Development of attention orienting to rapidly presented facial expressions from infancy to toddlerhood

• In Event: Attention and memory in infants and children

Abstract

Background: Prioritized processing of salient and biologically relevant information shapes functional specialization throughout early development. While little is known about the minimal amount of information needed to bias visual attention, previous work (Lasch, Thomas, & Elison, 2018) has established some visual attentional biases in 2-5-year-olds towards fear faces as compared to scrambled and neutral faces, but not inverted fear faces

Objective: The objective of these analyses was to elucidate the development of biases towards rapidly presented fear faces in the first years of life.

Methods: Two hundred and twelve infants completed an eye-tracking task at 1-7 visits from ages 1 to 60 months, a total of 837 visits. Data collection has closed. Of the collected data, 523 visits from 180 participants have undergone an initial preprocessing. Of those, 355 visits from 126 unique participants are ready to be analyzed, and 103 of those visits (from 40 participants) have already been imported to R for initial modeling (Figure 1). Analysis of the 355 files from 126 participants are extremely feasible before April 2021, as final data processing has recently been streamlined for remote completion.

In this variant of a spatial cuing task, competing cues are briefly presented (24 ms) within the parafoveal visual field and then backward masked (126 ms) before the presentation of a peripheral target. Facilitation and cost due to cue-target congruent/incongruent trial-types are captured via eye tracking (Tobii TX 300). Cue combinations included 1) a fearful face and a gaussian filtered ‘scrambled’ version of the fearful face; 2) a fearful face and the same fearful face inverted. (Figure 2).

Multi-level nested models (trials within visits within participants) will be run to predict response time to targets. Separate models will assess these effects in the scrambled and fearful conditions. Predictors in these models will include the main effects of participant age, target direction (left vs. right), fear cue location (left vs. right), and relevant interactions between these variables (most importantly the cue x target interaction that would indicate an attentional bias).

Expected Results: We expect to discover an age X cue X target interaction in the fear vs scrambled condition, such that the attentional bias, as indicated by the cue X target interaction, will increase with age. For this analysis, we anticipate a significant cue X target interaction as well, which will indicate that we can detect an attentional bias in our youngest participants. For the fear vs inverted fear condition, we anticipate an age X cue X target interaction as well, but no cue X target interaction. The results should indicate that an attentional bias is detectable only in older children, These findings promise to elucidate the nature of prioritized attentional selection early during the infant/toddler period.

Authors

• Carolyn Lasch, University of Minnesota - Twin Cities
  Presenting Author

• Jed T. Elison, University of Minnesota - Twin Cities
  Non-Presenting Author