Autonomous Agents & Multiagent Systems/Agent Theories, Architectures, and Languages, 2007
We investigate a framework where agents search for satis- fying products by using referrals from ... more We investigate a framework where agents search for satis- fying products by using referrals from other agents. Our model of a mechanism for transmitting word-of-mouth and the resulting behavioural effects is based on integrating a module governing the local behaviour of agents with a mod- ule governing the structure and function of the underlying network of agents. Local behaviour incorporates
We use an agent-based model to examine the influence of word-o f-mouth communication in environme... more We use an agent-based model to examine the influence of word-o f-mouth communication in environments where agents select products by using positive referrals from oth er satisfied agents as well as their own preferences and perce ptions. The role of referrals is illustrated using a complex decisio n environment that incorporates a number of psychological realities. Using the model, we
This paper presents the results of an experiment investigating the effects of using different for... more This paper presents the results of an experiment investigating the effects of using different formats for representing uncertain attribute evaluations on decision making. Study participants make a series of hypothetical choices using six uncertainty formats – probability distributions, expected values, standard deviations, three-point (minimum–median–maximum) approximations, quantiles, and scenarios – and effects on decision making are tracked in terms of the quality of the final choice, the specific characteristics of the selected alternatives, and the difficulty experienced in making a decision. The results provide insights into how subjects make single- and multi-criteria choices in the presence of uncertainty (and some format for representing uncertainty) but in the absence of any real facilitation. The use of probability distributions appeared to overload subjects with information, leading to poorer and more difficult choices than if some intermediate level of summary was used – in particular three-point approximations or quantiles.► How uncertainty is represented exerts a significant influence on decision making and the alternative that is eventually chosen. ► For the tasks performed in this study, the use of probability distributions appeared to overload subjects, leading to relatively poor choices. ► Subjects seemed to derive benefit from formats providing some intermediate level of summary of the full probability distributions, specifically quantiles and scenarios. ► Uncertainty format has a significant effect on the level of difficulty experienced, with the level of difficulty experienced increasing with the quantity of information provided. ► The use of standard deviations becomes more difficult relative to other uncertainty formats as the number of objectives increases.
Uncertainty is present in many decisions where an action's consequences are unknown because they ... more Uncertainty is present in many decisions where an action's consequences are unknown because they depend on future events. Multi-attribute utility theory (MAUT) offers an axiomatic basis for choice, but practitioners may prefer to use simpler decision models for transparency, ease of use, or other practical reasons. We identify some ‘simplified’ models currently in use and use a simulation experiment to evaluate their ability to approximate results obtained using MAUT. Our basic message is that avoiding assessment errors in the application of a simplified model is more important than the choice of a particular type of model, but that the best performance over a range of decision problems is from a model using a small number of quantiles.► We use simulation to compare a number of models using simplified representations of uncertainty with multi-attribute utility theory. ► For good results, avoiding assessment errors is more important than the choice of a particular type of decision model. ► Over a range of decision problems, the best performance is given by a model using a small number of quantiles. ► Models using explicit risk attributes like variances or probabilities of failure perform relatively poorly. ► Models using scenarios can give good results but this depends heavily on the construction of scenarios.
Autonomous Agents & Multiagent Systems/Agent Theories, Architectures, and Languages, 2007
We investigate a framework where agents search for satis- fying products by using referrals from ... more We investigate a framework where agents search for satis- fying products by using referrals from other agents. Our model of a mechanism for transmitting word-of-mouth and the resulting behavioural effects is based on integrating a module governing the local behaviour of agents with a mod- ule governing the structure and function of the underlying network of agents. Local behaviour incorporates
We use an agent-based model to examine the influence of word-o f-mouth communication in environme... more We use an agent-based model to examine the influence of word-o f-mouth communication in environments where agents select products by using positive referrals from oth er satisfied agents as well as their own preferences and perce ptions. The role of referrals is illustrated using a complex decisio n environment that incorporates a number of psychological realities. Using the model, we
This paper presents the results of an experiment investigating the effects of using different for... more This paper presents the results of an experiment investigating the effects of using different formats for representing uncertain attribute evaluations on decision making. Study participants make a series of hypothetical choices using six uncertainty formats – probability distributions, expected values, standard deviations, three-point (minimum–median–maximum) approximations, quantiles, and scenarios – and effects on decision making are tracked in terms of the quality of the final choice, the specific characteristics of the selected alternatives, and the difficulty experienced in making a decision. The results provide insights into how subjects make single- and multi-criteria choices in the presence of uncertainty (and some format for representing uncertainty) but in the absence of any real facilitation. The use of probability distributions appeared to overload subjects with information, leading to poorer and more difficult choices than if some intermediate level of summary was used – in particular three-point approximations or quantiles.► How uncertainty is represented exerts a significant influence on decision making and the alternative that is eventually chosen. ► For the tasks performed in this study, the use of probability distributions appeared to overload subjects, leading to relatively poor choices. ► Subjects seemed to derive benefit from formats providing some intermediate level of summary of the full probability distributions, specifically quantiles and scenarios. ► Uncertainty format has a significant effect on the level of difficulty experienced, with the level of difficulty experienced increasing with the quantity of information provided. ► The use of standard deviations becomes more difficult relative to other uncertainty formats as the number of objectives increases.
Uncertainty is present in many decisions where an action's consequences are unknown because they ... more Uncertainty is present in many decisions where an action's consequences are unknown because they depend on future events. Multi-attribute utility theory (MAUT) offers an axiomatic basis for choice, but practitioners may prefer to use simpler decision models for transparency, ease of use, or other practical reasons. We identify some ‘simplified’ models currently in use and use a simulation experiment to evaluate their ability to approximate results obtained using MAUT. Our basic message is that avoiding assessment errors in the application of a simplified model is more important than the choice of a particular type of model, but that the best performance over a range of decision problems is from a model using a small number of quantiles.► We use simulation to compare a number of models using simplified representations of uncertainty with multi-attribute utility theory. ► For good results, avoiding assessment errors is more important than the choice of a particular type of decision model. ► Over a range of decision problems, the best performance is given by a model using a small number of quantiles. ► Models using explicit risk attributes like variances or probabilities of failure perform relatively poorly. ► Models using scenarios can give good results but this depends heavily on the construction of scenarios.
Uploads
Papers by ian durbach