Master of Arts (MA)
In two experiments, I investigated whether providing accuracy feedback on recognition memory tests affects discriminability of encoded targets from lures. The primary hypothesis was that feedback is a source of criterion noise which leads to lower discriminability. Additionally, it was predicted that separate sources of criterion noise might have additive effects. In both experiments, the presence of feedback was manipulated within-subjects. In Experiment 1, participants completed two recognition tests in which they made either “old/new” decisions or responded using an 8-point confidence scale. Feedback lowered discriminability for both response type conditions, although a slightly larger deleterious effect was observed in the “old/new” response condition. Whether people responded either with “old/new” decisions versus on an 8-point confidence scale had no effect on discriminability. In Experiment 2, I manipulated the strength of study items whereby half of the items were studied once (weak) and the other half were studied four times (strong). At test, these targets were intermixed with an equal number of lures. Additionally, the presence of color cues indicating the expected strength of test items was varied between-subjects. Feedback decreased discriminability, although this was primarily for the strong items. The presence of color cues marking expected strength had no effect on discriminability. Taken together, these results suggest that feedback has a deleterious effect on recognition discriminability and that this may result via feedback introducing criterion noise into the recognition decision.
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Franks, Bryan, "Feedback as a Source of Criterion Noise in Recognition Memory" (2014). LSU Master's Theses. 437.