Science Books: Does Genre Matter when Teaching Kindergarteners Physical Science Concepts?

• In Event: Learning by Reading: Fostering Children’s Science Learning Through Storybooks

Abstract

Science books are effective tools for early science teaching because they can provide information that is not accessible through direct observation (Pringle & Lamme, 2005). Although both fiction and non-fiction books are used in primary science classrooms (Donovan & Smolkin, 2002), narratives are usually favoured by teachers because of the assumption that children understand stories before expository texts (Pappas, 1993; Duke & Kays, 1998; Donovan & Smolkin, 2001; Duke & Tower, 2004). Previous research examining the effect of genre on science learning has typically used trade books and were therefore unable to control for confounds such as length and content (Pappas, 1993; Duke & Kay, 1998; Mantzicopoulos & Patrick, 2010). Few studies have designed books to investigate learning from genres (Duke & Bilman, 2009). Only two studies have created comparable texts, but their findings were inconsistent: elementary students learned more from informational books (Cervetti et al., 2009), while middle school students learned more from narratives (Arya & Maul, 2012).

In three studies, we assessed learning from either realistic fiction or explanatory non-fiction books that progressed from highly matched genres to more distinct representations. A pre-post-test design measured children’s understanding of a science concept. All analyses controlled for receptive language and socioeconomic status.

Study 1 (N=156), addressed the belief that heavier objects fall faster than lighter ones. The books had identical illustrations, as well as similar readability and length. Four- and five-year-olds predicted what would happen if two objects were dropped at the same time. A mixed analysis of covariance (ANCOVA) controlling for age revealed a significant interaction between test phase and genre, F(2,150)=9.85, p<.001, ηp2=.12. Post-hoc tests indicated that compared to pretest, children had more accurate predictions at posttest for the realistic fiction and explanatory conditions (ps<.001), but not the control condition (p=.12).

Study 2 (N=96) addressed the belief that heavy objects always sink, and light objects float. Genre-specific aspects like the use of characters in the realistic fiction and headings in the explanatory book were differentiated. Five-year-olds identified which object in a pair would sink and which would float and then explained their predictions. A mixed ANCOVA revealed an effect of test phase, F(1,92)=0.25, p=.035, ηp2=.05. Post-hoc tests indicated that compared to pretest, children had more accurate explanations at posttest for both genres (ps<.001).

Study 3 (N=82) addressed the belief that objects balance in the geometric center. These books differentiated the types of images: illustrations versus photos. Five-year-olds placed pairs of objects on a beam so it would balance and explained their placements. A mixed ANCOVA revealed a significant interaction between test phase and genre, F(2,77)=7.17, p=.001, ηp2=.16. Post-hoc tests indicated that compared to pretest, children had more accurate explanations at posttest for the realistic fiction and explanatory conditions (ps<.001), but not the control condition (p=.43).

Together, these findings show that kindergarteners learn physical science concepts equally from realistic fiction and explanatory books.

Authors

• Vaunam Venkadasalam
  Presenting Author

• Nicole Larsen, University of Toronto
  Non-Presenting Author

• Patricia A Ganea, University of Toronto
  Non-Presenting Author