Investigation of the Factor Structure and Differential Item Functioning of the Child and Adolescent Mindfulness Measure (CAMM): Analysis of Data from a School-Based Cluster Randomised Controlled Trial

Abstract

Objectives: This study used data from a randomised controlled trial of a school-based mindfulness programme in the UK to investigate the structure and performance of the 10-item Child and Adolescent Mindfulness Measure (CAMM). Method: The study included 7924 children and adolescents aged 11 to 14 years. Participants provided CAMM data at pre-intervention, 7 months (post-intervention) and 1 year. Exploratory factor analysis (EFA) of pre-intervention data was undertaken. Multiple indicators multiple causes (MIMIC) models were fitted to pre-intervention responses to investigate differential item functioning across groups defined by gender, year group and ethnicity. Response shift resulting from receiving the mindfulness programme was investigated by fitting MIMIC models to compare item functioning between the intervention and control arms. Results: EFA results indicated that the 2-factor model was a good fit. Eight items were associated with the first factor, while the remaining two items, which specifically addressed avoiding unwanted thoughts and feelings, were associated with the second factor. MIMIC model findings indicated that girls scored lower (ostensibly less mindful) on 4 items than boys that had the same latent level of mindfulness; as a result of receiving the mindfulness programme, participants scored lower on one item (“At school, I walk from class to class without noticing what I’m doing”) after holding latent level of mindfulness constant. Conclusions: Findings indicate that the CAMM has a 2-factor structure in the UK in late childhood and early adolescence. While we did observe some differences in how individual items performed across groups, these differences were small compared to the overall variability in the CAMM scores. Preregistration: Current controlled trials ISRCTN86619085.