Drawing on a number of key case studies, this innovative book develops a novel model that highlights how experiments are one of the most important empirical data sources in cognitive linguistics. It covers three problem areas: the uncertainty of experimental data, contradictory experimental results and the relationship between experimental data and theories. It provides metascientific tools which make it possible to (re-)evaluate the reliability and the usability of experiments as data sources, to reveal the relationships between closely related experiments and to summarise their outcomes. By increasing meta-scientific awareness, this book allows researchers to conduct systematic and comprehensive re-evaluations of experiments, which will contribute to more conscious and transparent experimental practices of experimenting in cognitive linguistics, as well as in other fields of research. It is essential reading for researchers and advanced students in cognitive linguistics, psycholinguistics and other areas that involve the study of experimental linguistic data.