During the past decade, progress in research on human communication networks has been impeded by the discovery that self-reports of communication are not valid measures of independently observable behavior. Research on observable communication behavior has since languished because direct measurement of communication behavior by human observers is impractical. This report describes a method for using digital signal analysis of synchronous, independent-channel audio recordings to detect instances of message exchange behavior. The system achieved 95.1% agreement with a human observer in a laboratory test. Considerations for extension to the field are discussed.
All Science Journal Classification (ASJC) codes
- Sociology and Political Science
- Social Sciences(all)