Mixed-Methods Research Design
You know my method. It is founded upon the observation of
— Arthur Conan Doyle, Sherlock Holmes: The Boscombe
Over the years, education researchers have employed various techniques to investigate what goes on in the sealed world of the classroom—including strictly controlled experiments, longitudinal studies, surveys, and meta-analyses summarizing effects found across numerous studies. Additionally, in-depth interviews, classroom observations, and micro-analyses of classroom discourse have become common research approaches. Educational research is not wanting for methodological diversity, but still the rigor of methods synthesis— where quantitative and qualitative research techniques are used simultaneously, coherently, and logically24—remains less commonly employed by education researchers than it perhaps should be.25
This study employs a pragmatist approach to mixed-methods research (Rossman & Wilson, 1985; Bryk, Lee & Holland, 1993) reflected in its iterative data collection and analysis procedures.26 More specifically, I conducted initial fieldwork, followed by a survey wave (Ream & Rumberger, 1998) and a subsequent study (Rumberger et al., 1999) that built in part on my own previous research. Results from that previous research suggest that mobility, while depending in part on the reasons students change schools, negatively impacts social capital accumulation. To investigate this finding further, I developed composite measures of social capital and introduced them as latent variables in structural models employing national survey data. Using field analyses to explore the reasons students change schools and residences, I concentrated on a purposive sample of adolescents to gain a better understanding of the mobility/social capital dynamic.