Discrimination and Metacognitive Sensitivity in Boundary Extension:
Applying Signal Detection Theory to BE
Boundary extension (BE) is a robust phenomenon where people remember seeing things that might have been present beyond the edge of the original image. Participants in two experiments (N = 75) studied scenes in either a close (C) or wide (W) version. At test they saw half of the scenes in the same version (CC or WW), and the other half in a different version (CW or WC). They indicated whether the camera position was the same as or different from the one at study and their confidence about the judgment on a 6-point response scale. The SDT-based analysis uses d’ to measure people’s ability to discriminate old from new (discrimination sensitivity), and compares meta-d’ with d’ to assess how accurately their confidence reflects their performances (metacognitive sensitivity). We found d’ was significantly higher in the wide-study condition (WW and WC), suggesting participants had better discrimination sensitivity when studying the image in its wide version. The data didn’t reveal significant differences in metacognitive sensitivity. The findings suggest that discrimination sensitivity contributes to BE effect, and the metacognition of BE tasks is inefficient since meta-d’ is smaller than d’ in both conditions.