The main models used to study persistence are Astin’s Theory of Involvement in Higher Education, Pascarella’s General Model for Assessing Change and Tinto’s Theory of Student Departure. Tinto proposed that academic and social integration reinforce students’ commitment to their institution and educational goals improving retention. This claim was assessed with an algorithm for mutuality that evaluates social integration in a network. It is not known if standard academic records may be used for sociometric techniques applied in engineering education research. This paper will introduce this approach and, inparticular, discuss the social network parameter “mutuality” and study its relationship to persistence in engineering. Mutuality is an index that assesses the tendency for individuals in a group to reciprocate choices more frequently than would occur by chance, thus mutuality reflectsreciprocity beyond random grouping, due to students having free selection of groups.The records of the Multiple-Institution Database for Investigating Engineering Longitudinal Development were grouped to establish which students took classes in each other’s presence, a simplified mutuality algorithm was evaluated with this data and probability Weibull models were fitted for persisters and non persisters. The models for persisters shown larger mutuality scales with lower shapes than those for non-persisters, meaning that they paired with classmates more frequently than the students leaving.Results suggest that indexes for social networks may be calculated using standard academic records, facilitating the assessment of social integration and assisting the analysis of the departure puzzle’s factors and informing policy making processes in education.