This study adopted two formats of a two-tier test to diagnose sixth-graders’ alternative conceptions in electricity. One was a static representation, and the other a dynamic representation. The major purpose of this study are to: (1) explore students’ alternative conceptions or knowledge deficiencies; (2) investigate the impacts of dynamic representation to their answering behavior. Sixty-two students were matched by the score on an achievement test in electricity, and divided into two groups for the representation format. Each participant was individually interviewed using the two-tier test. A confidence rating was also used to indicate the strength of a student’s conceptual understanding. This study identified six typical responses, four being robust alternative conceptions in high confidence. In addition, answering behavior might be changed by a dynamic representation. The confidence of the lower score students is reinforced. It is also shown in the interview data analysis that lower score students might figure out the context of items with the assistance of dynamic representation. In contrast, that higher score students are confused with an inconsistent hint while it is different from their internal representations. AS well as the higher score students could obtain more hints from the inconsistency between their internal and external representations while answering. Finally, implications for the design of a two-tier test were discussed.