The purpose of this research was to explore the effects of two factor experimental designs in word problem solving (WPS) assisted assessment for third graders. The WPS skill diagnostic guidance and learning strategies were used as independent variables, different tasks and transfer performances of WPS comprised dependent variables, and we partialled out the effects of some control variables to promote the analysis accuracy. The findings are as followed - the diagnostic, prompting and classified function of skill diagnostic guidance could operate well, and the main effect and partial interaction of facilitation were found for skills diagnostic guidance with the adopted multiple-form learning strategies. In addition, the performance of WPS could be correlated with and predict performance effectively, the performances of skill diagnostic guidance could also be correlated significantly with the performance following prompting facilitation for third graders.