Academics within Lancaster University’s Psychology Department have been examining how babies and infants sample information about two unfamiliar, animal-like categories, using a new method that allows them to interact with what is shown on the screen via their looking behaviour. They found babies tend to repeatedly engage with one information source before switching to and exploring the other. However, they widely differ in their sampling patterns, from more exploitative to more explorative.

Infants are active learners who explore the world based on their curiosity, but little is known about how they structure their exploration. Previous research has focused either on general exploration patterns in free play observations or on engagement with presented stimuli in highly controlled passive looking studies.

Researchers from Lancaster’s Infant and Child Development Lab led by PhD student Elena Altmann under the supervision of Dr Marina Bazhydai and Professor Gert Westermann, developed a new paradigm to measure even young babies’ curiosity-based information sampling using gaze-contingent eye-tracking. This method uniquely bridges free play study designs and highly controlled passive looking designs by allowing the baby to interact with the computer program via their looks. Hence, their fixation acts like the finger pushing a button on a touchscreen.

In the study, babies were introduced to animal-like creatures that none of the babies had seen before. Two ‘families’, or categories, of these creatures, were associated with either of two identical houses on the screen. Looking at either house triggered the presentation of a new member from the respective category. Each trigger represented a sampling decision to stay with the current category (exploit) or switch to the other (explore). In this way, babies could sample based on their own, in-the-moment curiosity leading to differences in how many members they wanted to see from either family.

On the results of the experiment, Elena said: “Our findings suggest babies’ sampling choices are not random but informed by their exploration history in a way that allows them to take in the available information and robustly encode it. However, they also highlight the need to acknowledge and better understand individual differences, as some more exploitative babies might lose out on discovering other sources of information, and some highly explorative babies may not engage long enough with one source to process the information as deeply as necessary to construct new knowledge.”

In another funded project, the research team are developing a new caregiver questionnaire to measure trait curiosity in infants and toddlers, that is their tendency to interact with and explore their immediate surroundings to gain new information. With this, they will be able to explain some of the differences in exploration patterns observed in this study.

Elena’s PhD supervisor, Professor Gert Westermann added: “This innovative paradigm not only allows us to investigate fundamental aspects of babies’ moment-to-moment information sampling but also paves the way for new investigations of how specific characteristics of the available information affect curiosity-based, active exploration.”

Full details of the results can be found in the paper published in Cognition.