ABSTRACT
Experiments, alone or in combination with other methods of data gathering, are growing in popularity among sociolinguists because they can provide different kinds of data that contribute to our understanding of social variation in language use. A primary strength of production experiments is their efficiency. In an experiment, the target list of utterances is designed to elicit an answer to the research question and is established by the experimenter before any data are collected. In production experiments, the researcher has much more control over the materials, and carefully designed experiments can yield valid and highly reliable data. Perception experiments can also be used to determine how the social categories that we identify in our production data map onto the social categories of non-linguists. To capitalize on the strengths of experimental design and maximize the control over the materials, read speech is often elicited in production experiments and presented to listeners in perception experiments.
