ABSTRACT
This chapter compares three response methods commonly used in sentence acceptability experiments: yes/no, an n-point (Likert) scale, and magnitude estimation. An experiment was conducted in which participants were randomly assigned to one of these three response methods. The experiment examined three syntactic phenomena with different degrees of expected acceptability contrasts: presence or absence of inversion in wh-questions (an extremely clear contrast), the that-trace effect (a relatively clear contrast), and subextraction from embedded subjects and wh-subjects (a relatively subtle contrast). In addition, the inversion cases compared second-person pronoun, third-person pronoun, and lexical DP subjects (an extremely subtle contrast). The experiment shows that the three response methods provide strikingly consistent results across a wide range of contrast types, from very clear to very subtle. Given the well-known disadvantages of magnitude estimation, including difficulties both for experimenters and for participants, the yes/no and n-point response methods appear to be more reasonable choices for most sentence acceptability experiments.
