The false belief task is a psychological test designed to measure whether a person, usually a young child, can understand that someone else can hold a belief that differs from reality. It is the most widely used method for assessing “theory of mind,” the ability to recognize that other people have their own thoughts, knowledge, and perspectives that may not match your own. Children typically begin passing the standard version of this task around age 4, marking a key milestone in social and cognitive development.
How the Classic Task Works
The most famous version is the Sally-Anne test, first developed by researchers Heinz Wimmer and Josef Perner in 1983. A child watches a simple story acted out with dolls or props. Sally places a marble in a basket and then leaves the room. While she’s gone, Anne moves the marble to a different location, say a box. Sally comes back, and the child is asked: “Where will Sally look for her marble?”
The correct answer is the basket, because that’s where Sally believes the marble still is. She wasn’t in the room when it was moved, so she has no way of knowing it’s been relocated. To get this right, a child has to set aside what they personally know (the marble is in the box) and think about what Sally knows instead. That mental leap, separating your own knowledge from someone else’s, is what the task is testing.
In the original 1983 study of 36 children aged 3 to 9, none of the 3- to 4-year-olds pointed to the correct location in both test scenarios. Among 4- to 6-year-olds, 57% got it right. By ages 6 to 9, 86% answered correctly.
The Unexpected Contents Version
A popular variation is the “Smarties task,” sometimes called the unexpected contents task. A child is shown a familiar candy tube (like a Smarties box) and asked what they think is inside. Naturally, they say candy. The tube is opened to reveal pencils instead. Then the child is asked: what would another person, who hasn’t seen inside the tube yet, think it contains?
Children who have developed false belief understanding will say the other person would think it holds candy, because that’s what the packaging suggests. Younger children, who haven’t yet grasped that other people can hold mistaken beliefs, tend to say the other person would know about the pencils, as though everyone shares the same knowledge they do.
Why Age 4 Is a Turning Point
Most 2- and 3-year-olds consistently fail the standard false belief task. They point to where the object actually is rather than where the character believes it to be. Around age 4, children begin passing reliably, and performance continues improving through age 5 and 6. At age 3, only about 30% of children give the correct verbal answer in standard testing. By age 4, this number rises sharply, and it reaches a ceiling in adults.
The picture gets more interesting with younger children. Studies using eye-tracking technology, which measures where babies look rather than asking them to answer questions, suggest that infants as young as 13 to 15 months show some sensitivity to other people’s false beliefs. In one experiment, 15-month-olds watched an agent hide a toy and then observed the toy being moved while the agent was away. When the agent returned and reached for the wrong location, the infants stared longer, as if surprised. By 25 months, toddlers in nonverbal tasks can correctly anticipate where a person with a false belief will search for an object.
This creates a puzzle: if babies seem to track false beliefs before their second birthday, why do children still fail the verbal task until age 4? The gap between what infants appear to understand implicitly and what older children can express explicitly remains one of the most debated questions in developmental psychology.
What the Task Actually Measures
One major criticism of the false belief task is that it may not purely measure social understanding. To pass, a child also needs working memory (to hold the story details in mind), language comprehension (to understand the question), and impulse control (to resist blurting out what they know is true). When researchers reduce these extra demands, younger children’s performance improves significantly.
A particularly important factor is what researchers call the “curse of knowledge,” the natural human tendency to assume that others know what you know. Adults experience this too, but it’s much stronger in young children. Some researchers have argued that the difference between a 3-year-old’s performance and a 5-year-old’s performance reflects a change in how well children can overcome this bias, not necessarily a change in their underlying understanding of beliefs. In other words, a 3-year-old might grasp that Sally doesn’t know the marble moved but still be pulled toward answering with what they themselves know to be true.
This has raised questions about decades of research that used false belief performance as a predictor of social skills and other outcomes. Those findings may partly reflect differences in the ability to set aside one’s own knowledge rather than differences in social cognition itself.
First-Order vs. Second-Order Tasks
The standard Sally-Anne test is a first-order false belief task: “What does Sally think?” Second-order tasks go a step further, asking what one person thinks another person thinks. For example: “Where does Sally think Anne thinks the marble is?” This layered reasoning, sometimes described as “I think that you think that she thinks,” is considerably harder and typically develops later in childhood. Classic second-order tasks like the “Ice cream seller” scenario have been used to track this more advanced social reasoning in school-age children.
Connections to Autism
The false belief task gained particular prominence through research on autism spectrum disorder. In studies comparing children with autism (average age around 7.5 years) to age-matched typically developing children, the children with autism did not perform above chance level on false belief tasks, while typically developing children performed significantly better. Difficulty with false belief reasoning aligns with the broader social communication challenges that characterize autism, though it’s worth noting that many autistic individuals do eventually learn to pass these tasks, sometimes through different cognitive strategies rather than the intuitive social reasoning that typically developing children use.
What Happens in the Brain
Brain imaging studies have identified a consistent network of regions that activate during false belief reasoning. Two areas play central roles. One, located at the junction of the temporal and parietal lobes on the right side of the brain, is involved in tracking what another person believes, essentially building a model of someone else’s mental state. The other, in the middle front part of the brain, activates when evaluating outcomes, particularly when something violates either your own or another person’s expectations. These two regions appear to handle different aspects of the task: the first tracks beliefs as they form, while the second processes the results of those beliefs. This same network activates whether someone is deliberately reasoning about another person’s beliefs or doing so spontaneously, suggesting it reflects a core social processing system rather than a skill that only kicks in during effortful thinking.