ManyBabies5: A large-scale, collaborative investigation of the the Hunter & Ames Model of infant looking preference

• In Event: Updates from ManyBabies: Collaborative efforts to promote best practices and improve robustness in developmental science

Abstract

Since Fantz’s seminal studies in the 1960s, infants’ fixations to familiar versus novel stimuli have been widely used to make inferences about infants’ cognitive processing. With the rise of infant looking time methods in cognitive development research, the conditions under which infants show preferences for familiar or novel stimuli has been a key question in infancy research. In 1988, Hunter & Ames proposed a model of infants’ preferences for familiar and novel stimuli, suggesting that this preference is affected by infants’ age, amount of familiarization, and stimulus complexity. This influential model has been cited a total of 578 (google scholar, retrieved 2020-09-22) times by papers investigating visual attention, speech segmentation, object categorization, face recognition, quantification, memory, knowledge of the physical world, social cognition, and more. While this model provides widely-used guidance for what to expect given certain features of stimuli, there has not yet been a large-scale investigation directly testing the predictions of the Hunter & Ames model.
ManyBabies5 is a collaborative research project conducting a large-scale, cross-cultural investigation of the Hunter & Ames model of infant looking preference. The goal is to form a consensus-based experimental design that tests Hunter and Ames’ (1988) prediction that infant age, familiarization time, and stimulus complexity interact to predict the trajectory of infants’ looking to novel vs. familiar stimuli. Thus far, 192 researchers from 118 labs spread across 40 countries and 6 continents (Figure 1) have signed up to participate. The ManyBabies5 team has held its first meetings and has begun collaboratively working on experimental design, with the goal of finalizing design and analysis details and submitting a Registered Report by April 2021.
The currently proposed design of the task is modeled after past studies testing the Hunter and Ames model (e.g., Rose et al., 1982). In the task, infants will first be familiarized with a visual stimulus for a fixed familiarization time (Figure 2). Then, the familiarized stimulus will be presented alongside a novel stimulus matched in visual complexity. The main dependent variable of interest will be infants’ looking time to the familiarized vs. the novel stimulus. We will manipulate three main factors of interest: (a) infants’ age (between-participants), (b) how long infants are familiarized to a given stimulus (within-participants), and (c) the complexity of the visual stimuli (within-participants). The prediction (following Hunter & Ames) is that (a) infants should shift from a familiarity to a novelty preference faster as they grow older; (b) longer familiarization times should lead to shifts in infant looking towards a novelty preference; and (c) more complex visual stimuli should lead to shifts in infant looking towards a familiarity preference.
In our presentation, we will share the final version of our experimental design and procedure, informed by researchers from a variety of backgrounds and communities. We will also outline our plans for collecting data from a large, culturally-diverse sample of infants. The project aims to significantly advance our understanding of the factors influencing infants’ behavior in preferential looking paradigms, and their generalizability across participants and communities from around the world.

Group Authors

The ManyBabies 5 Consortium

Authors

• Jessica Elizabeth Kosie, Princeton University
  Presenting Author

• Martin Zettersten, Princeton University
  Non-Presenting Author

• Christina Bergmann, Max Planck Institute for Psycholinguistics
  Non-Presenting Author

• Dima Amso, Columbia University
  Non-Presenting Author

• Casey Lew-Williams, Princeton University
  Non-Presenting Author