<?xml version="1.0" standalone="yes"?>
<Paper uid="W02-0224">
  <Title>Grounding styles of aged dyads: an exploratory study</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
2 Method
</SectionTitle>
    <Paragraph position="0"> The details of the corpus, its collection, and the assignment of tags are given below.</Paragraph>
    <Section position="1" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
2.1 Dialogue Data
</SectionTitle>
      <Paragraph position="0"> Our data consist of 5 dialogues by older dyads (age range: 80-88, sd: 2.1) and 5 dialogues by younger dyads (age range: 23-39, sd: 5.2). Most of the older participants were therefore so-called &amp;quot;older-olds,&amp;quot; meaning that they were between 75 and 85 years of age. Familiarities of the older dyads were strong: they had known each other since they were children or after they moved into the current areas of residence, although they started talking each other frequently after they joined an common local association several years ago. The younger dyads were also familiar with each other: they had worked in the same department of an institute for at least 3 months and up to 5 years and had talked with each other almost every working day.</Paragraph>
      <Paragraph position="1"> Dialogues of older dyads were videotaped in a community center that the subjects regularly visit for activities of their association. Videotaping of younger adults took place in a lounge space of their work place. The dialogues were casual conversations in Japanese, involving no specific objectives or tasks. The dialogue topics between older dyads included community events, common friends, their girlhood, and garden work, while the topics between younger adults included their job, their children, their cars, and common friends. The length of a dialogue in either group was from 30 to 50 minutes. Older adults participated in our data collection on the voluntary basis, while we paid a small amount of money for participation of younger adults.</Paragraph>
    </Section>
    <Section position="2" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
2.2 Transcription and coding
Transcription
</SectionTitle>
      <Paragraph position="0"> We transcribed a 5-minute portion after 3 minutes from the beginning of each dialogue. All words and word fragments audible to the ear were transcribed, including overlapping speech, nonlexical fillers (such as &amp;quot;uh&amp;quot;), and other vocalizations (such as laughter and whistle). After each dialogue was transcribed, we checked the videotape to add data of all noddings and salient gestures (facial expressions and hand movements) to the transcription. The transcriber solicited the help of some local residents to understand some of the dialectical expressions contained used in dialogues by older dyads.</Paragraph>
      <Paragraph position="1"> Coding On the basis of Traum's finite-state transition model of grounding, we divided all speech in our dialogue data into utterance units (UUs), namely, &amp;quot;continuous speech by the same speaker, punctuated by prosodic boundaries (including pauses of significant length and boundary tones)&amp;quot; (Traum, 1994). One of the present authors then classified each utterance unit into one of seven categories of grounding acts, according to its contribution to a grounding process. Table 1 shows the seven categories of grounding acts and their definitions.</Paragraph>
      <Paragraph position="2">  An acknowledgement claiming or demonstrating understanding of a previous utterance. It may be either a repetition or paraphrase of all or part of the utterance, an explicit signal of comprehension such as &amp;quot;ok&amp;quot; or &amp;quot;uh huh&amp;quot;, or an implicit signaling of understanding.</Paragraph>
      <Paragraph position="3">  To check the reliability of our coding scheme, we asked two independent subjects to code a portion of our data and compared their results with our own. As an instruction to these independent coders, we produced a coding manual that lists Traum's definitions of the seven categories of grounding acts (see Table 1), sample utterances of each category, and some general notes on treatments of marginal cases such as fillers, simultaneous utterances, and utterances with both Init and Ack functions. The subjects did some exercise coding before actual coding, and they were encouraged to ask questions about the coding standard during the exercise period. Both transcribed text and audio recording were available to the subjects during the actual coding. The subjects coded two partial dialogues one from older dyads and the other from younger dyads. Each dialogue was about 2.5 minute in length.</Paragraph>
      <Paragraph position="4"> The interrater reliability among the original coder and these two subjects was excellent; there was an 83% agreement, with a Cohen's Kappa of .78, for dialogues by older dyads (UUs = 138, coders = 3, categories = 9) and a 78% agreement, with a Cohen's Kappa .77, for dialogues by younger dyads (UUs = 128, coders = 3, categories = 9).</Paragraph>
    </Section>
  </Section>
class="xml-element"></Paper>