Abstract
Serial reproduction reveals how information changes when communicated through a social network. Studies using it find that information is transformed to fit cultural norms, expectations, stereotypes, (e.g., Bartlett, 1932) audience attitudes (Higgins & Rholes, 1978); and simplified and polarized (e.g., Gilovich, 1987). Information content and the relationship between conversers also influences information spread (Granovetter, 1973; Hanson, 1999). Serial reproduction methodology could improve understanding of complex information flow in organizations. Knowledge is now the most important strategic resource in organizations, and its management is critical to organizational success (Ipe, 2003). Research using grapevine or similar methodology with verbal information spread in organizations (Bordia & Rosnow, 1998; Davis, O’Connor & Regis, 1976) has examined factors like the timing and extent of information spread and information change. In contrast, serial reproduction follows written information through a chain of people, recording how information changes in each step of a communication chain. It can be more or less structured, allowing for natural communication, or the creation of predetermined communication lines (simulating organizational lines of communication). Thus one can examine, control for or manipulate relationships and information. We present serial reproduction methodology and sample research, examining how stereotypical information changes depending on communicators’ relationship. Four-member serial reproduction chains were created, manipulating the relationship between the participants. Initial participants read a story created by the researchers including information varying in stereotypicality and complexity. After a distractor task participants recreated the story for the next participant. Each succeeding participant received the story written by the previous participant. Texts were analysed using MaxQDA, a qualitative tool for coding texts: coded for changes in stereotypicality, compared to the previous and the original text. Findings show substantial information change over chains. Stereotype consistent information tended to be added more than other types of information. Applications in organizational contexts will be discussed.