There is a growing volume of evidence coming from practitioners around the types of interventions that can make a difference to individual outcomes, but there are gaps in the data because most rigorous impact evaluations focus on programme and school level interventions, not on system wide processes. It is also predominately evidence gathered from the global north or by those from the global north. The medical model continues to be reflected in evaluations of projects that focus on the steps required for specific groups of children (often meaning children with sensory and physical impairments) to be included rather than on the barriers and system biases that exclude. There is an over-reliance on learning from pilots and small-scale interventions (inputs to teacher training, accessible learning materials or Assistive Technology for example) which are standalone projects, often targeting the easiest to accommodate (probably the greatest volume) and not taking into consideration those who have the most complex needs (who are often excluded by design). Context is also important because the location, timing or background conditions (emergency, crisis, rural etc) can impact on whether or not the evidence and ‘what works’ are genuinely transferable. More contextualisation of the learning offered to governments or implementing partners is needed to ensure that it appears more immediately applicable.
Despite there being a lot of evidence, there is a significant gap in learning in relation to what works to stimulate system level transformation. UNICEF is currently synthesising global evidence on system level interventions across seven countries which is starting to identify important themes. However, more direct alignment between qualitative and quantitative data is needed to understand the nuances required for including children with disabilities at scale.