Poster #28 - Moms matter in word-learning: The impact of maternal and others' CDS on novel word-learning

• In Event: Poster Session 16

Abstract

Infants who hear more child-directed speech (CDS) tend to have better language learning trajectories; however, the majority of the studies examining language input on infants' language development have only examined input produced by infants' mothers (e.g. Huttenlocher et al., 1991; 2010; Pan et al., 2005). For 2-year-olds the amount of CDS from all individuals in their environment is related to their later vocabulary development (Shneidman et al., 2013). However, it is unknown whether 12-month-olds also benefit from CDS spoken all individuals in their environment, or whether instead speech from their mother is most relevant. Two-year-olds are better at processing speech (Fernald et al., 2006), and have better memory and attention capabilities (Hayne et al., 1997; Rose et al., 2009) than 12-month-olds, which may allow them to better aggregate input from multiple speakers across time. Thus, it is likely that the features important for 12-month-olds differ from those that are important for two-year-olds. We tested whether all language input is important for infants' word learning and to what extent it matters whether the input is coming from the infants' mother or other individuals in their environment.

We collected recordings of language input in 12-month-olds' (N=50, Males=26) environment and coded how much CDS and overheard speech they heard from their mother and from other individuals. At 18-months infants completed an in lab novel word learning task.

To test how characteristics on infants' language environment were related to their novel word learning scores, a hierarchical regression was performed. Participant sex was entered in the first step and was not significant, p > .05 . At step two CDS from the mother, both tokens and Type-Token Ratio (TTR), were entered. Maternal input significantly increased model fit, ΔR2 = .16, ΔF (2, 42) = 4.08, p < .05. Mothers' CDS tokens significantly predicted infants' novel word learning, ß = .60, p < .05, but not the amount of TTR, ß = .41, p > .05. In the third step CDS from other individuals, both tokens and TTR, were entered, these factors did not significantly predict novel word learning or improve model fit, p > .05. In the final step, overheard speech, both tokens and TTR, were added to the model. These factors did not significantly predict novel word learning or improve model fit, p > .05.

We found that the number of CDS tokens from infants' mothers was related to infants' novel word learning, but not the number of CDS tokens from the other individuals in the infants' environment or the number of overheard speech tokens. One reason why mothers' CDS tokens was related to infants' novel word learning could be that mothers provide infants with more similar learning events of word-referent mappings, which may be especially important for the early stages of word learning. Overall, our findings provide further support for the role of language input on infants' word learning and suggest that input from the individuals in infants' environments differently impact their language development.

Authors

• Abbie Thompson, Valparaiso University
  Presenting Author

• Jill Lany