Effects of fantastical and anthropomorphic features in children's picture books on young children's learning and transfer: Systematic review and three-level meta-analysis protocol
Year: 2026
Type: preprint
Study design: Study Protocol
Abstract: Picture books are among the most common cultural tools through which youngchildren acquire knowledge about the world, yet most commercially availablepicture books contain elements that violate the rules of the real world:talking animals, characters who dress and behave like humans, and events inwhich physical laws are suspended. This raises what the literature calls theyoung reader's dilemma: how does a child decide which parts of the informationencountered in a fictional world can be carried into the real one? Experimental work on this question has grown rapidly since 2008, but thefindings have not converged. Some studies report that anthropomorphic picturebooks reduce factual learning about animals and increase the attribution ofhuman characteristics to real animals; others find that anthropomorphicillustration improves the quality of biological explanation, or thatanthropomorphic language does no harm to factual learning. The picture isequally mixed for transfer outcomes. All existing syntheses in this area are narrative reviews; a preliminarysearch located no meta-analysis of this question. This protocol describes a systematic review and three-level meta-analysis ofthe effects of fantastical and anthropomorphic features in children's picturebooks on the learning and real-world transfer of children aged 36 to 108months. The review models four candidate sources of the conflicting findings:the type of fantasticality manipulated, the outcome domain, the age of thechild, and whether the fantastical element is presented before, simultaneouslywith, or after the target content. Effect sizes are synthesised with a three-level random-effects model(metafor::rma.mv) with cluster-robust variance estimation (CR2, Satterthwaite)at the sample level. Effect sizes are never averaged within studies. Theprimary analysis is conducted separately by outcome domain rather than as asingle pooled estimate. Reporting follows PRISMA 2020 and PRISMA-P 2015.Registration is on OSF Registries using the Generalized Systematic ReviewRegistration template. --- Correction, 23 September 2026: versions 1 and 2 of this record carried, inerror, the abstract of a different protocol (narrative-based early mathematicsinstruction). The description above replaces it. No methodological decision isaffected. Version 1 cannot be edited and remains accessible with the incorrectdescription.
Authors: KELEŞ, OĞUZ, KARAGÜL, SEDAT
Language: Turkish
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7204873022