Poster #105 - Infant Vocalizations Shape the Sensitivity of Infant Educators’ Language

• In Event: Poster Session 17

Abstract

Background: In infant-toddler classrooms, educators’ sensitive responses to infant cues create opportunities for infant learning (Girolametto & Weitzman, 2000). For example, educators simplify their speech when responding to infant vocalizations (Authors, under review), which may facilitate word-object associations. Here, we further investigated the characteristics of educator speech to understand how educators’ response profiles might differ when responding to infant vocalizations compared to talking at other times. We also consider how the characteristics of infant vocalizations impact the type of responses educators provide. Understanding the relative salience of vocal quality and infant gaze will add nuance to our understanding of how infants construct their learning environments in infant classrooms.

Method: We analyzed video recordings of 34 educator-infant dyads in a five-minute free play session with toys. The sample came from a larger dataset on toddler socio-emotional development (Woldarsky et al., 2019). Infants were 10-15 months at the time of recording (M=12.6 months). First, educators’ speech was transcribed and categorized as contingent if produced within 2 seconds of an infant’s vocalization or non-contingent (all other speech) (Authors, under review). In a second pass of the data, we classified educator utterances into one of six mutually exclusive categories: descriptions, questions, affirmations, imitations, play prompts, and explore prompts (Tamis-LeMonda et al., 2001). We also coded infant vocalizations for vocal quality (e.g., phonology: vowels or consonant-vowels; Oller, 2000) and infant gaze while vocalizing (e.g., object-directed, educator-directed, or undirected). We then compared the proportion of contingent and non-contingent utterances that received each response category and examined differences in contingent speech types by vocal quality and gaze types.

Results: The response profiles of educators’ utterances differed when responding contingently to infant vocalizations compared to speech at other times (Figure 1). When examining non-contingent speech, descriptions were the most likely form of educator speech (40%), followed by questions (15%) and play prompts (20%). In contrast, when responding contingently, educators were equally likely to respond with descriptions (30%) as affirmations (28%). This higher rate of affirmations suggests infants are more likely to receive less specific information following their vocalizations.
Infant gaze and the vocal quality of the vocalization impacted some aspects of educators’ contingent responses. For example, undirected vocalizations were more likely to elicit explore prompts (p=.003). Consonant-vowel vocalizations directed at objects were also significantly more likely to be imitated by educators’ than undirected vocalizations or those containing only vowels (p=.004; Figure 2).

Implications: Our results suggest that infant vocal cues organize contingent and sensitive responses from educators. Infants’ vocalizations elicited more imitative and affirmation responses from educators than non-contingent speech. While imitations may model closer word approximations for infants, affirmations provide limited information for learning. Understanding the specificity of educators’ contingent responses is important because we know that the information provided after a vocalization is privileged for infant learning. Consequently, if infants receive vague input following a vocalization, then opportunities for learning may be missed. Our results demonstrate that infants actively construct their learning environments by where they look and how they babble.

Authors

• Melissa Geary, Lebanon Valley College
  Presenting Author

• Rachel R Albert, Lebanon Valley College

• Claire D Vallotton, Michigan State University