Collaborative Research: Investigating Compensatory and Noncompensatory Decision Making Strategies in Dynamic Task Environments

在现代、不断发展的工作场所中，需要支持工人适应新的需求和机会。传统上，人类判断的特点是加权策略，其中在决策标准之间进行权衡。例如，大学招生委员会考虑申请者的多个方面，通常，平均成绩差的人可以通过优异的入学考试成绩来弥补。然而，在存在时间压力或高工作量的情况下，人类的判断通常是基于规则的，其中因素没有加权。例如，空中交通管制员可能会根据一些显著的线索在高工作量的情况下诊断潜在的问题。这项研究将增加关于决策环境如何影响决策者使用的策略的日益增长的知识体系。虽然加权策略和基于规则的策略等特征是从大量的实验研究中推导出来的，但还没有框架来模拟和预测个体决策者判断策略的变化。这项研究的中心主题是建立一个数学框架，在工作量和压力水平不同的工作环境中，模拟从加权策略到基于规则的策略的转变，反之亦然。这项研究将建立在现有工作的基础上，创建一个从人类数据中推断判断规则的模型。多个实验将比较基于规则的模型和常用的加权模型在不同的工作量和时间压力情况下的性能。假设增加的工作量和时间压力将促进从加权战略到基于规则的战略的系统性转变。这项工作将对理解复杂、动态环境中的决策做出有价值的贡献。这项研究的一个潜在意义是设计了一种基于动态工作环境条件的、适应用户需求的辅助机制。

In the modern, evolving workplace there exists a need to support workers in adapting to novel demands and opportunities. Traditionally, human judgments are characterized by weighted strategies where tradeoffs between decision criteria are made. For example, college admissions committees consider multiple facets of an applicant and, often, a poor grade point average can be compensated by a superior entrance exam score. However, on occasions in which time stress or high workload exists, human judgments are typically rule-based where factors are not weighted. For instance, an air traffic controller is likely to diagnose potential problems in high workload situations based on a few salient cues.This research will add to the growing body of knowledge concerning the way in which a decision environment influences the strategy used by decision makers. While characterizations such as weighted and rule-based strategies have been deduced from a multitude of experimental studies, no framework exists to model and predict shifts in judgment strategy in individual decision makers. The central theme of this research is to create a mathematical framework to model the shift from a weighted to a rule-based strategy, and vice versa, in work environments where workload and stress levels vary. This research will build upon existing work to create a model which infers judgment rules from human data. Multiple experiments will be conducted to compare the performances of the rule-based model with a commonly-used weighted model under varying workload and time stress situations. It is hypothesized that increasing workload and time stress will promote a systematic shift from weighted to rule-based strategies. This work will make a valuable contribution toward understanding decision making in complex, dynamic environments. One potential implication of this research is the design of an aiding mechanism that adapts to the needs of the user based on conditions of dynamic work environments.