Home Help
 Academic Programs
 Research
 Libraries
 Medical Center
 Athletics
 Arts
 Events Calendar
 Prospective Students
 Students
 Faculty & Staff
 Alumni
 Neighbors
 About Columbia
 A–Z Index
 E-mail & Computing


Columbia News
Search Columbia News
 
Advanced Search
News Home | New York Stories | The Record | Archives | Submit Story Ideas | About | RSS Feed
Peer Pressure the Wildcard in Predicting Hit Single

According to sociology Ph.D. student Mathew Salganik, it is a myth that "music executives can create stars at will." In a study he conducted of 14,000 music-loving teenagers, songs were more likely to do well if listeners thought that their peers admired them -- something that is too complicated to predict by any expert measures.

Salganik simulated the effects of peer pressure by asking the study's participants to visit a Web site with 48 songs by relatively unknown bands. They could listen to songs, rate them and decide whether to download them. In the control group, listeners had no information about which songs had been downloaded by others, whereas in various "social influence" groups, they could see which songs had been popular with other users.

In general, participants tended to give higher ratings to the songs that had been downloaded often, and to download these songs themselves, creating a snowball effect.

For instance, "Lockdown," by the Milwaukee punk band 52metro, ranked in the middle among listeners with no social context, but then rose to the top and sunk almost to the bottom in two different sample groups that had access to other people's opinions.

Salganik's study, coauthored by associate professor of sociology Duncan Watts and ISERP research scientist Peter Dodds, appeared in the Feb. 10 issue of the journal Science.

Published: Mar 08, 2006
Last modified: Mar 09, 2006