A vast literature investigating behavioural underpinnings of financial bubbles and crashes relies on laboratory experiments. However, it is not yet clear how findings generated in a highly artificial environment relate to the human behaviour in the wild. It is of concern that the laboratory setting may create a confound variable that impacts the experimental results. To explore the similarities and differences between human behaviour in the laboratory environment and in a realistic natural setting, with the same type of participants, we translate a field study conducted by reference (Sornette, D.; et al. Econ. E-J. 2020, 14, 1–53) with trading rounds each lasting six full days to a laboratory experiment lasting two hours. The laboratory experiment replicates the key findings from the field study but we observe substantial differences in the market dynamics between the two settings. The replication of the results in the two distinct settings indicates that relaxing some of the laboratory control does not corrupt the main findings, while at the same time it offers several advantages such as the possibility to increase the number of participants interacting with each other at the same time and the number of traded securities. These findings pose important insights for future experiments investigating human behaviour in complex systems.

中文翻译：

---

现场和实验室实验资产市场的行为影响和市场动态

大量研究金融泡沫和崩盘行为基础的文献依赖于实验室实验。然而，目前尚不清楚在高度人工环境中产生的发现如何与野外人类行为相关。令人担忧的是，实验室环境可能会产生影响实验结果的混杂变量。为了探索实验室环境和现实自然环境中人类行为之间的异同，我们使用相同类型的参与者，翻译了一项通过参考进行的实地研究（Sornette, D.; et al. Econ. EJ. 2020, 14, 1-53），每轮交易持续六天，实验室实验持续两个小时。实验室实验复制了实地研究的主要发现，但我们观察到两种环境之间的市场动态存在重大差异。结果在两种不同设置中的复制表明，放松一些实验室控制不会破坏主要发现，同时它提供了几个优势，例如增加参与者之间互动的数量的可能性同时和交易的证券数量。这些发现为未来研究复杂系统中人类行为的实验提供了重要的见解。同时，它提供了一些优势，例如可以增加同时相互交互的参与者数量和交易证券的数量。这些发现为未来研究复杂系统中人类行为的实验提供了重要的见解。同时，它提供了一些优势，例如可以增加同时相互交互的参与者数量和交易证券的数量。这些发现为未来研究复杂系统中人类行为的实验提供了重要的见解。