Counting nodes is easy to grade and easy to game. Proposition quality is harder and more useful.
Ask a teacher to rank two concept maps and the instinct is to count. More nodes, more links, more visible work. The study behind this article suggests that instinct is unreliable.
The team compared two extension concept mapping methods with 55 students. One let students build freely. The other had them recompose a teacher's map from a fixed kit of components, then extend it. The second produced higher quality propositions even though students wrote fewer of them.
A map can be large and still be wrong. Propositions can be scientifically meaningless, or can link two ideas that do not belong together. Counting treats every statement as equal, and the authors argue that the quality of the interrelationships between concepts is what carries the learning signal.
There is a practical consequence. If size is the metric in a rubric, students learn to pad the map, and the padding looks like progress right up until they are asked to explain it.
In the kit-build condition, students first rebuild the teacher's map from provided parts. That step forces attention onto how an expert connects ideas, before the student adds anything of their own. It is a rehearsal of structure, not just a task.
The perceptions data supports that reading. Students rated the kit-build tool higher on ease of use and usefulness, and reported a stronger intention to keep using it. The authors single out the recomposition step as the part that makes the activity feel workable.
Grade the quality of the propositions, not the number. Give students an expert structure to rebuild before asking them to extend anything. And expect a good map to look smaller than a free-form attempt, because a clear structure does not need filler.
Prasetya et al. (2022). Analysis of quality of knowledge structure and students' perceptions in extension concept mapping. Research and Practice in Technology Enhanced Learning, 17(14).
