Examining Mindfulness Training for Teachers: Theoretical and Methodological Extensions of Intervention Effectiveness

This dissertation seeks to extend the field of mindfulness-based interventions (MBIs) for teachers, both theoretically and methodologically. The first study is a systematic review conducted of randomized controlled trials examining the effectiveness of MBIs for school teachers. The purpose of Study 1 was to determine theoretical and methodological next steps for the field. Results of the theoretical review indicate more empirical evidence is needed examining mindfulness practices used in the interventions, distal impacts on classrooms and students, and mediation tests connecting proximal and distal outcomes. The methodological review indicates that more studies should focus on measuring the fidelity of implementation of the intervention, include longer follow-up times of measurement and larger sample sizes, and utilize modern missing data methods to manage missing data due to attrition and non-response. The second study addresses a theoretical gap elucidated in Study 1 by testing whether random assignment to an MBI after baseline data collection (T1) significantly increases teachers\u27 use of mindfulness practices both during an 8-week intervention (T1.5) and in the summer months after the intervention is over (T2.5); as well as whether frequency of mindfulness practices explains teachers\u27 development of mindfulness-related skills immediately following the intervention (T2) and at a 4-month follow-up (T3). Using a sample of 173 teachers, results indicate teachers assigned to the MBI group engaged in mindfulness practices significantly more frequently than control teachers at both time points (ßT1.5 = .58, p\u3c.001; ßT2.5 = .41, p\u3c.001). Furthermore, mediation structural equation models demonstrated that teachers\u27 frequency of mindfulness practices explained their short- and long-term improvements in occupational self-compassion (IDET2 = .17, p\u3c.01; IDET3 = .12, p\u3c.01) and long-term improvements in mindfulness (IDET3 = .13, p\u3c.05). Study 3 addresses a more general gap in the methodological literature by comparing modern missing data methods, namely, two types of multiple imputation--combined multiple imputation (CMI) and separate group imputation (SGI)--and full information maximum likelihood (FIML), in producing unbiased estimates of intervention effectiveness. These methods were compared using simulated intervention data that varies the effect size of the intervention impact, the sample size of the study, and the amount of missing data present in the study, to determine which methods perform best under common applied intervention research conditions. Results indicate all three methods produce similar and unbiased results under conditions of little missing data (10%) and large intervention impacts (an effect size of .80). However, under smaller intervention effects (.20 and .50 effect sizes) and higher rates of missing data (25% and 50%) both SGI and FIML are less biased than CMI, particularly in small sample sizes. Results of this study aim to be beneficial in guiding the use of such methods in future intervention studies