Information processing under memory constraint

Abstract

Human memory is limited and humans tend to distort early information subconsciously in many memory recall situations. This paper aims to address the situation of memory distortion by constructing a model that will yield the optimal distortion in information processing, minimizing the cost in the context of limited memory. Two mathematical models will be built using economics statistical concepts to understand human behaviours in the recalling and processing of information, given the scenarios of perfect memory and limited memory. Results show that more weight will be placed on the signal with higher accuracy and less distortion is done when individuals are familiar with the true state. When there is a higher probability of forgetting the first signal, there will be higher distortion to include more of the first signal to obtain a more informed decision. However, when there is cost of distortion involved, the level of distortion will be lower. In addition, the models explain implications where prior expectations affect experience and first impressions matter. Through these implications, companies will be able to better understand the importance of advertising and packaging as expectations and impressions will affect the decision making of consumer.