Abstract
Getting-acquainted conversations for 55 dyads were tape recorded. Speaker affect was evaluated by speakers, conversation partners and observers. On-off vocal activity was analysed to yield 14 interaction tempo variables. Factor analysis reduced these to four factor scores. Multiple regression analyses used these interaction tempo factors to predict speaker affect ratings made by speakers, partners, and observers. Speaker evaluations were significantly related to predictable patterning of talk within speaker and durations of interruptions and pauses. Partner ratings of speaker affect were significantly related to these two factors and to speaker talkativeness. Outside observer ratings were related to all four interaction tempo factors. Speakers who were more talkative, had longer interruptions and shorter pauses, whose amount of talk was more closely coordinated with partner talk, and whose amount of talk was predictably patterned over time, were rated more positively. A higher percentage of variance in observer ratings (44%) was explained than in speaker (16%) or partner (19%) ratings; participant ratings were probably influenced by many other types of information while observers had only voice to base their evaluations on.
Get full access to this article
View all access options for this article.