Poster #58 - The Role of Emotional Variability in Storybooks for Young Children’s Word Learning

• In Event: Poster Session 13

Abstract

Emotional variability is common in daily experience. We often use emotional information to make predictions about others’ likely behaviors and to understand others’ goals, but it remains unclear whether the plethora of emotional information present in daily life can influence other types of learning as well. Language is one of the most important things that young children learn. Further, prior research has shown that variability in physical aspects of learning environments (e.g., the room someone is in or the cloth behind an object) can impact word learning. However, it is less clear whether variability in more social types of contexts can similarly influence word learning. Given the variability in emotion in daily experience, we sought to examine whether variability in emotional context influences children’s word learning.

Ninety-six 4-year-olds (Mage=4.46 years, SDage=0.34 years, 48 female) participated in the present study, which was conducted entirely over Zoom. Children learned eight novel words presented throughout a virtual storybook. Each novel word was presented on four separate pages, and each page involved either a happy, sad, angry, or afraid emotional context (see Figure 1 for example page). Children were randomly assigned to either the no variability (each word presented in only one type of emotional context), low variability (each word presented in two emotional contexts), or high variability (each word presented in all four emotional contexts) condition. Importantly, all children heard all four emotional contexts throughout the story. Only the pairing of a specific word to emotional context varied by condition. After the story, children participated in eight generalization test trials, one for each novel word. Each test trial included four choices, and children were asked to identify which object matched the word.

A between-subjects ANCOVA was conducted to compare the number of correct post-test trials (out of 8 possible) across the three study conditions while covarying for child age. Results revealed a statistically significant difference in post-test performance across the three emotional variability conditions (F(2,92)=4.14, p=.019), with the average number of correct responses per condition as follows (and as depicted in Figure 2): No variability=3.59 (SD=1.90); Low Variability=2.97 (SD=1.56); High variability=4.09 (SD=1.65). We followed up with post-hoc pairwise comparisons using a Bonferroni correction, which revealed that children had a higher number of correct responses in the High Variability condition compared to the Low Variability condition (t(92)=-2.83, corrected p=.017), but that the No Variability condition did not significantly differ from either the Low Variability (t(92)=1.45, corrected p=.452) or High Variability condition (t(92)=0.17, corrected p=.516).

Taken together, these results suggest that variability in emotional context can influence word learning. In particular, we found that high emotional variability led to the best performance in the storybook context, significantly better than low variability. This may be because high emotional variability provided substantial support for decontextualization, while low variability did not provide sufficient support for either aggregation or decontextualization in this learning context. Overall, these results suggest further important links between emotion and cognition (specifically, word learning) in early development.

Authors

• Marissa Ogren, Rutgers University - Newark
  Presenting Author

• Michelle Lynn Luna, University of Notre Dame

• Catherine Sandhofer, University of California - Los Angeles