The rise of online information pools---or information repositories comprised of individual's unique contributions---has sparked much research regarding individuals' motivations to share information in these contexts. However, the effect of non-contributors on individuals' motivation to share their own information has been understudied, in spite of the prevalence of non-contribution. Using the bystander effect as a theoretical lens, the present study explored the effect of others' contributive behavior on individuals' decision to share information on an online review site. In contrast to prior research, this study experimentally manipulated and measured the three psychological mechanisms underlying the bystander effect, allowing for a better understanding of the relationship among these mechanisms in the context of online information pools. A between-subjects 4 (bystander volume; 6, 49, 242, 831) x 2 (anonymity; anonymous, not anonymous) x 2 (contribution appropriateness; 4% contributing, 73% contributing) factorial design was used (N = 243).

Although bystander volume is typically related to greater feelings of diffusion of responsibility, the present study found no significant between bystander volume and diffusion of responsibility; in fact, bystander volume was positively related to contribution amount, though this relationship was not linear. However, consistent with prior research, diffusion of responsibility was negatively related to contribution amount. Although prior research has found that individuals' behavior is influenced by the observed behavior of others, individuals' contribution amount and length in the current experiment was not affected by the social norms displayed in the information pool. Consistent with past research, individuals' feelings of anonymity were significantly, negatively related to evaluation apprehension, though neither anonymity nor evaluation apprehension was significantly related to contribution amount.

Together, the findings of the present study suggest that the bystander effect and its underlying mechanisms operate differently in the context of an online review site than they do in face-to-face contexts or more traditional mediated contexts, such as email. Possible explanations for these differences and suggestions for future research are presented.