We study peer influence of production and consumption of projects in the Scratch community, an online platform developed by MIT Media Lab and targeted for young children, where users collectively learn programming by creating and sharing projects. We investigate if Scratchers are influenced by the popularity of their peers’ projects and their peers’ preferences for consuming from specific baskets of projects. We find that the popularity of Scratchers’ projects is significantly influenced by the production popularity of their peers. Testing for heterogeneity in influence, we find that Scratchers are not influenced by specific peers who might have highly popular projects, instead it seems that they are influenced by just the aggregate popularity of all peers. We find that Scratchers who have a minimum activity of one month on the platform are more susceptible to peer influence. Scratchers with high tendency to create projects by rebuilding on existing projects on the platform tend to have significant improvements in their future production popularity (due to influence from peers’ production popularity) only in the short term and not in the long run. We also disentangle a self decision making mechanism from other mechanisms that might explain the channel of influence: we find that a significant proportion of the estimated influence from peers is mediated via Scratchers’ decision to create new projects. This highlights Scratchers’ subsequent behavioural decisions in response to existing popularity of peers’ projects. We find evidence of polarized consumption patterns on the platform, i.e., there are certain groups of projects (discovered in an unsupervised manner based on co-consumption patterns) for which Scratchers have high specificity. We do not make claims about how such groups form on the platform - for example, whether it is a conscious choice or is a result of the way the platform is organized. However, we find that such polarization is not a consequence of Scratchers being influenced by their peers’ consumption patterns.

Peer influence of production and consumption behaviour in an online social network of collective learning

Samantray A.;Riccaboni M.
2020-01-01

Abstract

We study peer influence of production and consumption of projects in the Scratch community, an online platform developed by MIT Media Lab and targeted for young children, where users collectively learn programming by creating and sharing projects. We investigate if Scratchers are influenced by the popularity of their peers’ projects and their peers’ preferences for consuming from specific baskets of projects. We find that the popularity of Scratchers’ projects is significantly influenced by the production popularity of their peers. Testing for heterogeneity in influence, we find that Scratchers are not influenced by specific peers who might have highly popular projects, instead it seems that they are influenced by just the aggregate popularity of all peers. We find that Scratchers who have a minimum activity of one month on the platform are more susceptible to peer influence. Scratchers with high tendency to create projects by rebuilding on existing projects on the platform tend to have significant improvements in their future production popularity (due to influence from peers’ production popularity) only in the short term and not in the long run. We also disentangle a self decision making mechanism from other mechanisms that might explain the channel of influence: we find that a significant proportion of the estimated influence from peers is mediated via Scratchers’ decision to create new projects. This highlights Scratchers’ subsequent behavioural decisions in response to existing popularity of peers’ projects. We find evidence of polarized consumption patterns on the platform, i.e., there are certain groups of projects (discovered in an unsupervised manner based on co-consumption patterns) for which Scratchers have high specificity. We do not make claims about how such groups form on the platform - for example, whether it is a conscious choice or is a result of the way the platform is organized. However, we find that such polarization is not a consequence of Scratchers being influenced by their peers’ consumption patterns.
2020
Causal mediation mechanism
Computational social science
Homophily
Human behaviour
Online social network
Peer influence
Social influence
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.11771/17193
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
social impact