Patterns – Laboratory – DCTs

In her assessment of different research methods, Yuan (2001) also assessed the usefulness of discourse completion tests. She administered the tests both orally and in writing. She gives a detailed description of the set-up:

The procedure of the oral DCT is this. First, the instructions as well as the 24 DCT scenarios were tape-recorded by a male and a female native speaker, both in their early 30s, of Kunming Chinese. The male voice recorded the 12 compliment scenarios and the female voice recorded the 12 scenarios of CRs. Informants were invited to the researcher’s residence individually, at a time of their choice. They listened to the scenarios one by one and responded to each scenario orally. A second tape-recorder was kept running to record the oral sessions in their entirety. All the recordings were done by the researcher herself. Respondents of the written DCT questionnaire, however, were given the freedom of filling out the questionnaire at home at a time of their convenience. No time limit was set. (Yuan 2001: 274)

She recorded 87 subjects who filled in 24 scenarios. That means that her data consisted of 2088 compliment sequences. Her aim was to assess the reliability of the research method. She investigated the length of the responses (in number of characters), and several features that she claims to be very frequent in everyday conversations, viz. exclamation particles, repetitions, inversions, and omissions. In this way, she tries to assess the oral quality of the responses in order to compare the written and the spoken responses. Not surprisingly, she finds that oral responses are much longer than written responses (2001: 278). The number of exclamations, repetitions, inversions and omission all have higher frequencies in the oral DCT data than in the written data (2001: 279).