Back to History

Fields Changed

Registration

Field Before After
Trial End Date July 31, 2019 August 01, 2019
Last Published June 29, 2019 04:24 PM July 17, 2019 03:28 AM
Intervention End Date July 31, 2019 August 01, 2019
Planned Number of Clusters 10 markets 20 markets
Planned Number of Observations 90 students 180 students
Sample size (or number of clusters) by treatment arms 5 control markets, 5 treatment markets 10 control markets, 10 treatment markets
Intervention (Hidden) With the help of this randomized controlled experiment, we intend to investigate the influence of social interactions on asset markets. The study will provide an empirical test of the hypothesis that social interactions increase asset price bubbles. The experimental design follows the bubble game experiment by Caginalp et al. (2001) and Andrade et al. (2016) with two critical variations. 1. The experiment will not be executed in the lab, but instead make use of a web-based app, which allows participants to take part in the experiment with their smart phones. As noted by Giamattei and Lambsdorff (2019) this can be employed for lab-in-the-field experiments, allowing the investigation of behavior in natural settings. 2. Different from Andrade et al. (2016) the treatment condition will not make use of videos shown to investors, but instead allow investors within the market to interact with each other using typical social media features such as likes. Thus, while the control condition replicates the standard bubble game by Caginalp et al. (2001), the treatment condition additionally allows participants to like the trades of their peers with typical social media features. With the help of this randomized controlled experiment, we intend to investigate the influence of social interactions on asset markets. The study will provide an empirical test of the hypothesis that social interactions increase asset price bubbles. The experimental design follows the bubble game experiment by Caginalp et al. (2001) and Andrade et al. (2016) with a critical variation: Different from Andrade et al. (2016) the treatment condition will not make use of videos shown to investors, but instead allow investors within the market to interact with each other using typical social media features such as likes. Thus, while the control condition replicates the standard bubble game by Caginalp et al. (2001), the treatment condition additionally allows participants to like the trades of their peers with typical social media features.
Pi as first author No Yes
Back to top