Methods

This method measures where and for how long infants look at different experimental displays. Using precisely designed stimuli, infants’ visual attention patterns reveal what kinds of changes they notice, what types of events they expect, and how they process visual information. 

We use two different types of visual attention techniques in our research. The first technique is live looking time coding, in which a trained observer sits out of view of the infant and records how long the infant looks at different events using a computer program. The second technique makes use of an eye tracker to record detailed information about where infants are looking within a particular display. The eye tracker emits a harmless infrared light and measures the reflections from the infant’s cornea and pupil to precisely record the infant’s gaze patterns and pupil dilation. These behaviors reveal new insights into infants’ perception and understanding of their environment.

Infants’ reaching behavior can be as informative as their visual attention. Choice paradigms present two objects to infants simultaneously and use the infants’ reaching behavior as an indication of their preference for one object over the other. 

Many of our studies use a “time-to-touch” paradigm in which we measure infants’ latency to touch individual objects that are presented serially, as well as measuring their subsequent exploration of each object. By comparing how infants touch different types of stimulus objects, we can make inferences about the types of objects infants avoid. In general, we present all of our stimulus objects for a given study to each participant to account for individual differences in reaching behavior. We also randomize the order in which objects are presented across participants to mitigate order effects.

Video recordings of infants’ and young children’s behavior can be rich sources of data. To standardize the types of behaviors we are interested in, we develop coding schemes that precisely operationalize what is counted as, for example, a “look” or a “touch” in a particular study. 

With the help of behavioral coding software, our research assistants comb through video recordings to identify how often and long behaviors of interest occur, as well as the time sequence. This technique can be used with videos of our tightly controlled laboratory studies, as well as with videos of infants and young children in naturalistic settings.