Following Friends vs. the Crowd – Information Cascades in Online Movie Ratings
Information cascades, the idea of following the crowd and joining in its behavior regardless of one’s own private information/signals, is a concept that has been widely explored in the context of the social media revolution. However, recently, researchers have explored the application of information cascades with respect to user behavior when submitting online movie ratings, specifically focusing on the difference in herding behavior found among friends’ movie reviews as well as among strangers’ reviews. The researchers found that there was indeed a significant difference in behavior when user’s responded to the reviews of their friends compared to the reviews of strangers.
It was found that friends’ ratings always induced an information cascade among other friends in the friend group, implying a sense of social pressure and conformity when subsequent friends in the cascade submitted their reviews of a movie. On the other hand, with respect to a user’s response to the reviews of strangers, there was a lower yet still significant likelihood that a user would follow the reviews of the crowd/strangers. For example, it was found that as the popularity of a movie increased, the number of subsequent reviewers that provided negative ratings in response to positive prior ratings decreased, highlighting herding behavior. For movies that were extremely popular, it was found that subsequent reviewers had a much higher likelihood of contributing positive ratings, once again demonstrating an information cascade through a crowd of strangers.
However, demonstrating the value of a user’s friends’ ratings over that of strangers’ reviews, it was found that the number of friends’ prior ratings of a movie helped to reduce the herding effect on a user. For example, if a user saw that a movie had generally negative reviews from the crowd, but the user had a higher number of friends who had differing opinions (i.e. these friends had positive reviews of the movie), the user himself would be less influenced by the crowd and more influenced by his friends. Reinforcing this idea, it was found that the herding effect on a user was stronger as there were more prior ratings by friends.
Overall, these findings highlight the bias and different social dynamics and pressures involved in the postings of online movie reviews. We see how information cascades occur in the context of movie reviews most often when those reviews are from friends. In this way, it implies that subsequent users are much more strongly influenced by the prior reviews of their friends compared to the reviews of strangers. A combination of pressure to conform combined with a user’s increased sense of validity in a friend’s movie review highlights how an information cascade of friend’s opinions of a movie can outweigh that of the crowd.
Source:
Young-Jin Lee, Kartik Hosanagar, Yong Tan (2015) Do I Follow My Friends or the Crowd? Information Cascades in Online Movie Ratings. Management Science 61(9):2241-2258. http://dx.doi.org/10.1287/mnsc.2014.2082