Information Cascade Effects in Online Event Scheduling
In a research paper written by Daniel M. Romero, Katharina Reinecke, and Lionel P, Robert Jr., they concluded that sequential group decision-making processes like online event scheduling are under social influences when the decisions depend on individual preferences. Previous research has shown that scheduling polls that make all respondents’ answers visible have a greater chance of succeeding, which demonstrates the impact of social influence and coordination. In this paper, they discovered that cascading effects indeed play a critical role in event scheduling, and early respondents have a particularly dominating influence on the poll result.
In the study, the researchers examined over 1.3 million polls generated by Doodle users. Their study produced three main findings: i) Early poll participants have the greatest influence on poll results, and their availability largely determines the final success of the poll. ii) Poll respondents are impacted the most by their preceding neighbor, and this localized influence generates global cascading effects. iii) Poll participants are more likely to find a time that works for everyone when a small number of respondents with the lowest availability are hidden.
The purpose, investigation topic, and results of the study are all highly relevant to the content of INFO 2040 and the study directly applies what we learned about information cascades to a real-world experiment. The first finding suggests that event scheduling is a subjective rather than objective choice, with its success contingent on group dynamics and repondents’ willingness to prioritize the entire group’s goal over their individual interests. The second finding explains that the availability of the nearest neighbor is viewed as a relatively accurate representation of the group’s overall availability. An alternative explanation is the recency effect, in which respondents assume their nearest neighbor before them has already taken others’ availabilities into consideration and thus perceive copying the neighbor’s decision as a group-optimizing strategy. Lastly, the third finding states that the negative cascading effects can be reduced by hiding 35%-50% of people with the fewest available times.
In conclusion, the research paper discusses information cascades in online event scheduling and illustrated that poll participants are especially influenced by the earliest respondents as well as their nearest preceding neighbor. The findings give some insights into how event scheduling systems can be designed and improved to streamline the process of finding a mutually agreeable time.
Source: “The Influence of Early Respondents: Information Cascade Effects in Online Event Scheduling”