Perceived seriousness of offense: the ignored extraneous variable

A widely used method for eliciting written speech act data is to ask speakers how they would respond in hypothetical situations. The seriousness of the offenses described in these situations has been shown to influence responses, but the possibility of variation in respondents' assessments of seriousness of offense has not been investigated systematically. In this study, participants wrote responses to various situations and rated the seriousness of the offense involved. In five out of ten situations, there was considerable variation in ratings. Most of this variation could not be traced to speaker characteristics, such as gender. On the other hand, respondents' perceptions of seriousness of offense seemed to influence responses to the elicitation task. For instance, in four out of ten situations, those who considered an offense more serious were significantly more likely to request redress. Therefore, the results show that considering individual variation in ratings of seriousness of offense can help explain patterns in speech act data that would otherwise seem unsystematic. (C) 2003 Elsevier B.V. All rights reserved.