A values-affirmation intervention changed biology classroom friendship networks
Kate M. Turetsky, Valerie Purdie-Greenaway, Jonathan E. Cook, Joseph P. Curley, Geoffrey L. Cohen
Science Advances
Network design
Nodes
students in a gateway college biology class, embedded within the complete roster network
Ties
directed, weighted friendship nominations measured at baseline and again near the end of the course
Methods
The researchers compared randomized groups on network size, degree and closeness, modeled continuation into the next biology course, and used Bayesian mediation analyses to examine whether friendship-network changes could carry the treatment association.
Source: Science Advances
Reviewed summary
In a gateway biology course, students were randomized to a 15-minute values-affirmation or control writing task in week three. Among 226 students with baseline and end data, treated students reported about 29.2 percent more friends and were 11.7 percentage points more likely to enroll in the next course. The experiment provides unusually strong evidence that a brief psychological intervention can alter relational integration, connecting individual experience with classroom network structure and later academic participation.
Affirmed students reported larger friendship networks, higher degree and closeness, and greater subsequent course enrollment; mediation estimates were consistent with friendship growth as one pathway, especially for outgoing nominations. Of 552 invited students, 328 consented, 290 received an intervention, and 226 completed both network waves, so attrition and missing nominations matter. The work concerns one course, outgoing and incoming ties differed, and statistical mediation does not by itself prove the social mechanism.
SNA
How SNA was used
The study defined students in a gateway college biology class, embedded within the complete roster network as nodes and directed, weighted friendship nominations measured at baseline and again near the end of the course as ties. The researchers compared randomized groups on network size, degree and closeness, modeled continuation into the next biology course, and used Bayesian mediation analyses to examine whether friendship-network changes could carry the treatment association.
Source
Science Advances, 6(45), eaba9221
DOI: 10.1126/sciadv.aba9221
