Using Social Network Analysis to Investigate Sociocultural Competence in Two-Way Immersion Programs

• In Event: Defining and Operationalizing Sociocultural Competence in Dual-Language Education

Abstract

Objectives
We used Social Network Analysis (SNA) to examine sociocultural competence (SCC) among students in one Spanish-English two-way immersion (TWI) kindergarten classroom. We examined a friendship network for ‘homophily,’ the tendency for people to create ties with people like themselves. We posited that students with higher levels of SCC would exhibit higher levels of ‘heterophily,’ meaning they would be more likely to name friends from different language groups. We investigated measures of homophily/heterophily across levels of analysis (classroom, language group, and individual student).

Conceptual Framework
Early work on SCC in TWI contexts focused on positive cross-cultural attitudes of students (Christian, 1996; Lindholm, 1990), especially towards speakers of other languages. More recent work carried out in international education and the European context, (Deardorff, 2006, 2011; Council of Europe, 2018) suggests that SCC manifests in inclusive attitudes, knowledge and skills that lead to appropriate and effective behavior with cultural others. We propose SCC includes attitudes and behaviors that facilitate friendship formation between students across different language groups.

Methods and Data Sources
We interviewed all 21 kindergarten students in the class and asked them to list their friends, yielding a directed binary network (Figure 3.1). We categorized students based on language dominance (English, Spanish, or Bilingual) and used UCINET 6.730 (Borgatti, Everett, & Freeman, 2002) to calculate homophily measures for student ties at the classroom, language group, and student level. For classroom and student level data, we used the standard E-I measure (Krackhardt & Stern, 1988) which ranges from -1 (perfect homophily - a student only naming friends from the same language group) to 1 (perfect heterophily - students only naming friends NOT in the same language group). We explored homophily at the group level by investigating density by groups. Akin to correlation, each cell reflects a percentage of ties that exist between groups, where higher numbers mean members of the group on the row frequently named friends in the group on the column.

Findings
The overall classroom friendship network has an adjusted E-I measure of -0.063 (p < 0.001), suggesting a small degree of homophily, and is fairly integrated in terms of named friendships. However, the density by groups analyses provide more nuanced insight into student relations. The greatest number of ties were between English dominant students naming other English dominant students as friends. The least prevalent ties were Spanish dominant students naming English dominant students as friends. The high degree of homophily among English dominant students, clear in Figure 3.1 and Table 3.1, was obscured at the whole network measure by the relative lack of homophily displayed by the rest of the class. The E-I measures for each student (see Table 3.2) revealed tendencies towards homophily that were consistent with the group level data.

Scholarly Significance
This study introduces SNA as a novel way to investigate SCC. We operationalize homophily/heterophily as an outcome measure of SCC, and provide a nuanced understanding of SCC outcomes across levels of analyses in one classroom (classroom, group, and student).

Authors

• Erika Feinauer, Brigham Young University

• Rich DeJordy, Fresno State University

• Elizabeth R. Howard, University of Connecticut