Poster #99 - Intersensory Processing of Social Events Mediates Relations Among Maternal Well-Being, Language Input, and Child Language

• In Event: Poster Session 14

Abstract

Maternal well-being (MBWB) is a multi-faceted construct, encompassing psychological (e.g., depression, anxiety, self-acceptance, and satisfaction) and social processes (e.g., family and community support; Zafar et al., 2014), with implications for child development. A sizeable literature has demonstrated that lower levels of maternal depression, greater maternal sensitivity (emotionally positive, prompt, and contingent responsiveness), and greater maternal language input (MLI) predict better child language outcomes (e.g., Albright & Tamis-Lemonda, 2002; Hart & Risely, 1995; Sohr-Preston & Scaramella, 2006). Recently, we demonstrated that better child language outcomes are also predicted by infant and child attention to audiovisual speech and intersensory processing (IP) of faces and voices (Bahrick et al., 2018; Edgar et al., 2022). The present study thus assessed relations among child IP of faces and voices, MWB, MLI, and, in particular, whether IP of faces and voices would mediate relations among MWB, MLI, and child language.

Children (N=97; 50 F) participated at 36 months of age (M=36.77 months, SD=0.93) as a part of a larger, ongoing longitudinal study. IP of Faces and Voices was assessed using the Multisensory Attention Assessment Protocol, a new individual difference measure of attention to audiovisual events (Bahrick et al., 2018). Trials consisted of a 3s central event (morphing geometric shapes) followed by two 12s lateral social events (women speaking), one of which was synchronous with its natural soundtrack (see Figure 1). IP was calculated as the proportion of total looking time to the sound-synchronous event (face-voice matching) on each trial and averaged across trials. Diversity of MLI was calculated as the number of unique words mothers spoke to their child during an 8-minute lab-based play session. MWB was assessed by having mothers complete the Ryff Scales of Psychological Well-being (Ryff & Keyes, 1995), which measures six aspects of well-being, including self-acceptance and purpose in life. Child Expressive Language was assessed using the Expressive Vocabulary Test (Williams, 2007). Maternal education served as our index of Socioeconomic Status (SES) and was used as a covariate.

We created a latent variable of MWB using a confirmatory factor analysis of the six Ryff sub-scales. Preliminary analyses revealed correlations among IP, MLI, child expressive language, and SES, ps<.01. Also, the MWB latent factor was correlated with IP and child expressive language, ps<.01. To address our main research question, we tested a mediation model using SEM (MWB & MLI → IP → language) and found that this model had very good fit to the data (see Figure 2). Greater MWB and more MLI predicted greater child IP of faces and voices, ps<.05, which in turn predicted greater child expressive language, p<.01, holding SES constant.

Findings reveal new developmental pathways to child language outcomes, in which maternal predictors (well-being, language input) and child predictors (face-voice matching) contribute to language outcomes. Further, the present findings highlight the important, mediational role of child face-voice matching: maternal variables contributed to child language outcomes through face-voice matching. Findings build upon our previous work emphasizing the importance of early face-voice matching as a gateway to language outcomes.

Authors

• James Torrence Todd, Florida International University
  Presenting Author

• Bethany K Ramirez, Florida International University

• Kaitlyn Testa, Florida International University

• Elizabeth V. Edgar, Yale University

• Lorraine E. Bahrick, Florida International University