The present study examined the effect of addition and deletion of stim
uli to the current environment on perceivers' performance of memory ta
sk. Further, the robustness of the asymmetric effects of addition and
deletion was investigated. 48 objects in a room were grouped into four
schematicaily consistent or inconsistent conditions. In the experimen
t, participants viewed a series of an original scene and a target scen
e with the placement of an object changed from the original scene befo
re the recall and the recognition tasks. Participants were asked wheth
er they noticed change from the original scene to the target scene. Th
e recall tasks were performed both verbal and written. In the recognit
ion task, each of the 24 objects (17 objects were actually present in
the scene, but seven mere not) was visually presented on the computer
display along with confidence ratings. Analysis showed that participan
ts were more likely to recall placement of objects classified as highl
y expected and highly salient according to an hypothesized schema of p
lacement. There was also a significant effect of addition on recall pe
rformance when objects were classified as low in expectation but highl
y salient were added to the target scene. The effects of addition on s
chema-inconsistent information were discussed in terms of informativen
ess of the information.