Dataset Take your Time

Despite recent trends towards graphical represent information, little is known about its impact on decision consistency in multiple criteria decision problems. In an online experiment, we query the performance ratings of six subordinates and examine the influence of information representation (graphs vs. tables vs. graphs with data labels) on decision consistency. We further address the (unavoidable) subjectivity of weighting multiple criteria by applying the analytic hierarchy process that is not yet applied in this research field. This methodological approach overcomes weaknesses of previous studies examining decision accuracy based on objective weights. Somewhat surprisingly, our results show that tables lead to the highest decision consistency and effort levels. They suggest that decision-makers prefer to minimize effort, i.e. decision time and the share of analytical judgment processes, if they are confronted with a mixed information representation (graphs or graphs with data labels). We caution against an unreflective use of graphical information representations in management reports and presentations lead to worse decision consistency and are no panacea for higher decision consistency in multiple criteria decision problems.尽管当前信息可视化呈现已成研究与应用趋势，但学界对其在多准则决策问题中对决策一致性的影响仍鲜有深入探究。本研究依托一项线上实验，采集六名下属的绩效评分数据，探究三种信息呈现形式——图表、表格、带数据标签的图表——对决策一致性的影响。针对多准则权重赋值过程中不可避免的主观性问题，本研究采用了当前该研究领域尚未广泛应用的层次分析法（Analytic Hierarchy Process，AHP），该方法弥补了过往基于客观权重开展决策准确性研究的固有缺陷。颇为意外的是，实验结果显示，表格呈现形式能带来最高的决策一致性与任务投入程度。研究结果还表明，当决策者面对混合式信息呈现形式（纯图表或带数据标签的图表）时，更倾向于降低认知投入——具体体现为缩短决策时长、减少分析性判断流程的占比。本研究特此提醒：切勿盲目在管理报告及演示文稿中使用图表类信息呈现形式，这类形式会削弱决策一致性，并非提升多准则决策问题中决策一致性的万能解决方案。