Researchers in the field of linguistics have a wide range of methodologies at their disposal. One approach that has seen a marked increase in recent decades is experimental research, which in the wider social sciences is open to a large number of possible designs (Kirk 2003, for example, lists forty for quantitative research alone; see also Lavrakas 2008). In this chapter, we provide an overview of experimental design options available to linguistics researchers, as well as a brief overview of mixed methods, an increasingly common option for investigating complex research questions. The discussion reviews general principles in experimental design, with examples from a selection of subfields. (For further details of experimental methods in theoretical linguistics and psycholinguistics in particular, see Chapters 3 and 8.) The chapter concludes with a discussion of common data collection techniques relevant to experimental designs that are used in a variety of subfields in linguistics.
Fundamentals
Researchers adopting a quantitative approach seek to investigate phenomena by collecting numerical data and analyzing those data statistically. To facilitate this statistical analysis and to control for extraneous variables, quantitative researchers typically recruit a large number of participants and carefully design all aspects of the study before collecting data. In this design process, the quantitative researcher faces a number of questions, including: Do I need more than one group? If so, how many groups are needed to address the research question(s)? How should participants be placed into groups? How will data be collected from the participants, and how often? If an experimental approach is adopted – for example, observations or measurements to be collected under relatively controlled conditions – what will the treatment consist of (e.g., stimuli, timed response, feedback)? How will extraneous variables be addressed?