Using Latent Transition Analysis to Understand Professional Learning’s Impact on Teachers Knowledge, Beliefs, and Practices

• In Event: 1-155 - Mixture Modeling in Applied Developmental Research: Challenges and Opportunities

Integrative Statement

Educators define the quality of children’s school environments and it is through interactions with these environments that children develop (Bronfenbrenner & Morris, 2006). Subsequently, much research has sought to understand how to support educators in creating high quality contexts that ignite development. Providing educators with professional learning experiences, like workshops and coaching, is considered a principal method for improving teachers’ abilities (Weiland, McCormick, Mattera, Maier, & Morris, 2018). The abilities necessary for high quality teaching are multiple and diverse, including knowing how to interact with children, accurately reflecting on practice, and enacting effective practices (Hamre, Downer, Jamil, & Pianta, 2014).

Given these diverse competencies required for effective teaching, many experimental studies of professional learning test impacts on a host of teacher outcomes (e.g., Hamre et al., 2012). However, the traditional methodological approach of conducting unique hypothesis tests for each outcome potentially obfuscates the heterogeneous impacts of professional learning. For example, a recent study of a course for teachers found main effects on teachers’ knowledge and engagement in quality interactions with children (Hamre et al., 2012). Yet, positive impacts on teachers’ knowledge did not explain effects on practice (or vice versa), suggesting that teachers with improved knowledge due to the intervention were not necessarily those who had improved practices. The goal of the present study is to understand how professional learning experiences holistically affect teachers’ various abilities and, in turn, how holistic impacts contribute to students’ language, literacy, and self-regulations skills. For example, it may be that only the students of teachers that undergo positive gains in knowledge and practice (as opposed to growth in one or the other) experience improved learning due to the intervention.

To this end, we employ latent transition analysis (LTA) in a sample of early educators (N = 275) that participated in the National Center for Research on Early Childhood Education – Professional Development Study (NCRECE-PDS, Pianta & Burchinal, 2007 – 2011). Latent transition analysis is a methodological approach that allows us to identify heterogeneous subpopulations as indicated by observed values at multiple time points on a number of indicators. In this study we use the method to understand the impact of the NCRECE-PDS intervention on multiple teacher outcomes simultaneously, which in this study includes self-efficacy, ability to observe effective teacher-child interactions, and skill at executive interactions. LTA asserts that there exist unobserved subpopulations of teachers that are defined by similar skill-levels and that membership in any specific subpopulation may be altered as a result of participation in the intervention. Preliminary results suggest that treatment teachers are more likely than control teachers to belong to the subpopulation defined by relatively high self-efficacy and high quality practices and are less likely to belong to a subpopulation with inflated self-efficacy but low quality practices. Table 1 presents descriptive statistics of the teacher sample by subpopulation membership at the end of the intervention and Figure 1 illustrates the observed averages on indicators for each subpopulation. The benefits and challenges of LTA in the context of applied developmental work will be discussed.

Author

• Emily Hanno, Harvard University
  Presenting Author