A Comparison of Item- and Context-Noise Models in Recognition Memory

Abstract

Item-noise models (b; Shiffrin and Steyvers 145-166) assert that recognition memory performance depends on item information of other traces in the list. REM therefore proposes that performance decreases with increasing list length (list-length effect). Context-noise models (BCDMEM; Dennis and Humphreys 452-478) assert that the main factors in determining memory performance are the contexts in which the stimulus has been learned. Therefore, context-noise models assume null list-length effect when the potential confounds are controlled. Due to the contradictory predictions, list-length effect is critical to test recognition memory models. There is a significant amount of literature demonstrating the importance of both item and context information in recognition. However, further research must be conducted to understand in which level of recognition these two types of information have an impact on. Especially in the observation of output interference, literature has not been revealed as to whether the main cause of a decrease in memory performance through test phase is item-noise, context-noise or other confounding variables.

Keywords

• Tanıma Belleği
• Bellek Modelleri
• Liste Uzunluğu Etkisi
• Testin Bozucu Etkisi