Rethinking boundary extension: The role of source monitoring in scene memory

Boundary extension (false memory beyond the edges of a view; Intraub & Richardson, 1989) is thought to be a source monitoring error that occurs when memory for the highly constrained amodal continuation of a picture is mistaken for visual memory (Intraub & Dickinson, 2008). Based on the source monitoring framework, three series of experiments tested the prediction that increasing the similarity between two sources of information will increase boundary extension. Observers viewed a sequence of 12-16 pictures for 14 seconds each. At test, the same view of each was presented and rated as being the same, closer up or farther away on a 5-pt scale. Series 1 and 2 examined if enhancing the representation of the scene beyond the boundaries of the picture through language and sound effects increased boundary extension. Series 2 and 3 examined if decreasing the quality of the visual representation, by varying the visual characteristics of the stimuli and increasing the retention interval between study and test increased boundary extension. The results showed that boundary extension always occurred but was affected by factors that influenced the visual representation but not by enhancing the larger context. These findings suggest that boundary extension is a source monitoring error that is specifically tied to the visual information. How much of the entire scene representation is misattributed to vision is modulated only by factors that affect the quality of visual memory.