|
Title:
|
AGGREGATING DIFFUSE INFORMATION WITH SUBGROUPS |
|
Author(s):
|
Kay-yut Chen , Tad Hogg |
|
ISBN:
|
972-98947-8-7 |
|
Editors:
|
Nitya Karmakar and Pedro Isaías |
|
Year:
|
2004 |
|
Edition:
|
Single |
|
Keywords:
|
Experimental economics, information aggregation, knowledge markets, subgroup. |
|
Type:
|
Oral Presentation - 30 minutes |
|
First Page:
|
45 |
|
Last Page:
|
54 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Decisions based on information spread among a group require mechanisms to efficiently uncover and aggregate this information. The growth of computer networks dramatically lowers the transaction costs of such mechanisms, potentially enabling the broad use of suitably designed methods. Information markets are one approach to information aggregation, but can perform poorly when applied to relatively small groups. In such cases, prior studies have shown the benefits of using direct queries to the individuals combined with a reward for correct answers. Since such queries can be costly, we propose and experimentally evaluate mechanisms using only a subgroup of the individuals. The experiments show queries to a few members of the group can be sufficient to enable correct decisions, even if they are selected randomly. Better performance is seen by selecting subgroup members based on their performance with respect to prior queries in a repeated-game context. Computer-mediation readily maintains such information and can automatically select appropriate subgroups. |
|
|
|
|
|
|
