Your research brought together a number of different approaches and subjects that we study here at MACSS- digital experiments, statistical inference, network efforts and social contagion. Thank you!
How could the set up of your experiment be extended so as to model the following behaviour:
1) Virality (across social media) of challenges aimed at social causes- Movember for Breast Cancer awareness, the Ice Bucket Challenge, effective Kickstarter campaigns
2) Herd Behaviour- Incorrect and self-aggravating 'wisdom' around correlations in financial markets, finally leading to crashes and bubble bursts - ranging from Holland's Tulip Mania in 1637 to the 2007-08 crash in the United States?
Your research brought together a number of different approaches and subjects that we study here at MACSS- digital experiments, statistical inference, network efforts and social contagion. Thank you!
How could the set up of your experiment be extended so as to model the following behaviour: 1) Virality (across social media) of challenges aimed at social causes- Movember for Breast Cancer awareness, the Ice Bucket Challenge, effective Kickstarter campaigns 2) Herd Behaviour- Incorrect and self-aggravating 'wisdom' around correlations in financial markets, finally leading to crashes and bubble bursts - ranging from Holland's Tulip Mania in 1637 to the 2007-08 crash in the United States?