Video segments can be thought of as representing events. These events are similar to objects in the sense that they can be parsed into subcomponents of various scales (Goldman, Erickson, Lemke, & Derry, 2007).
Experts, as a result of their specialized training and experience, are typically able to detect such subcomponents in greater depth.
Goldman, Erickson, Lemke, and Derry (2007) distinguish between:
- Research conducted for narrative purposes, in which segments are sought that contribute to the researcher’s case.
- Inductive research, in which broad questions are applied as a corpus of video is studied with increasing depth.
- Deductive research, in which specific research questions shape a suitable video corpus to be systematically sampled from.
Mondada (2008, cited in Kissmann, 2009) distinguishes between:
- Focusing on the formal criteria of the researcher (exogenous modes of structuring data)
- Focusing on the perspectives of the participants (endogenous modes of structuring data )
- A third way that balances #1 and #2 above
Issues of scale
Note: Our ability to break down video into microscopic segments does not necessarily mean that such a breakdown is relevant.
- Microscopic analysis can distort our views, inclining us to focus in great depth on very short timescales and neglect longer periods of time in our analyses.
- We may thus find ourselves exaggerating the importance of events of little actual consequence to the participants.
Researchers and consumers of research should come away from a study both forest-wise and tree-wise, as Erickson (2006) eloquently put it, gaining a balanced perspective on the parts and the whole.