written by ChatGPT and those written by more experienced teachers.
“This was good news,” Waterfield said. “It shows that teachers can gain the skills they need to produce high-quality goals.”
High quality goals are only one element of what Waterfield calls “legally defensible IEPs.” They must also include goal monitoring, measurement rubrics and other elements. Then comes the obvious question: Can AI improve the quality of these elements of IEPs?
This question is one of many Waterfield and her colleagues are curious about.
Another possible role for AI in special education is reducing the overall time spent on paperwork, often cited by special education teachers as overwhelming. Did using AI reduce the time these teachers spent writing those quality goals? Can it be used to reduce other elements of the workload, such as monitoring a student’s progress toward their goals.
“For example, progress monitoring is really difficult to do,” Waterfield said. “Imagine a teacher has a student with a math goal. To get there, the teacher might create worksheets for the student to do each week to monitor their progress.”
The “individual” part of the IEP is what can make these education plans so labor intensive. Each goal for each individual student determines the specific milestones of monitoring: what materials are needed to test a student’s progress and how often they are used. Which has Waterfield asking, could AI help streamline each student’s assessment materials? Can AI help analyze the assessment data?
“Reducing the time teachers spend on paperwork can then increase their face time with students, which we know improves student outcomes,” Waterfield said. “We knew that before AI was here.”
And better outcomes for students is the whole point.
While researchers are just beginning to answer some of these questions, many schools and teachers are already using AI.
“Districts and schools should have guidance for AI use,” Waterfield said. “Some districts have entered into partnerships with specific AI tools. But if they don’t have specific guidelines or guidance, teachers should check in with their administrators before using any AI tools.”
What makes this especially tricky in special education, Waterfield notes, is that these teachers are responsible for understanding the legal side of using AI as it relates to students with disabilities.
As research continues in how AI can be used effectively in special education, Waterfield urges researchers and those training teachers to use AI to center the most important voices.
“Student, teacher, and, when appropriate, family voices and their experiences need to be centered in this work to truly reflect and respect each unique circumstance,” she said. “And teacher preparation programs and professional development opportunities should also use this perspective, in addition to adhering to legal requirements and other policies.”
Ultimately, Waterfield is optimistic about how AI can support special education teachers in their work making truly individualized, equitable learning experiences for the students with disabilities they serve.
Read more research news from School of Education.
This article was originally published here: https://education.virginia.edu/%3Ca%20href%3D%22/news-stories/ai-and-ieps-can-technology-improve-quality-and-reduce-special-educators-workload%22%20hreflang%3D%22en%22%3Eview%3C/a%3E