Longitudinal Associations Between Emotional Intelligence, Behavior Difficulties and Academic Attainment in Early Adolescence: A Study of Developmental Cascades

Developmental cascades refer to, “the cumulative consequences for development of the many interactions and transactions occurring in developing systems that result in spreading effects across different levels, among domains at the same level, and across different systems or generations” (Masten & Cicchetti, 2010, p.491).  The study of these processes has gained increasing prominence in recent years, making important contributions to developmental theory and offering insights for prevention and intervention.  In this paper we use a developmental cascades model to examine the longitudinal inter-relationships between emotional intelligence, academic attainment and behavior difficulties in early adolescence.  We test three key hypotheses adapted from Moilanen, Shaw and Maxwell (2010) - adjustment erosion/fortification, academic (in)competence and shared risk (aka ‘common cause’).

The study utilizes secondary analysis of data from a large, government-funded study of social and emotional learning.  Our sample comprises a nationally representative sample of nearly 4,000 adolescents (aged 11/12 at the outset of the study) drawn from 41 secondary schools across England.  Data was collected on an annual basis (T1, T2 and T3; except academic attainment: T1 and T3 only).  Emotional intelligence was assessed using the Emotional Literacy Assessment and Intervention battery (Southampton Psychology Service, 2003).  Behavior difficulties were assessed using the conduct problems subscale of the Strengths and Difficulties Questionnaire (Goodman, 1997).  Academic attainment was measured using data from national assessments in core curriculum subjects. Finally, two shared risk variables – socio-economic deprivation and special educational needs (SEN) – were derived from England’s National Pupil Database.

A developmental cascade model was built using structural equation modeling. Following Moilanen et al. (2010), two versions were tested.  The first version was a cascade model which accounted for temporal stability in emotional intelligence, behavior difficulties and academic attainment; concurrent correlations between these variables; and cascade pathways across domains over time.  The second version tested the shared risk hypothesis by adding deprivation and SEN as predictors of the other variables at each time point.  We will present these models and discuss their implications for the adjustment erosion/fortification, academic (in)competence and shared risk hypotheses.  Additionally, we will also consider the application of this work to the school-based prevention literature (e.g. the social-emotional learning logic model; CASEL, 2007).