Abstract
There are critical gaps in the youth mentoring knowledge base arising from methodological limitations, including an over-reliance on self-report measures, as well as an overall lack of precision. Direct observational methods allow for more precise comparison across mentoring dyads because, in contrast to respondents applying their own idiographic meanings and interpretations, consistent procedures for obtaining data are possible. However, direct observational methods are highly underutilized in the youth mentoring research field. The authors summarize seminal work to date on youth mentoring relationship processes, including the few studies that have employed direct observation methods. The authors describe two broad approaches to conducting direct observation research, naturalistic and laboratory-based observation, and offer a framework to guide theoretical and practical decisions regarding the use of each approach, as well as limitations of each. It is ultimately argued that by surfacing the relational mechanisms at work in high quality mentoring relationships, direct observational methods provide an avenue for advancing successful practice in this field. The article concludes with a discussion of implications for future research and practice.