Information valuation and confirmation bias in virtual communities: Evidence from stock message boards

JaeHong Park, Prabhudev Konana, Bin Gu, Alok Kumar, Rajagopal Raghunathan

Research output: Contribution to journalArticlepeer-review

82 Scopus citations


Virtual communities continue to play a greater role in social, political, and economic interactions. However, how users value information from these communities and how that affects their behavior and future expectations is not fully understood. Stock message boards provide an excellent setting to analyze these issues given the large user base and market uncertainty. Using data from 502 investor responses from a field experiment on one of the largest message board operators in South Korea, our analyses revealed that investors exhibit confirmation bias, whereby they preferentially treat messages that support their prior beliefs. This behavior is more pronounced for investors with higher perceived knowledge about the market and higher strength of belief (i.e., sentiment) toward a particular stock. We also find a negative interaction effect between the perceived knowledge and the strength of prior belief on confirmation bias. Those exhibiting confirmation bias are also more overconfident; as a result, they trade more actively and expect higher market returns than is warranted. Collectively, these results suggest that participation in virtual communities may not necessarily lead to superior financial returns.

Original languageEnglish (US)
Pages (from-to)1050-1067
Number of pages18
JournalInformation Systems Research
Issue number4
StatePublished - Oct 22 2013


  • Confirmation bias
  • Investment decisions
  • Overconfidence
  • Stock message boards
  • Virtual communities

ASJC Scopus subject areas

  • Management Information Systems
  • Information Systems
  • Computer Networks and Communications
  • Information Systems and Management
  • Library and Information Sciences


Dive into the research topics of 'Information valuation and confirmation bias in virtual communities: Evidence from stock message boards'. Together they form a unique fingerprint.

Cite this