Abstract
This study uses a repeated measures design to examine the effects of the self on recognition and source attribution. Working in pairs, 30 undergraduate psychology student participants generated words themselves or listened to their partner's generated item, and made metamemory judgments – judgments of learning and judgments of source – about their likelihood of recognising both the item and its source a week later. The generation task required them to produce in succession names of towns and names of occupations or hobbies of particular significance to themselves. At test, participants recognised, as old or new, words generated by themselves or their partner previously, along with a unique experimenter-generated list of distracter items. Additionally, they characterised each old response as being “remembered” (having memory qualities associated with sensory and contextual information) or “known” (seeming familiar but lacking contextual information). False alarm rates for remembered words suggested that false alarms could be recollectively experienced. Subjects correctly predicted that they would recognise self-generated items better than other-generated items, and that they would be more accurate in remembering the source of self-generated items compared to self-generated items. A new finding is that judgements of source can be sensitive to a variable (self) that influences source discrimination.