Mayer Lukas W, Bocheva Desislava, Hinds Joanne, Brown Olivia, Piwek Lukasz, Ellis David A
Department of Cognitive Sciences, UCI School of Social Sciences, University of California Irvine, Irvine, CA, 92617, USA.
Decisions and Operations Division, School of Management, University of Bath, Claverton Down, Bath, BA2 7AY, UK.
Behav Res Methods. 2025 Sep 17;57(10):289. doi: 10.3758/s13428-025-02745-9.
Across disciplines, research often relies on groups of people to participate in experiments or attend events at the same time. Typically, researchers try to maximize attendance by manually identifying a set of times that suit the diaries of many individuals. However, this is inefficient, is prone to error, and can lead to a final sample that is not large enough to provide meaningful inferences. While current scheduling tools are useful for individual-based research, enabling participants to select times convenient to them within a researcher's preset parameters, they are less useful in research that requires specific or flexible group sizes. In response, we present Optimeet, a web application that allows researchers to upload participants' availability data and generate an optimal allocation schedule for multiple groups. We describe the function of the underlying applet, which identifies a schedule to maximize attendance by treating it as a computational problem involving combinatorial optimization (Experiment 1). Our solution relies on an empirical comparison of parameter-free heuristics to make allocation decisions that make the best use of participants' availabilities and the derivation of appropriate performance metrics. Of the algorithms evaluated, one consistently outperformed comparable versions of existing tools, which we verified in a further exercise (Experiment 2) involving a large human sample (N = 5,289). We consider the methodological utility and practical value of these developments, and include detailed documentation, code, and a video tutorial so that researchers can rapidly employ Optimeet to support group research.
在各个学科中，研究通常依赖于一群人同时参与实验或参加活动。通常，研究人员会通过手动确定一组适合许多人日程安排的时间来尽量提高参与人数。然而，这种方法效率低下，容易出错，并且可能导致最终样本量不足以提供有意义的推断。虽然当前的调度工具对于基于个体的研究很有用，能让参与者在研究人员预设的参数范围内选择方便自己的时间，但在需要特定或灵活分组规模的研究中，它们的用处就较小了。作为回应，我们推出了Optimeet，这是一个网络应用程序，允许研究人员上传参与者的可用时间数据，并为多个小组生成最优分配时间表。我们描述了底层小程序的功能，它将确定时间表以最大化参与人数的问题视为一个涉及组合优化的计算问题（实验1）。我们的解决方案依赖于对无参数启发式算法的实证比较，以做出能充分利用参与者可用时间的分配决策，并推导适当的性能指标。在评估的算法中，有一种算法始终优于现有工具的可比版本，我们在涉及大量人群样本（N = 5289）的进一步实验（实验2）中验证了这一点。我们考虑了这些进展的方法学效用和实际价值，并提供了详细的文档、代码和视频教程，以便研究人员能够迅速使用Optimeet来支持小组研究。
