What can SEAMLESS do?

There are many potential applications of SEAMLESS technology to improve access to and use of IEPs. Click below for a few examples.

Nearly all districts (97%) currently conduct period reviews of completed IEPs, more than half of whom (55%) formally assess the quality of some or all IEPs based on a checklist or rubric. Additionally, state education agencies often include a review of IEPs as part of their monitoring of districts on implementation of the IDEA. These tasks are primarily done through manually reviewing IEPs, a time-consuming and expensive endeavor that limits the number of quality reviews a district or state can undertake. With SEAMLESS, all or part of this review could instead be completed in minutes. 

Reviewing IEPs is a major component of investigating complaints that students with disabilities are being denied their rights under the IDEA. When IEPs are investigated for due process hearings or compliance audits, this process typically involves high costs and weeks of time spent reviewing hundreds of documents. The initiation of these investigations is often significantly delayed as folks wait to get approval to access IEPs and other documents that contain protected student information. SEAMLESS can allow folks to clear that hurdle of access by automatically redacting this protected student information, substantially speeding up and lowering the costs of this work. Moreover, SEAMLESS could allow for more thorough analyses of substantive compliance questions at scale, like:

  • To what extent do present levels of observed performance statements describe both academic and functional/behavioral performance?
  • To what extent are descriptions of the type, amount, and location of IEP goals, services, and accommodations aligned to the student’s needs (e.g., not based on class period or overall schedule time segments), logically and appropriately connected, and designed to offer more than de minimis benefit? 

IEPs contain descriptions of the services and accommodations that will be provided to a student to help them reach their goals. These descriptions list the frequency with which services will be delivered (e.g., hours/minutes per week, number of times per week), as well as the staff member responsible for overseeing and implementing these services. This information could be collated across IEPs to forecast staffing needs, particularly of specialists and related services personnel who may serve multiple schools or districts. Moreover, services and accommodations could be compared across different students who have similar disabilities and IEP goals to ensure that all student needs are being appropriately met. 

Imagine being able to use IEPs to quickly flag and discontinue inefficient or insufficient services, goals, or accommodations. SEAMLESS can harness the power of AI to sort patterns in large data, helping schools, districts, or states discover what works and for whom. For example, we are investigating the extent to which SEAMLESS can track: 

  • The proportion of IEP goals that are considered clear and well defined, that include statements about measurement and timeframes, and that clearly link the goal back to the student’s statement of disability-related needs 
  • Whether students with similar disability profiles have similar IEP goals throughout their schooling 
  • Whether and which IEP goals, services, and accommodations are associated with more growth over time 

IEPs should consider how the cultural or linguistic background of a student affects their learning process and progress. Insights from SEAMLESS could be used to help identify districts where additional professional development or technical assistance may be needed to support non-native English speakers. For example, SEAMLESS can be trained to scan and flag IEPs for the presence or absence of the following elements for students learning a second language: 

  • Statements addressing how acquiring a second language impacts the student’s needs or goals, interpretations of assessments, and the student’s progress expectations 
  • Whether the student should be assessed or evaluated for disability in their native language 
  • Statements or insights from the student’s family or caregivers who can provide their perspective on what constitute appropriate IEP goals and expectations given the student’s values, attitudes, beliefs, and customs 
  • Inclusion of translated IEP passages into the student’s and family’s native language 

Understanding patterns in how IEPs are written across groups of students or schools could help identify targeted areas for technical assistance or professional development. Districts could disaggregate IEP data by student characteristics like race or ethnicity, sex, free or reduced-price lunch status, language status, and so on. Using a tool like SEAMLESS to redact and then summarize IEP data might enable these types of analyses in a way that has not yet been possible. 

We are exploring how capably SEAMLESS can automatically pull data needed for SPP/APR reporting in at least three areas: 

  • Placement decisions or educational environments: Can SEAMLESS automatically flag how many IEPs in a district list students’ placement decisions as inside the regular class 80% or more of the day? 
  • Primary disability category by racial or ethnic identity: Can SEAMLESS auto-populate a database indicating how many students are served in each IDEA disability category by race or ethnicity? 
  • Secondary transition: Can SEAMLESS flag whether an IEP includes measurable IEP goals related to the student’s transition services and needs that have been updated at least annually? Can SEAMLESS automatically code the transition services outlined on a student’s IEP, including courses of study?  

Students with disabilities are legally entitled to a free appropriate public education (FAPE) in the least restrictive environment (LRE). Instead of manually documenting LRE placement decisions into a database or file for SPP/APR reporting, we could explore the capabilities of SEAMLESS to automatically compile this information into a report across IEPs. This could facilitate deeper analysis of important questions about the characteristics of the services and supports provided to students receiving special education services, like: 

  • What specific services are students receiving in general versus special education settings?  
  • What amount of each type of service are students receiving outside of general education, and does this align with their LRE placement?  
  • To what extent do these service or educational environment patterns vary across disability categories or student characteristics?