In this research, a new methodology for the conceptual evaluation of students of a Massive Open Online Course (MOOC) is addressed. The electronic circuits MOOC course "Bases de Circuitos y Electrónica Práctica" (Fundamentals of Circuits and Practical Electronics) was used as a model. The conceptual evaluation of the course was measured by the semantic proximity between ten relevant concepts of the course. This semantic structure involves conversion and reduction processes through Pathfinder associative networks and the minimum spanning tree. The gain between the start and the end of the course is calculated by statistical tests considering the similarities between the students' networks and the teacher's network, as a reference network. The results show significant gains, particularly when evaluating courses with badges. We propose this evaluative model as a recommended generic model that can be integrated into MOOCs due to its immediate and online response.