Poster #110 - Mother-Child Reminiscing Across Levels of Abstraction and Preschoolers' Language

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Language, Communication

Integrative Statement

Researchers studying parent-child reminiscing in the preschool years have often focused on elaborative talk (i.e., provision of unique details). We propose a novel conceptualization of parent-child reminiscing by examining four levels of abstraction (LoA; i.e., a continuum of literal to inferential language) used in parent-child book reading studies (Hammett, van Kleeck, & Huberty, 2003) and utterance function (i.e., wh-questions, yes/no questions, statements) given the relevance of LoA and utterance function in the book reading context (Collins, 2016). The purpose of this study was to examine how mothers’ LoA and utterance function 1) elicited children’s immediate responses within the same LoA and 2) predicted children’s story comprehension. Story comprehension was targeted because of the relation between mothers’ reminiscing and story comprehension (Reese, Leyva, Sparks, & Grolnick, 2010) and the importance of narrative input for preschoolers’ oral language skills (Rowe, 2012). We also compared LoA and elaborations coding.

Fifty preschoolers (Mage = 4.42 years) participated twice six months apart with mothers. Mothers discussed three past events with children at each time. Narratives were coded for elaborations and LoA (see Table 1 for LoA coding). Children’s vocabulary was assessed with the PPVT (Dunn & Dunn, 2007). Story comprehension was assessed at each time with 10 questions, including five literal (e.g., characters) and five inferential (e.g., causal inference; Paris & Paris, 2003).

Regarding the first aim, we examined the contingent relations between mothers’ input and children’s immediate responses using sequential analysis (Bakeman & Quera, 1995), which determines whether certain adult utterances and child responses occur more frequently than expected by chance. Because some contingencies lacked normality, levels 1 and 2 were collapsed into a literal category and levels 3 and 4 were collapsed into an inferential category. Contingencies within a LoA were significant for all utterance types (e.g., when mothers’ utterances were inferential, children’s subsequent utterances tended to be inferential, not literal). Relations were stronger for mothers’ questions and weaker for statements; mothers’ wh-questions are presented as an example. The Yule’s Q values are presented in Figure 1, which are interpreted like a correlation coefficient (e.g., > ±.70 is very strong; Bernard, 2000). Contingency patterns were remarkably similar across time for all utterance functions.

Regarding the second aim, controlling for child age and vocabulary, children’s story comprehension was significantly correlated with mothers’ yes/no questions at the lowest LoA at Time 1 (r = .30, p < .05) and mothers’ statements at the highest LoA at Time 2 (r = .30, p < .05). Finally, mothers’ LoA and elaborative discourse did not vary systematically, suggesting that these are distinct conceptualizations.

Although the sample size was modest and duration brief, this study suggests that parent-scaffolded reminiscing (in addition to book reading) can encourage children’s inferential talk. It also suggests that for older preschoolers, providing inferential statements may facilitate children’s story comprehension or that children with greater story comprehension have mothers who use more inferential talk (e.g., provide explanations) with children. Future research with larger samples, longer durations, and experimental designs are needed to disentangle these relations.

Author

• Virginia Tompkins, The Ohio State University
  Presenting Author