Bringing home Baby Euclid: Evaluating infants’ basic shape discrimination using the online platform Lookit

• In Event: Infants’ learning about object properties and categories in diverse environments

Abstract

Discriminating shape information is essential to learning about objects (Quinn & Eimas, 1997; Quinn et al., 2001; Smith, 2009) and is at the foundation of school-readiness in STEM (Verdine, Golinkoff, Hirsh‐Pasek, & Newcombe, 2017). Studies conducted in highly controlled lab settings have revealed that 7-month-old infants already discriminate basic 2D figures that differ in shape information, even when those figures are also varying in their scale, orientation, and sense information (Dillon, Izard, & Spelke, 2020). It is nevertheless unknown whether such information drives infants’ shape discrimination in the complex environments of everyday life.
Following a preregistered design and analysis plan, the current study adapted Experiments 1 and 2 of Dillon et al. (2020) to investigate whether infants discriminate shape information when 2D figures are presented in the experimenter-free and uncontrolled home environments afforded by the online platform Lookit (Scott, Chu, & Schulz, 2017; Scott & Schulz, 2017). Experiment 1 (N = 48) analyzed the proportion of 7-month-old infants’ looking to triangles changing in shape and area compared to triangles changing in area alone using a change-detection paradigm (Figure 1). Unlike the two experiments testing such shape discrimination in Dillon et al. (2020), infants on Lookit did not look significantly longer to figures changing in both shape and area (t(47) = .273, p = .786). Post hoc analyses (after Scott & Schulz, 2017) revealed that almost half of the infants tested (N = 23) experienced at least one distracting event during the test trials, which may have affected their performance: The non-distracted infants (N = 25) showed a trending positive preference for the shape changes (t(24) = 1.59, p = .125) while the distracted infants did not (t(22) = -1.10, p = .283).
To explore how distraction may have affected infants’ performance in Experiment 1, Experiment 2 significantly improved the instructions provided to parents and prespecified an exclusion criterion based on distraction. It also follows a sequential Bayes Factor Design Analysis (Stefan, Gronau, Schönbrodt, & Wagenmakers, 2019): After every 8 infants, a Bayesian t-test is conducted until sufficient evidence for H0 (BF < 1/6) or H1 (BF > 6) is found or until a prespecified maximum sample of infants (N = 48) is collected. The analysis of the current sample (N = 32) provides stronger evidence in favor of H0 (BF10 = .321), suggesting no difference in looking times to figures changing in both shape and area (M = .51, SD = .06) compared to area alone (M = .49, SD = .06).
The current results provide evidence that infants’ shape sensitivities revealed by lab studies may not drive shape discrimination in uncontrolled home environments. At home, not only distractions, but also variable lighting, sounds, and testing equipment, including different sized screens, may reduce infants’ interest, attention, and sensitivity to the basic properties of shapes compared to the lab. These differences, moreover, may change infants’ behavior (Saffran, Pollak, Seibel, & Shkolnik, 2007) and ultimately what types of knowledge they can acquire about object shapes and objects in everyday, complex environments.

Authors

• Agata Bochynska, New York University
  Presenting Author

• Kimberly Scott, Massachusetts Institute of Technology (MIT)
  Non-Presenting Author

• Moira R. Dillon, New York University
  Non-Presenting Author