The Effectiveness of Robotics to Foster Collaboration Among Children with Autism

Wainer, J., Ferrari, E., Dautenhahn, K., & Robins, B. (2010). The effectiveness of using a robotics class to foster collaboration among groups of children with autism in an exploratory study. Personal and Ubiquitous Computing, 14(5), 445–455.

In this 2010 study, Wainer, Ferrari, Deutenhahn, & Robins present a compelling argument for the effectiveness of collaborative robotics activities for developing the often impaired social skills of children with autism. The study seeks to answer two question: “whether interacting with robots over an extended period of time in a group-based setting would result in an increase in collaborative behaviours displayed by the children” (p. 446) and “whether the children would be able to generalize their collaborative behaviours from a robotics class into another, less structured domain” (pg. 447). The authors ran weekly robotics classes for children with autism over 4 months, and analyzed the interactions and behaviours of 7 students who attended at least 60% of the classes. Sprinkled throughout the study, the researchers also engaged the students in free-form drawing activities to test for the above mentioned cross-disciplinary generalization of any growth in social behaviours. A variety of data collection techniques were used. After each class, students and their parents were asked to complete 7 point Likert scale questionnaires to measure students’ enjoyment of  the class, their perceived ability to work with others that class, and their parents’ perceptions of their child’s behaviour outside of class. Video analysis was used to catalog moments when three specific behaviours took place concurrently, which acted as the definition for potentially collaborative moments in this study. In this context, the authors were looking for moments when children were displaying one or more of three social behaviours (robot related speech, pointing, and shared positive affect such as smiling or laughing together) while both in close proximity to, and looking at the same object as their group mates. Finally, semi-structured interviews with the parents were used to further discuss changes in their child’s collaboration and social skills. The results of this study indicate that programs like the one implemented by these authors can positively effect the collaborative and social skills of children with autism. Of note, the data collected indicates that student enjoyment of class, more so than the raw amount of time a student spends with a given group, is strongly correlated to their gains in social development.

This is the kind of paper I can get behind. On the surface, it is well written, coherent, and has clearly defined goals and scope. The authors don’t dwell on situating the study within a melange of theoretical frameworks, as some others might, but get right into the details after a brief look at past work in this field and some descriptive text on the nature of autism. They go into just the right amount of detail on the methods used in the study to give me faith in the outcomes of the study without boring me with minutia. The study itself is very interesting, and the questions asked lend themselves to conclusions which are potentially actionable in informing policy and program development. As such, the conclusions are cautious, indicating only a correlation between student enjoyment of the robotics classes and their social development, while it remains unclear whether the enjoyment comes from learning to collaborate or that learning to collaborate is a result of enjoyment of the class.

This is an great example of design-based research, as over the course of many classes the researchers freely to integrated tools and practices, such as “turn taking wheels” (p. 448) into they learned more about the context they were working in. Though this paper was 4 years too early to draw from Enyedy & Stevens’s (2014) framework for analyzing collaboration (who would likely categorize this as a study in collaboration-as-learning), care could have been taken to further define what the authors considered “collaboration” beyond simply laying out their 3 criteria.

While this study shows promise in the benefits of collaborative robotics activities for children with autism, it does remain to be seen if this effect is specific to the domain of robotics. The literature review at the beginning of the paper indicates that some key aspects of robotics in particular are exceptionally engaging practice for children with autism, but future studies could attempt to replicate these results in disciplinary practices other than robotics.


Wainer, J., Ferrari, E., Dautenhahn, K., & Robins, B. (2010). The effectiveness of using a robotics class to foster collaboration among groups of children with autism in an exploratory study. Personal and Ubiquitous Computing, 14(5), 445–455.

Enyedy, N. & Stevens, R. (2014). Analyzing collaboration. in K. Sawyer (Ed.) The Cambridge handbook of the learning sciences (191-212). New York, NY: Cambridge University Press.