This study analyses the impact of an instructional intervention based on Scratch for teaching statistical knowledge in Year 5 of Primary Education, with the aim of improving the learning of statistics (traditionally less addressed in the classroom) and promoting more active methodologies in the area of mathematics. To this end, a quasi-experimental design with a pre-test–post-test and a control group was conducted, with a sample of 47 students from a school in the Autonomous Community of Extremadura. The experimental group (n = 23) worked on the contents through Scratch-based activities, while the control group (n = 24) followed a traditional methodology supported by the textbook. The intervention addressed data collection and organisation, graphical representation, and the calculation of measures of central tendency, integrating elements of computational thinking. Likert-type questionnaires and a written test common to both groups were used for data collection. Data analysis was carried out using inferential statistical tests. The results show that students in the experimental group achieved significantly higher academic performance in learning statistical knowledge. In addition, students who used Scratch produced higher-quality definitions of fundamental statistical concepts and rated the tool as more usable and understandable than the textbook. Therefore, this study confirms Scratch as an effective instructional resource for teaching statistics in Primary Education, as it enhances academic performance, understanding of basic statistical knowledge, and a positive perception of the learning experience.