Most metaphor theories are based on evidence collected by two widely used methods: analytical methods and behavioural methods. Analytical methods target the structures and functions of metaphor use in order to identify and describe and sometimes evaluate metaphors as part of language use and as part of discourse events. Behavioural methods focus on the processes and products of psychological and social behaviour that is related to metaphor use in order to describe and explain these processes and products as part of general language and discourse behaviour. In recent decades, new methods have arisen in cognitive science, which have also begun to impact research on metaphor in action: biological methods collect and analyse data about the state of the brain and the body; and computational methods specify structures and functions as well as operations upon them which can model which behavioural processes occur in the mind and between people in interaction.

All of these methods can be regarded as empirical in the broad sense of the term. They help in examining data that are collected in order to research testable claims about metaphor in language use and discourse. In the behavioural sciences, the quality of these methods is traditionally measured in terms of reliability and validity. In DMT this is thematized in various ways. In this way, DMT has also fostered methodological research, which focuses on developing appropriate methods for the study of metaphor in action.