The effects of participantand observer perspectives on the quantity and quality of information recalled from conversations were examined. The observers were exposed to the conversation through videotape, audiotape, or transcript. No differences were found between participants and observers on measures of recall quantity. However, analysis of the composition of recall protocols indicated that participants reported more thematic and evaluative statements while observers produced more errors and elaborations. Observers of videotaped conversations also reported more descriptions and made more references to motiws. Instruction set was also found to influence subjects' recollections.tudies of conversational memory are based on two differing orientations. Some researchers have focused on measure-S ment of the quantity of content recalled from conversations.