The Power of Silence: An Analysis of the Aggregation and Reporting Biases in User-Generated Contents

User-generated contents (UGC) such as online reviews are inherently incomplete since we do not capture the opinions of users who do not write a review. These silent users may have systematically different online experiences than those who speak up. These differences can be driven by users‟ differing sentiments towards their online experiences as well as their disposition to generate UGC. Indiscriminately aggregating UGC across different sentiment levels can lead to an aggregation bias and overlooking the silent users‟ opinions can result in a reporting bias. We develop a method to rectify these two biases simultaneously through an inverse probability weighting (IPW) approach. In the context of users‟ movie review activities at Blockbuster.com, we found that the average probability for a customer to post a review is 0.06 when the customer is unsatisfied with the movie, 0.23 when indifferent, and 0.32 when satisfied. A user‟s reporting probability with positive experience first order stochastically dominates the one with negative experience. We then adjust common UGC measures such as review volume and sentiment using these estimated reporting probabilities as weights. We show that these rectified measures yield superior predictive power, as opposed to the raw ones. Our proposed approach provides a realistic solution for business managers to paint a more complete picture from the inherently incomplete UGC.