Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Amy H. Criss


Diffusion Model, Memory Models, Output Interference, Recognition Memory

Subject Categories



The strength based mirror effect (SBME) refers to higher hit rates and lower false alarm rates for strongly encoded items. The SBME has been explained by two alternative mechanisms: differentiation and criterion shift. The differentiation account posits that as the memory traces are strengthened, the memory for the items is less noisy and therefore are less confusable with the stored memory traces. The criterion shift account, on the other hand, suggests that the tendency to endorse a test item differs between strong and weak test lists. When participants receive a test after studying a strong list of items, they require more evidence to endorse an item. This account suggests that the SBME is observed due to the decision processes rather than memory. One strong piece of evidence for the criterion shift account is that the SBME is observed after studying a mixed list in which half of the items are strengthened and the other half are not. The differentiation mechanism can not account for the SBME observed after a mixed study list. However, recent findings of output interference (OI) in recognition memory have been explained by encoding at retrieval, which suggest that differentiation can also take place at retrieval. This suggests a third possible explanation for the SBME might be encoding during retrieval. In a series of experiments, we investigated the effect of strengthening items at study on OI in pure and mixed list paradigms. The results from the pure list paradigm (Experiment 1) showed both OI and the SBME as well as an interaction between OI and the SBME, as the item-noise models predicted. The results from the mixed list paradigm (Experiment 2-4) showed that the SBME was observed only when participants were informed of the strength of the list that they would be tested on. This finding adds constraints to the criterion shift account which states that participants shift their criterion to endorse a test item according to the type of the test list and that shift in the criterion causes the SBME. These results suggest that both criterion shift and differentiation account could be playing a role in the SBME after studying pure lists, and that only criterion shift account plays a role in SBME after studying a mixed list, but only when participants are explicitly aware of the testing conditions.


Open Access

Included in

Psychology Commons