Using Machine Learning to Predict Patterns of Employment and Day Program Participation.
A simple random-forest model using survey answers spots adults with IDD who are ready for competitive work 89% of the time.
01Research in Context
What this study did
The team fed NCI-IPS survey answers into a random-forest computer model.
Goal: guess if an adult with IDD had a real job or only attended a day program.
They tested the model on a fresh slice of data to see how often it was right.
What they found
The model guessed correctly 89 times out of 100.
The top clue was whether the person listed community employment as a goal.
How this fits with other research
Préfontaine et al. (2024) did the same trick for preschool kids. Their ML tools beat random guessing when forecasting how much autistic toddlers will progress in early ABA.
Song et al. (2022) also used ML, but to spot ID inside autism. Together these papers show the same math works across ages and questions.
Holwerda et al. (2013) used old-school logistic regression on young adults. They found family support predicts work, while the new ML finds goal statements matter more. The methods differ, but both say personal factors drive jobs.
Why it matters
You can add the 89% model to your transition plan today. Run the free NCI-IPS items through the forest before the annual ISP. If employment is not flagged as likely, add self-advocacy goals and community work trials. The tool costs nothing and gives you a data-backed reason to fight for real jobs instead of default day programs.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Score last year’s NCI-IPS survey for each adult client; flag anyone the model predicts as “employed” and schedule a community job try-out this month.
02At a glance
03Original abstract
In this article, we demonstrate the potential of machine learning approaches as inductive analytic tools for expanding our current evidence base for policy making and practice that affects people with intellectual and developmental disabilities (IDD). Using data from the National Core Indicators In-Person Survey (NCI-IPS), a nationally validated annual survey of more than 20,000 nationally representative people with IDD, we fit a series of classification tree and random forest models to predict individuals' employment status and day activity participation as a function of their responses to all other items on the 2017-2018 NCI-IPS. The most accurate model, a random forest classifier, predicted employment outcomes of adults with IDD with an accuracy of 89 percent on the testing sample, and 80 percent on the holdout sample. The most important variable in this prediction was whether or not community employment was a goal in this person's service plan. These results suggest the potential machine learning tools to examine other valued outcomes used in evidence-based policy making to support people with IDD.
American journal on intellectual and developmental disabilities, 2021 · doi:10.1352/1944-7558-126.6.477