【正文】
t have been contributed for different movies. This implies variability in the purchasing population’s average propensity to review different movies online.5. Conclusions We find that a purchasing population’s average propensity to review a movie online has: (1) a Ushaped relationship with the average valence of that movie’s reviews。 (2) a positive relationship with that movie’s marketing budget。 (3) a positive relationship with the level of disagreement among professional reviews for that movie。 (4) a negative relationship with the number of screens where the movie is exhibited。 (5) a positive relationship with the volume of previously posted reviews for the same movie.We thus find support for Hypothesis 1, 2, 3, 4 and 5b. Our findings reject the hypothesis of altruism/concern for others as the primary motivation for posting online product reviews. A collective reading of our empirical findings points, instead, to self expression/self confirmation motives as being the dominant force behind online review contribution: consumers are more likely to review very good and very bad movies (because these movies trigger strong feelings that consumers feel the urge to express), controversial movies (because then their review will be read with more attention by other consumers who seek to reduce their uncertainty about these movies), lesser known movies (because this allows consumers to showcase their eclecticism and sophistication), and movies around which they perceive that there is a lot of online buzz (because this implies that other people are more likely to read their review, or simply because it’s fun to do what everybody else is doing). Interestingly, we find that moviegoers appear to be more eager to review movies that are less widely released and that are lowerranked in the box office. This result remains robust even after we control for differences in the intrinsic munication habits of populations attracted by different movies. The result has implications for marketing campaigns that are interested in generating “online buzz” about a product. Another intriguing finding is that large numbers of previously posted reviews increase subsequent viewers’ propensity to post more reviews for the same movie even though, past a certain volume, additional reviews are unlikely to add little to what has already been said. This result has potential implications for the design of online product discussion munities as it suggests that, from society’s perspective, there might be inefficiencies in the way consumers allocate their online reviews. 8