The purpose of this study was to determine an effective method of teaching measurement before experimental manipulation. To achieve this purpose, during a lesson on “Burning Substances in Air” in a 6th grade science class, 152 pupils were divided into three experimental groups. Different teaching methods were used to teach each of these groups. Empirical verification indicated that the teaching method that prompts pupils to discuss experimental design and instructions afterwards is more effective in improving both skill and cognition, which are the two aspects of understanding. Moreover, it became clear that while using the teaching method that merely involves instruction, pupils’ understanding tends to be shallower. These results clearly favor organizing activities to enable pupils to design experiments proactively in school science classes.