Immediate scores flatter almost everyone. Retention is where teaching methods separate.
An immediate post-test is a generous instrument. Students have just finished the activity, the material is still warm in memory, and almost any method looks effective. A delayed test does not have that luxury. It asks what actually survived.
That is why the design of this study is worth noting. Both groups were tested right after the mapping activity and then again 14 days later, with the same questions shuffled into a different order. The shuffle matters. It stops students who memorized the sequence from looking better than students who understood the content.
Neither group held its immediate score. The control group fell from 81.02 to 69.91. The experimental group fell from 87.50 to 78.57. Both declines were statistically significant, which is the ordinary shape of forgetting.
The interesting part is not that scores fell. It is that the ranking did not change. Whatever the kit-build group gained during the activity, they kept relative to the other group. That is a stronger claim than a single post-test can support.
If you are choosing a learning activity, an immediate quiz will rarely help you choose well, because most activities pass it. A delayed quiz costs one extra session and tells you whether the activity built something durable. The gains here also concentrated on content students had to reconstruct rather than read.
The honest caveat is scale. Fifty-five students in one course is a single data point, and the paper notes that retention has been under examined for extension mapping. This is a good argument for running delayed tests more often, not a final answer.
Prasetya et al. (2022). Investigating the memory retention in extension concept mapping. Improving Assessment and Evaluation Strategies on Online Learning.
