Temporal dynamics of choice behaviour: an examination of pre- and post-choice latencies in rats and humans

摘要

Animals and humans face choices every day. Survival depends on whether the choices we make are adaptive and reflect the demands of the environment. One feature of the choice environment is that the outcomes associated with our choices are stochastic (probabilistic). Two other features are that choice outcomes are not directly indicated but must be learnt through experience, and that choice outcomes can vary over time. The process of choice thus consists of various stages within which complex features of the environment are integrated to guide choice. Previous research has mainly focused on how animals and humans allocate their choices. However, choice allocation does not capture the entire choice process. This thesis investigates choice behaviour beyond choice allocation by assessing pre- and post-choice latencies when animals and human choose between two probabilistically rewarded options. The experiments presented here used a discrete-trial procedure which allowed the choice process to be decomposed into three distinct behavioural stages. We investigated rat choice behaviour in Chapters 2 and 3, where we analysed response latencies within these three distinct choice stages and investigated how they related to changes in choice allocation. In Chapter 4, we sought to identify the differences and similarities in choice behaviour between rats and humans.In Chapter 2, we simulated two types of environments to investigate how closely rats approximated optimal foraging. We found that rats approximated the optimal strategy across the two environments that differed in the average probability of reward as well as the dynamics of the depletion-renewal cycle. In Chapter 3, we presented rats with reward probabilities that were dynamic across three phases: reward probabilities changed in size (large or small) and direction (increase or decrease). We found that a large decrease in reward probabilities rapidly produced the greatest change in choice proportions. In Chapter 4, we compared choice behaviour between rats and humans under two paradigms. Firstly, we used a cued choice paradigm where subjects were presented with a cue that directed them as to which of two options to respond for rewards. Secondly, subjects were presented with an un-cued task, where they were free to choose between two options in order to procure rewards. An interesting difference between rats and humans was observed: choice behaviour for humans, but not rats, was sensitive to free-choice aspects of the tasks.In this thesis, choice was examined across a variety of environments, and across rats and humans. Throughout all the experiments reported here, the analysis of behavioural latencies provided information about environmental contingencies which were not available from an analysis of choice allocation alone. This thesis thus demonstrates novel methods for assessing choice behaviour with the potential to also elucidate the underlying neural mechanisms.