Virtual Exchanges (VEs), facilitate online, collaborative work between students across national boundaries, contributing to gains in intercultural competency (ICC) and second language (L2) skills among others. To make VEs more effective as a tool for improving ICC, implementors of VEs need methods to quantify changes in ICC and identify factors associated with such changes. This longitudinal study investigates associations between changes in ICC scores as measured using the Beliefs, Events and Values Inventory (BEVI), individual factors (such as participation, gender, and L2 proficiency), and group factors (such as ICC scores of group mates) after two successive eight-week L2 VEs. Pre- and post-data for a cohort of 202 first-year participants in the two VEs at a Japanese university were collected. For 38 participants with initially low ICC scores (below the cohort median), meaningful gains (more than 5 points on the BEVI) were observed after the two VEs. Binary logistic regression analysis revealed that the meaningful gains in ICC scores were associated with participation and initial ICC scores of group mates but not gender or L2 proficiency. The implications of these findings for implementors of L2 VEs are discussed.

Keywords: Virtual exchanges, intercultural competency, BEVI, second language proficiency, gender, individual factors, group factors.