ABSTRACT
Traditionally, the study of psycholinguistic processes has been separate from the study of dialogue and the study of linguistic corpora. This chapter shows that two classic on-line psychological measures, reaction time and error rate, can be defined for a coded corpus of spoken dialogues. There is good reason to believe that reaction time and error rate measures can be found for dialogue. In laboratory paradigms, reaction time, measured from the presentation of a stimulus to the onset of a verbal response, is used to study perception, comprehension, and problem solving. Inter-move interval functions as a general measure of cognitive difficulty with independent contributions made by order, task difficulty, comprehension, production, and interpersonal factors. For both inter-move interval and disfluency, multiple regression results confirm the findings of studies exploring single cause-effect relationships. Models of speech production, for example, will have to consider why disfluency rate is insensitive to attested difficulties of comprehension.
