Scientific literacy in Croatia is developed through the learning outcomes of the subject Science and Society in primary classroom teaching (grades 1–4, with one teacher for all subjects). In subject teaching (grades 5–8, where each subject has its own teacher), it continues to evolve within the subjects Science and Biology, ensuring systematic scientific literacy from an early age. Given the transition between these two stages, a study was conducted to analyze the development of students' scientific literacy after completing Science and Society in the first stage, that is, after entering subject teaching. The study included 949 fifth-grade students (the average age is 10.5). To assess scientific literacy, a written knowledge test was constructed, consisting of 13 questions of varying cognitive levels. The questions assessed different scientific literacy skills, including observation, classification, experimentation (use of instruments, identification of research questions, recognition of cause-and-effect relationships), data interpretation (predicting results, reading graphs and drawings), and scientific reasoning based on results across different biological concepts. The selected concepts were derived from the Science and Society curriculum and continue to be taught within the Science subject. The data were analyzed using descriptive statistics to determine the students' performance at different cognitive levels. A gender-based analysis was also conducted, and a t-test was used to determine whether there was a statistically significant difference between boys and girls in solving questions of different cognitive levels, particularly those involving graphical and tabular representations and drawings (visual questions). A one-way ANOVA with post hoc analysis, depending on the homogeneity of variance, was used to determine differences in solving questions of varying cognitive levels and visual questions based on the final grade in Science and Society. Students performed poorly on higher cognitive-level questions compared to lower-level ones. On first-level questions, out of a maximum of two points, students achieved an average of 1.7 points. In second-level tasks, out of a maximum of seven points, students achieved an average of 4.15 points. In third-level tasks, out of a maximum of 13.50 points, students achieved an average of 4.23 points. Comparing the mean scores across different cognitive levels, no statistically significant difference was observed between boys and girls. However, in visual second-level questions, boys achieved statistically significantly better results than girls (boys: 2.14 points on average, girls: 1.99 points; t (947) = 2.511; p = 0.012). An analysis of results based on final grades showed that students with different final grades differed statistically significantly in solving questions of various cognitive levels (F(3, 945) = 90.3, p < 0.01). Differences were also found in solving visual questions (F(3, 945) = 84.45; p < 0.01). An analysis of incorrect answers across all categories based on total test scores revealed that students, regardless of their achievement level, struggle with understanding the principles of animal classification, the influence of abiotic factors on plant growth, and the application of knowledge in tasks involving tables, graphs, and result-based predictions. The written test results highlight the need to improve teaching practices for developing scientific literacy.

Keywords: Cognitive levels, visual representations, fifth-grade students, scientific reasoning, educational outcomes.