This essay defines and critiques "methodocentrism", the belief that predetermined research methods are the determining factor in the validity and importance of educational research. By examining research in science studies and posthumanism, the authors explain how this methodocentrism disenables research from taking account of problems and non-human actants that are presumed to be of no importance or value in existing social science research methodologies, both qualitative and quantitative. Building from a critique of these methods as profoundly anthropocentric, the authors examine three crucial problematics in which methodocentrism functions in educational research: the institutionalization of graduate training, a wide-spread privileging of the visual (part and parcel of empiricism), and the seeming necessity of "data" in social scientific research methodologies. Ultimately, this article does not reject the necessity of particular studies having methods--rigorous, philosophically grounded approaches to problems in the world--but it argues that the belief that methods must be selected from existing options and assembled "before" approaching the "objects" of study is not only a form of "bad science," it is also deeply implicated in anthropocentric and colonialist politics. Instead, what research requires today is a thorough rethinking of the very distinction between subject and object and a renewed questioning of how agency functions in specific research settings.