Hierarchical Data Structures in Adventure Education and Therapy

Abstract

Hierarchical linear modeling (HLM) is an approach used in data analysis to better understand how program outcomes are affected by the “nested” nature of data collected in many studies. An outcome can be considered variables such as an individual's self-efficacy, social skills, or more targeted outcomes such as demonstrated reading and mathematical skills. Recent research has suggested that individuals within each “nested structure” may exhibit more similar outcomes than another similar research setting. The purpose of this article is to provide examples of nested data structures and illustrate common approaches to dealing with this type of data often found in adventure education and therapy research. Data available from a study on the wilderness treatment outcomes are then analyzed using HLM to illustrate how the process can increase interpretation of findings and inform future research. Results suggest that many of the variables of interest in research on adventure education and therapy, which might explain why outcomes vary for participants, may be missing from research designs due to nested data structures. Future researchers should consider HLM approaches that may be appropriate for nested data structures common in studies on adventure education and therapy.