A quantum-like information processing model with memory noise for question order effect

This paper aims to better explain a typical quantum-like phenomenon in human uncertain decision-making, i.e., question order effect (QOE), by characterizing the evolution of human preferences for decision-making during information processing. It is achieved mainly by introducing two mechanisms into the information processing model with quantum mathematical formalism: immediate feedback and noise disturbance. The immediate feedback mechanism is based on the projection postulate and non-commutative operators, providing the model with the basic ability to explain QOE. This is essentially consistent with one existing well-known quantum-like model. Moreover, the noise disturbance mechanism from memory is proposed for the first time, which formalizes the evolution of human preferences from the unstable state formed by immediate feedback to the stable preferences represented by long-term memory. This mechanism can weaken QOE by partially offsetting the immediate feedback. In applying to five datasets, the model performs better in explaining QOE than three types of existing quantum-like models. Besides, the response replicability effect, which typically contradicts the interpretation of QOE in other quantum-like models, can be explained by the proposed model owing to the noise disturbance. The quantum-like information processing model considering memory noise provides an innovative and noteworthy insight into the evolution of human preferences in information processing, especially the quantum-like phenomena in uncertain decision-making.