Emergency Management Systems (EMS) are vital to communicating information between agencies, operations centers, and field personnel before, during, and after a disaster. However, when Emergency Management personnel sit down with software developers to gather requirements to create this bespoke software, they often overlook factors that would make the software more user-friendly and intuitive. Insufficient involvement of end-users in the design process and usability tests exacerbates these problems. Recent studies on user experience have identified recurring usability problems. Unlike past research, this study is a meta-analysis of usability methodologies using machine learning to determine the most common analytical approaches. The research compiles anonymized resource materials from published articles from the last five years, and uses machine learning to look for keywords, or tokens, that are common throughout the work. To determine what the most common research methodology is employed, the authors created four tasks. The first task creates a Python script to search for a curated set of tokens, the second task uses the Gensim library to produce a token list, and the third task uses NVivo to create the list. The authors review the results of these tasks and use them to adjust the last task. The results reveal that the most used research methodology in recent years involves having conversations with users, while field research to directly observe user behavior not frequently employed. The study concludes with suggestion for improvements.