Detecting Toddler Emotion Regulation from Sequential Patterns of Everyday Emotional and Verbal Communication

• In Event: Leveraging Wearable Sensors to Access Processes of Children’s Everyday Behavioral and Physiological Regulation

Abstract

Emotion regulation is a crucial skill that emerges largely in the context of early parent-child interactions. Yet traditional methods often measure child emotion regulation using brief, structured laboratory tasks, which differ greatly from everyday life and thus may limit ecological validity (Buhler-Wassmann & Hibel, 2021). These tasks may not tax self-regulatory capacities as everyday stressors do, may constrain variability in child behaviors, and often require time-intensive coding.

We use Language Environment Analysis (LENA), a wearable audio-recording technology, to address these challenges by sampling toddlers’ emotion and verbal communication in everyday life. LENA automatically generates a daylong, continuous record of parent-toddler communication, including toddler (pre)-verbal vocalizations (VOC), parent speech, and¬— a less utilized feature– toddler cries (CRY). Previously, we identified communicative sequences that might reflect maternal responsiveness to distress (CRYMOM) and co-regulation of toddler negative emotion (CRYMOMVOC). These sequences related meaningfully to maternal sensitivity to distress and toddler characteristics (BLINDED FOR REVIEW). Extending this work, we identify communicative sequences that might reflect a) toddler self-regulation of emotion (CRYVOC) or b) toddler breakdown in self-regulation (VOCCRY). We hypothesized that the sequence reflecting toddler self-regulation (CRYVOC) would be positively associated with factors known to support emotion regulation development (maternal sensitivity to distress, toddler effortful control, toddler language ability) whereas the reverse would be true for the sequence reflecting self-regulation breakdown (VOCCRY).

Twenty-five one-year-olds (Mage=16.60 months, 92% White) wore a LENA recorder for a full day. They participated in semi-structured parent-child interaction tasks, from which maternal sensitivity to distress was coded (NICHD ECCRN, 1999; ICC = .81). Mothers reported on toddler language ability (MCDI; Fenson et al., 2007) and temperament (ECBQ; Putnam et al., 2006).

Controlling for toddler age, multiple regression analyses showed that toddler characteristics and maternal sensitivity to distress conjointly contribute to the conditional probability of the target sequences occurring (Table 1). Toddlers with higher expressive vocabulary were more likely to shift from crying to (pre)-verbal vocalization (CRYVOC). This was also true for toddlers with high effortful control, but only if they had highly sensitive mothers (Figure 1a). In the context of low or average maternal sensitivity to distress, higher effortful control predicted lower probability of toddler regulatory breakdown (VOCCRY), whereas children with highly sensitive mothers exhibited low likelihood of regulatory breakdown, regardless of effortful control (Figure 1b). Thus, maternal sensitivity to distress appears to harness toddlers’ effortful control in service of emotion regulation, and buffer against regulatory breakdowns for toddlers with lower effortful control. Interestingly, although expressive language and effortful control are positively correlated in our sample and the broader literature (Guedes & Cadima, 2022), toddler language functioned differently; for toddlers with less sensitive mothers, higher expressive vocabulary predicted higher likelihood of regulatory breakdown.

Thus, the sequential patterns identified by LENA hypothesized to reflect toddler self-regulation (or self-regulatory breakdown) show meaningful associations with independent measures of toddler characteristics and maternal sensitivity. This highlights the promise of using ambulatory, automated processing technology to capture indicators of dynamic self-regulatory processes in daily life. We discuss the implications and potential applications of this approach.

Authors

• Margaret A. Fields-Olivieri, University of North Carolina at Greensboro
  Presenting Author

• Kristin Jennings, University of North Carolina at Greensboro

• Crystal Thinzar, University of North Carolina at Greensboro