Comparing Human Video Modeling to Animated Video Modeling for Learners with Autism
Animated video modeling works as well as human videos for many kids with autism—try both and let the data pick the winner.
01Research in Context
What this study did
Bloh et al. (2025) asked a simple question: do kids with autism learn better from a real person on video or from a cartoon?
They showed eight children two kinds of clips. One clip had a real adult doing a task. The other clip showed an animated character doing the same task.
The team used an alternating-treatments design. Each child saw both clip types and the order flipped across days.
What they found
Seven of the eight kids mastered the skill no matter which clip they watched first.
Some kids learned faster with the human video. Some learned faster with the cartoon. The group average landed in a tie.
How this fits with other research
Older studies also found ties. Bailey et al. (2010) saw no winner between live and video models. Richman et al. (2001) saw no gap between self and other videos. The new data say the same thing about human versus animated clips.
Yet a few papers report clear champs. Marcus et al. (2009) showed self-modeling beating peer modeling. Wilson et al. (2020) showed full-video beating video-prompting. The difference is the skill taught. When the task is simple imitation, format barely matters. When the task needs fine steps, the model type can tip the scale.
Two big meta-analyses back the overall safety of either pick. Hong et al. (2016) and Storch et al. (2012) both found solid gains across dozens of human-video studies. Bloh’s cartoon data now widen that same toolbox.
Why it matters
You no longer need to hunt for a volunteer to film every model. If an animated clip is easier to make or more fun for the child, use it. Run a quick probe with both, track the data, and let the learner’s performance choose. One kid may light up for cartoons while another tunes in to real people. Let data, not habit, drive the pick.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Film the same task twice—once with a staff member and once with a free animation app—then run two five-trial probes and chart which one hits mastery first.
02At a glance
03Original abstract
People with autism spectrum disorder (ASD) may have difficulty responding vocally with intraverbals and physically with motor imitation during conversations. Not responding with an appropriate word coupled with an absence of body language could compromise social opportunities. The literature lists scores of studies implementing human video modeling to increase skills of people with ASD but not much research has been conducted using animated video modeling (Kellems et al., 2020). This study compared human video modeling to animated videos to teach vocal intraverbal responding along with motor imitations of facial expression and body language to eight children with ASD. Seven of the eight participants acquired the target behaviors with one or both methods to some degree. Two participants demonstrated more of the target behaviors with the human video, three demonstrated more with the animated, and little difference in learning was observed for three participants. One participant only demonstrated target behaviors following the human video and another only demonstrated target behaviors following the animated video, suggesting that both methods could be effective and neither was conclusively superior.
The Analysis of Verbal Behavior, 2025 · doi:10.1007/s40616-025-00224-y