Presenter(s): Rennie Kendrick − Biology
Faculty Mentor(s): Dasa Zeithamova, Caitlin Bowman
Poster 32
Research Area: Cognitive Neuroscience
Memory allows us to link across multiple experiences to derive new information. For example, if we see a person walking a dalmatian, and later see another person walking the same dalmation, we may infer that the two people are married. There are two models of how we accomplish this associative inference. According to the flexible retrieval model, individuals retrieve separate memories (person 1-dalmatian; person 2-dalmation) and infer about their relationship (person 1-person 2) when needed. According to the integrative encoding model, we retrieve the memory of the first person we saw with the dalmation while seeing the second, and form an integrated memory that links the two. I hypothesize that how readily we integrate older memories with new experience depends on how well established the prior knowledge is. To test this hypothesis, participants encoded object pairs (AB and BC) that shared an object B (e.g., banana-clock, clock-keyboard). Each pair was repeated three times. Half of the AB and BC pairs were presented in a blocked format (AB, AB, AB, BC, BC, BC) and half in an interleaved format (AB, BC, AB, BC, AB, BC). Later, participants were tested on the indirect AC association (banana-keyboard). I predicted that participants would infer more quickly in the interleaved condition because the two episodes were already linked at encoding via integrative encoding. Preliminary data show faster inference in the interleaved condition. Further investigation into the effect of blocked vs. interleaved training on learning could lead to enhanced teaching methods.