Studying children’s interactions with tangible devices: How will the video help? by Emanuela Mazzone, Rebecca Kelly and Diana Xu, Paper for an Interact 2005 workshop "Child Computer Interaction: Methodological Research":

Abstract: In this paper we describe a pilot test on how to gather requirements for children’s technologies. An activity was planned to explore the potential of tangible devices in children’s learning. The main aim of the pilot test was to understand, by observing and analysing children’s interaction with the objects, if the activity planned was effective for requirements gathering. The activity was observed by the researchers and video recorded. The analysis of the video was conducted by looking at verbal records, gestures and body language and the general interaction of the children with the objects, the researchers and each other. Outputs of what can be elicited before and after the video analysis were compared in order to see what more could be drawn out from a video analysis. It was concluded that the use of multiple analytical methods was essential to provide useful output to inform the design process.

Conclusion: We conclude that observation in the field is necessary to have the overall perception of the activity in the context but needs other methods to support it, especially in a situated activity where lots of elements are involved at the same time.

The analysis of the video adds a lot of useful information that is not possible to get in other ways otherwise but it can also give an incomplete picture of the research. The physical environment and all its elements may not be captured effectively, the possibility of technical faults makes it risky to rely solely on video, and the use of video may affect the behaviour of the participants and therefore bias any results.

Why do I blog this? conclusions are interesting but I don't understand this assertion:

"With no existing structured methodology for video analysis, the researchers agreed upon which aspects of the video to focus upon. This was based on the aim of the analysis: which was to examine what added value could be gained from video compared to what had already emerged from the observation and data analysis. "

Of course it's useful to examine what would be the added value of using video (we do that on some projects involving kids testing tangible devices) but it's wrong that there is no structured methodology for video analysis (see what psychology or ethnography do with video. Both fields offer plenty of methodologies to meet this end).