Wise or mad crowds? The cognitive mechanisms underlying information cascades
The concept of Information Cascades was one of the most interesting topics that we’ve gone over in class, in my opinion. It’s interesting to see how the opinion of an individual can slowly but surely influence the opinions of those in his or her network. As our world is more and more intertwined with technology, there’s been a clear divide between how information is portrayed. Whether it’s more traditional in the form of newspapers and television or more contemporary like social media, there is a phenomenon on how word travels and how that can be interpreted within different circles of people and opinion. This research paper on Science Advances considers how a single person’s decision is actually the influence of various others. From choosing to vaccinate to financial investments, people observe the decisions of others before them in order to ascertain their own. The focus of this paper is to understand the decisions and conditions behind both negative and positive information cascades. This is a very relevant topic in light of the results of the election and how the country is divided politically and ideologically.
The top part of the attached graph is an illustration of the social drift-diffusion model(DDM), a under many to model the evidence accumulation process during decision making. In the notion of the DDM, there’s a situation where people are given a choice between two options and they base the choice on an internal level of evidence. The initial level of evidence is what prior tendencies they might have that causes them to side with one choice more than the other. As they gather more information, there comes the idea of an evolving level of evidence. This is the jagged line in the diagram. There is variability as one weighs the possible decision and after a choice threshold is met, the decision is made. The upper threshold represents choice A and the lower represents choice B. It’s definitely interesting to see how many different considerations can be boiled down into a model. At the beginning of the line, the accumulated information is what the individual knows up to the present moment. That means that there’s no additional information and not yet a choice made but an individual can still start close towards one choice. One can observe that some individuals take much longer to arrive at their decision, while others need less for convincing. Each possible use of this model can allow for additional research on how different conditions can affect the length of when an individual makes a decision. Whether it’s geographic location, education, or any other measurable aspect.
I believe that this research paper has been referred to previously in a blog post. I didn’t realize until halfway through but I would like to add an additional perspective on how this relates to our class. Clearly we had a boiled down introduction to information cascades in the form of high and low signals. While this paper goes into various different models to test decision making in information cascades. What I walk away from reading this paper is that there are various ways to model social behavior. There can be more complex models used to highlight a specific influence of something but at the same time a basic model can be used to model something that needs less granularity. For example, a bank trying to figure out how to grow their returns for the next quarter requires much less individualized information than trying to model the investing decisions of Americans in different age groups. This paper is a great way to see how topics from class can be further used in the field of research and also how we make our own decisions day to day.
Source: https://advances.sciencemag.org/content/6/29/eabb0266
