This chapter can be summarized in the following three comments: First, no method is perfect. It is crucial to know what can be achieved with a particular technique or method of study; it is important to anticipate the benefits and limitations of a particular methodology as one embarks on a particular study; and it is completely unreasonable to expect that any given method could uncover all properties of the phenomenon that you want to study. Second, since heritage speakers are often reluctant to produce language, it is important to combine data on their production with data on their comprehension. It is helpful to think of (controlled) production data as the first pass, and comprehension data as the guide to a fuller picture. Third, whatever methodologies work well for a language study should also work for heritage language study. The chaoter discusses main approaches in the study of production and comprehension by heritage speakers and emphasizs the importance of comprehenson in the modeling of heritage grammatical systems.