The majority of statistical software are disseminated as packages, procedures or functions that allow users to implement specific statistical methods in "script" mode without graphical user interface support. Nonetheless, such "script" mode software may not be satisfying for systematic analysis that requires intensive software-user interactions. For example, it is well recognized that missing data analysis should focus on positing varying missing data mechanism assumptions and evaluate the robustness of the inference results. Thus it becomes critical for any missing data analysis software to be able to efficiently explore a wide range of missing data assumptions and visualize and summarize the sensitivity analysis results. Such requirements are often challenging to be implemented in "script" mode. In this talk, we will illustrate how to use web-based applications with graphical user interface to conduct missing data analysis in a user-friendly manner.