Babies cannot tell us what they see. So researchers watch how long they look. Show the same picture again and again, and looking time drops. That is habituation. Then show something new. If looking time jumps back up, the baby noticed the change. That jump is called dishabituation. Run the study on a model baby below.
The baby's screen
Press Show the picture to start trial 1.
Looking time, trial by trial
Trials0
First 3 average-
Last 3 average-
Step 1: habituateshow the same picture until looking drops by half
Model baby's age
Step 2: pick the test picture
How the study worksResearchers set a rule ahead of time. Here the rule is: the baby is habituated once the average of the last 3 trials falls below half the average of the first 3. Then the test picture comes on. The test buttons unlock when the rule is met.
This is a model, not real data. The numbers follow the general pattern found in infant studies. Real babies vary a lot. Some studies also find that a baby may look longer at something familiar before turning to new things.