Information Cascades with Social Media
With new and emerging technologies appear all the time, its very difficult to see what technologies will be adopted and which technologies will be abandoned. This study examines the initial adoption of technology through not only information cascades but also a variety of other theories including microblogging. Microblogging is a form of social media where one user sends a message to multiple users. This system is a derivative of an information cascade, as this system has a direct audience unlike in a standard information cascade where one’s observations were keen. In this study, Twitter was used as the social medium by which microblogging could take effect in. Here, a standard information cascade was set up for the users. Theres an order for all of them, and they are asked to evaluate a new technology that is given to them. Their thoughts and reactions were to go on twitte, where everyone can reply and see thoughts and opinions on this. However, order is important here because as the information cascade will show, individual ix’s decision will change for all i0 – ix-1’s choices and opinions. The only difference here is that the use of twitter can give a much clearer picture of each users thoughts and opinions on this technology instead of simply accepting or rejecting it. In the data the researchers concluded that even with an enhanced tool to give thoughts and opinions on adopting technology, only half of the participants ended up adopting the technology. After closer analysis of posting trends and responses on twitter, the researchers did indeed conclude that an information cascade was in effect in the very beginning of the experiment. Almost all of those who did ultimately adopt the technology where also the first individuals to cast their adopt or reject answer. However, this cascade quickly died off as more and more users were able to see more and more information about each person’s reasoning of adopting the technology. After this cascade died off, an entirely new cascade appeared where now the trend was leaning towards non adoption of the technology. This is rather strange of an information cascade to cause such a great divide between adopters and non adopters. Usually in a standard information cascade we would see and landslide in either direction but here the data stands still. What could have caused this double cascade?
What the researches concluded is that it was because of the nature of twitter. What happened was that once the initial information cascade took place and people stopped posting and talking about it, others who have not chosen yet were left only recent information on their feeds and could only account for a smaller chunk of the total population of those who did approve in the beginning. With such little posts and only recent information, the initial information cascade faded away as users had to rely more on their own opinions on the product they have been asked to approve or reject. With this, a shift occurred where now the later participants were rejecting the technology simply because of the lack of updated information about everyone else’s opinion. Here an information cascade once again occured in the oposite diection, this time consuming the rest of the particapants own opinions. The big takeaway from this study is that information cascades can be amplified through social media only if continuous amounts of information is provided for users not just initially looking at the product or item. With a continual information flow through social media, and information cascade can play effect, however if left unattended it can cause harm to the cascade. Quite an interesting study this was and I found it to be both informative and surprising in the fact that information cascades are very fragile indeed. In the future the researchers plan to pursue more extensive research on not just information cascades but also information visualization as a key factor in sustaining information cascades.
http://www.academia.edu/4491686/Explaining_technology_adoption_with_information_cascades_A_study_of_microblogging_data