[Paper Review] Online Actions with Offline Impact: How Online Social Networks Influence Online and Offline User Behavior

Tim Althoff, Pranav Jindal+1 more|arXiv (Cornell University)|Dec 9, 2016

Impact of Technology on Adolescents20 citations

TL;DR

This study investigates how social network features in a physical activity tracking app influence both online engagement and offline physical activity. Using a natural experiment and causal inference on 6 million users over 5 years, it finds that new social connections increase online activity by 30%, retention by 17%, and daily steps by 7% (≈400 steps), with social influence accounting for 55% of behavior change.

ABSTRACT

Many of today's most widely used computing applications utilize social networking features and allow users to connect, follow each other, share content, and comment on others' posts. However, despite the widespread adoption of these features, there is little understanding of the consequences that social networking has on user retention, engagement, and online as well as offline behavior. Here, we study how social networks influence user behavior in a physical activity tracking application. We analyze 791 million online and offline actions of 6 million users over the course of 5 years, and show that social networking leads to a significant increase in users' online as well as offline activities. Specifically, we establish a causal effect of how social networks influence user behavior. We show that the creation of new social connections increases user online in-application activity by 30%, user retention by 17%, and user offline real-world physical activity by 7% (about 400 steps per day). By exploiting a natural experiment we distinguish the effect of social influence of new social connections from the simultaneous increase in user's motivation to use the app and take more steps. We show that social influence accounts for 55% of the observed changes in user behavior, while the remaining 45% can be explained by the user's increased motivation to use the app. Further, we show that subsequent, individual edge formations in the social network lead to significant increases in daily steps. These effects diminish with each additional edge and vary based on edge attributes and user demographics. Finally, we utilize these insights to develop a model that accurately predicts which users will be most influenced by the creation of new social network connections.

Motivation & Objective

• To understand the causal impact of online social networks on both online app engagement and offline physical activity.
• To disentangle the effects of social influence from increased intrinsic motivation in users who join social networks.
• To identify which users are most susceptible to behavior change following the creation of new social connections.
• To develop a predictive model for identifying users likely to increase physical activity after forming new social ties.
• To use objective, sensor-based physical activity data to avoid biases inherent in self-reported behavior.

Proposed method

• Leveraged a natural experiment in delayed friendship formation to isolate causal effects of new social connections.
• Applied difference-in-differences and matching techniques from econometrics to estimate causal impacts on behavior change.
• Used a large-scale dataset from the Azumio Argus app with 6 million users, 631 million activity posts, and 160 million days of accelerometer-based step tracking.
• Distinguished between self-reported in-app activity (online behavior) and passive step counts (offline behavior) as separate outcome measures.
• Built predictive models using user demographics, pre-connection behavior, edge attributes, and temporal patterns to forecast susceptibility to social influence.
• Evaluated model performance using AUC, achieving over 78% AUC in predicting post-connection activity increases.

Experimental results

Research questions

• RQ1 To what extent do new social network connections causally increase online in-app activity and offline physical activity?
• RQ2 How much of the observed behavior change is due to social influence versus increased intrinsic motivation from joining the network?
• RQ3 Which user and edge characteristics predict greater susceptibility to behavior change after forming new social ties?
• RQ4 Can we accurately predict which users will increase their physical activity following the creation of a new social connection?
• RQ5 How do the effects of social influence vary over time and across different types of social connections (e.g., friend vs. follow)?

Key findings

• The creation of new social connections increases online in-app activity by 30% and user retention by 17%.
• New connections lead to a 7% increase in offline physical activity, equivalent to approximately 400 additional steps per day.
• Social influence accounts for 55% of the observed behavior change, while the remaining 45% is attributed to increased intrinsic motivation.
• Subsequent edge formations lead to temporary, diminishing increases in daily steps, with effects decaying over time.
• Predictive models incorporating demographics, pre-connection behavior, and edge attributes achieve an AUC of over 78% in forecasting activity increases.
• The predictive power varies by edge type and initiator, with the highest AUC (0.847) observed for follower-initiated connections.