Skip to content

A brief explanation of the “Smarties” and “Sally-Anne” tests of theory of mind

This post is a short discussion of the Smarties and Sally-Ann tests of theory of mind, written for a multimedia class blogging assignment. I hope readers may find it informative.

Theory of mind is both ability and modular cognitive framework (Baron-Cohen, 1995). On one hand it is the ability to infer mental states (e.g., thoughts, feelings, desires, etc., in oneself and others; Baron-Cohen, Leslie, & Frith, 1985; Baron-Cohen, 1995). On the other, it is conceptually comprised of several cognitive modules which work together to enable mental state inference (Baron-Cohen, 1995).

An array of studies have corroborated the theoretical claim of ToM – that autistic individuals have less of a capacity for mental state inference (see Baron-Cohen, 2001, for a review).

A few different tests of mental state inference have been designed over the years; perhaps most notable of these are the Sally-Anne and Smarties tasks. The Sally-Anne task involves two dolls – “Sally” and “Anne,” who represent two actual people in a social situation. In the experiment, the participating individual is first shown the two dolls and asked whether or not they know which is which.

Sally has a marble, which she puts into her basket before leaving the scene. Anne then walks over and puts Sally’s marble into her own basket. Sally returns. It is at this point that the experimenter asks the participant where the marble is. They then ask where Sally will think the marble is. While the correct answer is that Sally should think the marble is in her basket because that is where she left it, autistic participants (who often but not always have a weaker theory of mind) will in many cases respond that Sally will think the marble is in Anne’s basket, because they themselves know the marble is there but fail to take into account that Sally was not present when Anne moved the marble and therefore will not know this information. Two control questions are asked of the participant to ensure they know where the marble was in the beginning, and where it really was at the end.

Another test of theory of mind is the “Smarties” task. Smarties, a British candy similar to American M&Ms. Here, a candy container similar to a mini M&Ms tube, is shown to the participant, who is asked what they think the container will contain. The participant, of course, will respond expecting smarties or M&Ms. However, the experimenter has replaced the candy with pencils, to the participant’s surprise. The participant is then asked what another, hypothetical person (or an actual person in the room) will think is in the container. Again, as with the Sally-Anne task, the participant who is weaker in ToM will expect the other person to know what they themselves know (in this case, that pencils are in the container instead of candy), as they do not recognize that the other person has not actually seen the pencils in the container and therefore would expect candy.

Both of these tasks are demonstrated in the following video.



Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a “theory of mind”? Cognition, 21, 37–46.

Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. Boston: MIT Press/Bradford Books.

Baron-Cohen, S. (2001). Theory of mind and autism: A review. International Review of Research in Mental Retardation, 23, 169–184. Retrieved from

Published inUncategorized

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *