AQA Syllabus focus:
'Baillargeon’s explanation of early infant abilities, including violation of expectation research.'
Violation of expectation research examines whether infants look longer at events that break their understanding of the world. Baillargeon used this method to argue that infant cognition is more advanced than previously thought.
What violation of expectation research means
Violation of expectation research is based on a simple idea: if infants already have some knowledge about how the world works, they should pay more attention when that knowledge is contradicted. Because very young infants cannot explain what they know, researchers infer their understanding from looking time.
Violation of expectation research: A method in which infants are shown a possible or impossible event, and longer looking at the impossible event is taken to show surprise at a violated expectation.
In this approach, an impossible event is one that appears to break a basic rule of the physical world, such as one object passing through another or a hidden object seeming to disappear. A possible event matches what would normally happen in everyday experience. If infants look reliably longer at the impossible event, researchers argue that the infant expected something else to occur.
This method was especially important for Baillargeon’s explanation of early infant abilities because it suggested that infants may understand more than older theories assumed. Instead of judging knowledge only by what infants can physically do, violation of expectation studies examine what infants appear to mentally represent.
How the method works
Researchers usually begin by familiarizing infants with a display so that their attention becomes stable. This reduces the chance that longer looking is caused simply by the event being new.
Habituation: A decrease in attention after repeated presentation of the same or a similar stimulus.
After habituation, infants are shown carefully designed test events.
A simple habituation curve showing looking time decreasing across repeated trials. This visual reinforces how researchers operationalize habituation as reduced attention to a repeated stimulus, which helps ensure later increases in looking are not just due to novelty. Source
The logic is usually:
infants watch a display several times
they then see either a possible outcome or an impossible outcome
the researcher records how long they look at each event
significantly longer looking at the impossible event is interpreted as a sign of surprise
This procedure is useful because it can be used with preverbal infants. It does not require pointing, searching, or speaking. That makes it a sensitive method for studying early knowledge, especially when motor skills are still limited.
Baillargeon’s classic use of the method
Baillargeon became well known for using violation of expectation tasks to test whether infants understood hidden objects and physical constraints. A classic example involved a rotating screen and a box placed behind it.
In the familiarization phase, infants watched a screen rotate through a full arc. Then a box was placed in the screen’s path. Two kinds of test event were presented:
Possible event: the screen rotated until it reached the hidden box, then stopped
Impossible event: the screen appeared to rotate through the space where the box should have been
Infants tended to look longer at the impossible event. Baillargeon argued that this was because they represented the box as still being there even when they could not see it. If the screen moved through that hidden space, the event violated the infant’s expectation about how solid objects behave.
The importance of this finding was that infants showed understanding through attention, not through an overt action such as searching for the object. Baillargeon therefore claimed that young infants may possess early knowledge of the physical world that is not always visible in their behavior.
What the findings suggest about early infant abilities
Baillargeon interpreted longer looking as evidence that infants are not passive observers. Instead, they appear to build expectations about what should happen next. When an event breaks those expectations, they attend for longer.
This leads to an important idea in interpreting infant behavior: infants may show greater cognitive competence than their actions alone suggest. In other words, a young infant might understand a physical event but still be unable to demonstrate that understanding through coordinated movement.
Violation of expectation studies therefore shifted attention away from the question “Can the infant perform the task?” toward “What does the infant appear to know?” This is why the method has been so influential in research on early development.
Strengths and limitations of violation of expectation research
A major strength is that the method is well suited to infancy. Since the dependent measure is usually looking time, it avoids many problems caused by immature motor skills. It also allows researchers to test very young infants who would fail more demanding tasks for reasons unrelated to knowledge.
Another strength is the high level of experimental control. Possible and impossible events can be designed so that only one critical feature differs. This helps researchers isolate which expectation is being violated.
However, the method also has important limitations. The main criticism is that looking longer is only an indirect measure of understanding. An infant may look longer because an event is visually more interesting, more novel, or more complex, rather than because it is logically impossible.
Researchers try to deal with this by using matched displays, careful controls, and repeated replications. Even so, interpretation remains difficult because the infant cannot explain why they are looking.
A further limitation is that small changes in timing, angle, brightness, or movement can affect looking time. This means findings must be interpreted cautiously. Strong evidence comes not from one study alone but from a consistent pattern across many violation of expectation experiments.
For exam answers, the key issue is whether longer looking reflects genuine knowledge of the event or only a change in attention caused by perceptual differences.
Practice Questions
Outline what is meant by violation of expectation research. (2 marks)
1 mark for stating that infants are shown a possible/expected event and an impossible/unexpected event.
1 mark for stating that longer looking at the impossible/unexpected event is taken to indicate surprise or an existing expectation.
Explain what Baillargeon’s violation of expectation research suggests about early infant abilities. Refer to one example in your answer. (6 marks)
1 mark for explaining that looking time is the key measure.
1 mark for stating that infants tend to look longer at impossible events.
1 mark for describing a relevant Baillargeon-style study, such as the rotating screen with a hidden box.
1 mark for explaining why the event is impossible, for example the screen appears to move through the space occupied by the hidden box.
1 mark for stating that infants seem to represent the hidden object or physical constraint.
1 mark for linking the finding to Baillargeon’s claim that early infant abilities are more advanced than behavior-based tasks suggest.
FAQ
Researchers usually set a numerical criterion before the study begins.
Common examples include:
a drop in looking time to 50% of the initial level
a drop across a set number of consecutive trials
a moving average based on the infant’s first few looks
This matters because test trials should begin only after the display is no longer especially novel. If habituation is judged too early or too late, the violation of expectation effect can be weakened.
There are several methods, and many studies use more than one.
Typical approaches include:
a trained observer pressing a key while the infant looks
video recordings coded later
blind coding, where the coder does not know which event was shown
eye-tracking equipment that measures gaze direction and duration
Blind coding is especially valuable because it reduces observer bias. If coders know which trial is “impossible,” they may unintentionally overestimate looking.
Small effects can occur for practical and methodological reasons.
For example:
infants may be tested at slightly different ages
the display may look different from the original version
modern controls may remove accidental cues that helped the original effect
sample sizes may still be fairly small
A smaller effect does not automatically mean the original finding was wrong. It may mean the phenomenon is real but sensitive to procedure, or that the original estimate was larger than the true effect.
A novelty preference happens when an infant looks longer simply because one event is newer or visually less familiar.
A true violation-of-expectation response is narrower. It means the infant looks longer because the event conflicts with an expectation about what should happen.
Researchers try to separate the two by:
matching the visual features of possible and impossible events
changing only the logically critical element
counterbalancing order
including control conditions
Without these safeguards, a longer look may reflect novelty rather than surprise.
Using multiple coders improves reliability.
If two independent observers produce similar looking-time scores, researchers can be more confident that the result is not due to one person’s judgment. This is especially important because infants may glance away briefly, move suddenly, or appear to look without actually attending.
Researchers often calculate inter-rater reliability after coding. High agreement strengthens confidence in the data and makes the final interpretation of any violation-of-expectation effect more convincing.
