Text analysis approach to measuring text social information in children’s picture books

ABSTRACT

Text social information includes the cognitive processes and social communication skills that support real or hypothetical human thought or interaction. The current measure of text social information is genre. However, genre is a limited measure because of poor operationalization, limited specificity, and overlap with structural and linguistic differences. The purpose of this study was to develop an automated text analysis approach to measure text social information in children’s picture books beyond genre. Studies 1 and 2 found convergent and divergent validity for several measures that captured text social information with these measures being significantly higher in children’s fiction compared to nonfiction books and not correlated with other structural and linguistic text measures. Study 3 found two components of text social information, based on a principal component analysis. These two components captured a general socialness factor (i.e., theory of mind, emotions, and social relationships) and pragmatics/conversation. Study 4 provides preliminary evidence for the predictive validity of the text social information measures. Together, this study provides an initial set of continuous measures for measuring text social information that can begin to advance the field in determining how text social information impacts comprehension and social cognition.

Acknowledgments

The author thanks Jessica Bradshaw, Christian McKenzie, Kaleigh Schreiber, Hannah Sevart, and Lily Shaw for transcribing the children’s books, assisting with LIWC and Coh-Metrix reports, and early development of the text social information measures. Earlier versions of this work were presented in part at the 2018, 2019, and 2022 Society for Text and Discourse Conferences.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability

Summary data, scripts, and supplementary materials related to this study are available from Davidson, M. M. (6 September 2023). Children’s Books Text Social Information. Retrieved from osf.io/n5yr7.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/0163853X.2023.2291297.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1. The more common term used in the literature is typically developing; however, this term is problematic because researchers often do not conduct full enough evaluations to know that children are actually typically developing (see Bottema-Beutel et al., 2021 for more details).

2. Autistic children (i.e., identity-first language) as opposed to children with autism (i.e., person-first language) is used here as it is preferred by the autism community (Bottema-Beutel et al., 2021; Kenny et al., 2016).

3. The Affect category also includes swear words, but these do not appear in children’s books and were, therefore, excluded from the definition for this study.

Additional information

Funding

This work was supported using startup funds from the University of Kansas. The author receives a salary from the University of Kansas.