How Do Children Reason About Mirrors? A Comparison Between Adults, Typically Developed Children, and Children With Autism Spectrum Disorder

Abstract

The information about what one can see and what other people can see from different viewpoints is important. There are circumstances in which adults and children make systematic errors when predicting what is visible from their own or others’ viewpoints. This happens for example when reasoning about mirrors. We explored differences among three developmental groups: young adults (N=60) typically developing children (N=30); and children with autism spectrum disorder (ASD, N=30). We used an illustration of a top-down view of a room with a mirror on a wall (Room Observer and Mirror Perspective test: ROMP). Participants selected (circled on paper) which objects behind the observer in the room were visible, reflected from the mirror and from a given position (viewpoint). For half of each group, the observer in the room was described as a teddy bear; for the other half, it was described as a child. Overall, there were many errors in all groups, which we separate in errors of ignoring the viewpoint (same response to all three locations) and inversion errors (choosing objects on the left instead of the right or vice versa). In addition to the overall task difficulty, the ASD group made relatively more mistakes of ignoring the viewpoint compared to the other groups and underestimated how many objects were visible in the teddy bear condition that is when the viewpoint was an inanimate object. We suggest that this is related to a delay in theory of mind (ToM) development.

Both the domain-general and the mentalising processes affect visual perspective taking

“…For example, in Posner’s spatial cueing task (Posner & Cohen, 1984), which does not require any decisional process, Hayward and Ristic (2018) showed that a directional cue directs attention regardless of its social features. Conversely, in a task that engages only a decisional process, as in the Room Observer and Mirror Perspective test (ROMP, Bertamini & Soranzo, 2018; Soranzo et al, 2021) in which participants were asked to judge how many targets are visible from a given position indicated by a cue, an advantage emerges for social cues compared to non-social cues.…”