In this chapter, we set out to illustrate the theoretical and methodological distinctions used in Geluykens (this volume) by applying them to one particular type of speech acts, viz. directives. In directive speech acts, the speaker linguistically indexes their desire for an addressee to carry out a future action that the addressee would not have carried out without direction (cf. e.g. Searle, 1969, 1976). In our review, we include studies that have empirically investigated any directive speech act in spoken language. We provide an overview of the research on directives, categorize this research database as to whether they have focused on production or perception, which part of the directive sequence was investigated, which data type was employed (elicited vs. non-elicited vs. mixed), and which particular (pragmatic) subdiscipline or framework was applied. We furthermore provide a meta-analysis of directives research and outline the correlational patterns found between methodological choices and frameworks applied. Our meta-analysis shows that there is no clear preference for a specific data type overall; the data type correlates with the specific framework applied. While contrastive research designs make the use of elicited data more likely, there are still a number of studies which achieve variable control while using non-elicited data.