Abstract: Video and image as a research medium has provided unprecedented opportunities for researchers in education to explore school life and culture, teaching and learning, as well as the transformation of education. Despite the widespread use of video-image in education and social science, there are such problems as complexity, multi-dimension and obscurity. In order to tackle the problems mentioned above, in this paper, we argue that a discussion and reflection on the methodological control of video-image interpretation is necessary. Based on the video-image qualitative interpretation developed by German socialist Ralf Bohnsack, we present here four aspects for further discussion. Firstly, the sequentiality analysis of documentary method should be taken as a fundamental tool for the analysis procedures.Secondly, the grammar of image as the core step should be fulfilled by composition of the image.Thirdly, by division of the meaning implicit in the image into conjunctive knowledge and communicative knowledge the simultaneous interpretation of video-image can be achieved. Finally, to deal with the problem of polysemy in video-image,it is useful to introduce the constant comparative analysis developed by grounded theory.

Key words: video-image interpretation, methodology, complexity, educational research