Detecting joint attention events in mother-infant dyads: Sharing looks cannot be reliably identified by naïve

Kirsty E Graham1,2, Joanna C Buryn-Weitzel1, Nicole J Lahiff1

  • 1Department of Psychology, University of York, York, United Kingdom.

Plos One
|July 23, 2021
PubMed

Insights

Measuring joint attention in infants is challenging. Naïve observers could not reliably distinguish infant looks, suggesting subjective gaze analysis is insufficient for assessing shared attention.

Area of Science:

  • Developmental psychology
  • Cognitive science
  • Infant behavior

Background:

  • Joint attention, crucial for language development, lacks a standardized measurement in pre-linguistic infants.
  • Existing definitions vary, with some requiring coordinated gaze and communicative signals, while others propose gaze quality alone indicates shared attention.

Purpose of the Study:

  • To empirically test the reliability of subjective gaze categorization for assessing joint attention in infants, as proposed by Hobson and Hobson (2007).
  • To determine if naïve observers can reliably distinguish infant "sharing", "checking", and "orienting" looks.

Main Methods:

  • 32 naïve raters categorized 60 infant looks to their mothers based on Hobson and Hobson's definitions.
  • Inter-rater reliability was assessed, and agreement with the mother's judgment was analyzed.

Main Results:

  • Overall low agreement among raters was observed.
  • In most cases, a majority of raters did not agree with the mother's perception of the infant's look.
  • Look duration and maternal communication rate may have influenced raters' judgments.

Conclusions:

  • Subjective judgments of infant gaze types are unreliable for identifying mutual awareness in joint attention.
  • Objective behavioral measurements are recommended for accurately inferring shared attention in infants.
  • This approach is vital for understanding the origins of joint attention.