Target-language gains are frequently cited as the primary objective of study-abroad (SA) programs. While it is a common “folk belief” that mere presence in a host country facilitates language acquisition as if through osmosis, empirical evidence reveals significant variation in linguistic attainment. To understand these disparate outcomes, researchers have increasingly turned to Social Network Analysis (SNA) to investigate how students’ social interactions condition their progress.

However, traditional research in this area has faced four major limitations:

1. A narrow focus on interactions with native speakers, neglecting the significant role of peer-to-peer communication.

2. Reliance on egocentric networks without verifying the perspectives of the alters in the network.

3. The use of limited, global metrics rather than nuanced structural data.

4. A reliance on one-shot pre- and post-stay snapshots rather than longitudinal tracking.

We will demonstrate how these gaps can be overcome with sociocentric (whole-network) SNA. Unlike previous approaches, sociocentric SNA i) focuses on students’ interactions with their alma mater classmates as well as other agents, ii) reconstructs their complete network, iii) allows tracing the impact of each individual student’s position in the social graph using centrality metrics, and iv) may offer a dynamic developmental perspective with multiple measurement points, gauging the extent to which changes in the interaction networks translate to changes in progress along a range of dimensions.

We will explore the affordances and methodological nuances of computational SNA, drawing on diverse empirical examples including Erasmus+ exchanges, intensive summer programs, and language courses for refugees, in diverse geographical and target-language settings. The approach contributes novel methodology and rigorous insights into the dynamics of study-abroad SLA, and offers tangible recommendations for SA program stakeholders.