Improving coding literacy among children is becoming increasingly essential in today’s world. By learning coding, children are more likely to grow different skill sets, such as computational thinking, critical thinking, authentic thinking and creativity whereby they learn to understand a problem and leverage their logic and knowledge to form a solution.

This research intends to model personalized learning paths for each child to learn coding at his or her pace, whereby the pace may differ for each child. The unique learning paths could help sustain the child’s learning interest and improve their performance. In order to achieve this goal, the Knowledge Space Theory (KST) concept is incorporated in the study to define knowledge states based on one or a sequence of Scratch code blocks. Subsequently, a set of coding challenges are tailored to evaluate each of the knowledge states delineated. In order to ensure the efficiency of the challenges, the challenge questions and solutions were designed with and reviewed by domain experts in block-based coding. Also, a web application system was developed for the automatic scoring of coding challenges.

The study emphasizes Penang public primary school children aged between 10 to 12. An entry test was conducted to identify their competency levels in coding. The children were then distributed accordingly into groups for the learning path experiment on fixed learning and personalized learning. Upon completion of the experiment, the children were required to complete an exit test in order for their performance to be monitored from the competency scores. At the final stage, a knowledge graph for learning coding alongside methods to evaluate children learning using personalized learning paths is developed. These learning paths could act as teaching modules to enhance children's learning experience and performance, in terms of block-based coding knowledge using the Scratch platform.