Academic libraries serving teacher education carry a dual mandate: supplying current readers while training future selectors of children's books. This study asks whether a collection can appear balanced by broad category yet remain narrow on the dimensions that matter most for that second mandate. This research evaluated 60 books suitable for upper elementary readers (grades 4 to 6) from Universitas Pendidikan Indonesia's library collection of 257 children's titles, using a children's literature framework to analyze genre, theme, character, setting, and illustration. The collection is weighted toward nonfiction (38.3%) and away from traditional literature (1.7%), and its single traditional literature title is a German rather than an Indonesian folktale, meaning Indonesian oral tradition is effectively absent from a category meant to carry it. Character representation shows a similar asymmetry: adult and religious figures are nearly absent (1.7% and 3.3%), and the only explicitly gender marked character category is Princess (6.7%), with no equivalent male-marked category. Illustration quality varies sharply by publication origin: locally published titles show 28.6% good-quality illustrations against 71.4% poor, compared to 97.8% good among internationally published titles, a gap that makes quality filtering and support for domestic publishing genuinely conflicting acquisition goals rather than complementary ones. These patterns indicate that genre and format count alone are insufficient evaluation criteria for teacher-education library collections; cultural specificity, authority modeling, gender symmetry, and publication-origin parity require explicit tracking. This study offers a replicable method for surfacing such imbalances and provides baseline evidence for collection development in Indonesian academic libraries.
Copyrights © 2026