The thesis herein explores the relationship between early and late information in judgement and decision-making and tests a quantitative model of this relationship based on contemporary dual-process theory. The first chapter reviews literature regarding early information as a potential biasing factor in judgement and decision-making, the neglect of dual-process theory in the domain and the tendency to rely on static modelling techniques derived from economic theory. The first empirical chapter concludes that a synthesis of a static-economic decision model (prospect theory) with contemporary dual-process theory principles can better predict choice behaviour than either one approach alone. I conclude that dual-process theory provides a strong theoretical basis for understanding the cognitive processes involved in early-informational biases, but also that the quantitative approaches to modelling choice behaviour can provide valuable additional insights. The third chapter acts on this conclusion by developing a dynamic-stochastic choice model (based on a sequential sample process) which reflects four contemporary dual-process theory concepts that are relevant to early-informational biases. Simulation results of the model are presented in order to demonstrate the choice behaviour predicted by this approach. The rest of the thesis is dedicated to empirical studies designed to test the implications of these simulation results and these predicted behaviours. The empirical studies cover a range of domains including biased predecision processing during evidence gathering, stereotype bias in multi-attribute decision-making under time-pressure and the impact of expectation and accuracy motivation on visual-search decision-making. I conclude that the dynamic-stochastic modelling approach demonstrates some clear value in understanding the cognitive processes involved in these domains and the results support the use of contemporary dual-process theory as a framework for understanding judgement and decision-making. Based on this conclusion I outline some future developments for a more nuanced dynamic model including integration with a more sophisticated way of modelling type 2 processing and expansion to account for hypothetical thinking principles. I also suggest future research domains for application of the model such as expert decision-making and multi-alternative decision problems.