Nowadays, the human-computer interaction is being a hot topic. In such a way, several methods have been proposed to deal with multi-granularity when people with different knowledge level express their preferences on the same concept using linguistic notation, that is, words instead of numbers. This is a critical problem in group decision making scenarios, but all the existing approaches have their own advantages and drawbacks. Therefore, some work better in certain environments than others. In such a way, choosing the best method in each situation is critical for obtaining good quality results. In this contribution, an analysis on the different fuzzy linguistic multi-granular modelling approaches is presented in order to provide the reader some advice of what method should be chosen depending on the problem and the quality of results that the user expects to obtain.