The development of some new statistical disclosure control (SDC) methods for tables in recent years provides statistical agencies with additional ways to protect the confidentiality of their data. However, it also means that these agencies need to make choices about which method should be applied to a given table or set of (possibly linked) tables. In this paper, we analyze several important SDC methods with respect to a set of important factors that an agency needs to consider in choosing a method. The dominant factor in this decision process should be the nature of the confidentiality pledge that an agency is required to enforce. Another important factor is the development time and runtime of the software associated with each method. Another factor, one which has received too little attention, is the way in which the tables will be used by table users. Of course, this factor is difficult to analyze since different users use the tables in different ways; e.g., some users simply want to look up a few cell values; others are interested in sophisticated statistical modeling. In this paper we propose some statistical methods for analyzing SDC methods with respect to these factors. Hopefully, this analysis will help agencies select the best SDC method to use with a given set of tables.