Michael WigelsworthUniversity of ManchesterPhD EducationA multi-level approach to assessing the impact of Social Emotional Learning:Secondary SEALDespite an expanding interest in the concept of Emotional Intelligence (EI),difficulties in definition, measurement and reconciling competing models hasled to an argument that the practical application of EI has overtaken currentlevels of understanding and research (Matthews, Roberts, & Zeidner, 2004a;Zeidner, Roberts, & Matthews, 2002). This is particularly relevant withineducation where a large range of social and emotional learning (SEL)programmes, designed to increase EI in pupils, vary drastically in theirintended outcomes and methods, quality of material and the frequency andquality of evaluation (Hoffman, 2009). To date, the majority of research hasbeen US based and the small quantity of UK research has been focusedeither at the primary level, or has assessed the perception of impact. Thismeans the potential success for SEL to positively improve UK secondary agedpupil outcomes is untested.The aim of the current study was the assessment of the SEALprogramme, a National Strategy for English secondary schools designed topositively influence a range of pupil outcomes, including increased emotionalliteracy, better behaviour and improved mental well-being. Additionally, thevalidity of the underlying relationship between EI and favourable outcomes,beyond identified socio-demographics, was measured. The study utilised apredominately quantitative design with a final sample of 22 schools(approximately 2360 pupils) implementing the SEAL programme, and 19'matched comparison' schools (approximately 1991 pupils), selected on thebasis of similar school level characteristics. Pupils from every schoolcompleted annual self-rated assessments of their emotional literacy (using theELAI), mental well-being and pro social behaviour (using the SDQ) over athree year period. A small case study element (9 SEAL Schools) wasselected from the larger quantitative sample to provide context to thequantitative results. Multi-level modelling (a statistical technique for examininghierarchically clustered data) was used to analyse the results.After controlling for socio-demographic factors, results indicated amarginal non-significant effect in pupil's emotional literacy and mental healthdifficulties as a result of attending a SEAL school, however no effect on prosocial behaviour was found. Results also indicated a differential effect on thebasis of the pupil variables of SEN provision, gender and ethnicity, which wasconsistent with very little variation at school level reported. A significantrelationship was found between emotional literacy, mental health difficultiesand pro social behaviour, indicating a valid theoretical framework, despite thelack of a significant effect of the SEAL programme. Further examinationrevealed that the relationship between the variables may be more complexthan originally theorised, although difficulties with high degrees of confoundbetween the factors limit this interpretation. Implications and directions forfuture research are also discussed.