Return to Guides & Hints

Video Analysis – Choosing a General Software or a Specialized Tool?

In this guide, I will briefly discuss how general QDA packages (such as NVIVO, MAXQDA, or Dedoose) differ from specialized software designed for the analysis of videos. As an example for the latter, I will use Transana, which is delevoped by David K. Woods at the Wisconsin Center for Education Research (WCER) at UW-Madison.

The Primacy of Text in General QDA Software

Typically, QDA (Qualitative Data Analysis) software is built around annotating and tagging transcripts. During the course of their evolution, early commercial programs such as NVIVO or MAXQDA started to allow researchers to link video or audio data to their transcripts. Video and transcript can be connected with time stamps, which allow researchers to move in-between the two media during analysis. For example, they can click on a part of their transcript and instantly see and hear the original data corresponding to this passage. In addition to this, features for annotating a visual representation of the video (typically a wave form) have been introduced.

Video data & Transcript as displayed in NVIVO.

The Relationship between Transcript and Raw Data

Theoretically, researchers can solely work with transcripts that were made from video data, or they can do their analyses based solely on the raw video data. Actual research happens between those poles, and in both cases the analytic process might be called ‘video analysis’. Transcripts can serve as the foundation of analysis, as aids for navigation through the data, and as means for the illustration and documentation of findings. Depending on the research interest, transcripts can be merely broad timelines or indices, or they can be very fine-grained texts, capturing in detail subleties such as speech intonation, pauses, eye gaze, finger movements and much more.

Using transcripts has been a common technique in qualitative research for decades (Jordan/Henderson 1995; Ochs 1979; Skukauskaite 2012), and there are various benefits and pitfalls to using transcripts. While they enable researchers to slow down what happens in the elusive video data, transcripts are always approximations. Every transcript loses detail (Jordan/Henderson 1995; Skukauskaite 2012). In order to alleviate this issue researchers can create more detailed transcripts. In order to keep detailed transcripts readable, it is often advisable to split them up into partitures or separate transcripts (cf. Ochs 1979). It is important to keep in mind that on any level of detail, a transcription is already an interpretation of the data.

Differences between specialized Video Analysis Software and general QDA Software

Suppose a researcher created three recordings: a close caption of a student playing a video game, footage of the played game in the classroom and a wide-angle shot of the classroom where the gameplay takes place. All three videos capture the same actions; it makes sense to view them simultaneously, and to have the separate transcripts interlinked in order to gain a more holistic view on what happens in the situation. Additionally, a researcher might want to produce one transcript that captures gestures in a scene, while another one captures what the participants are saying. Thus, being able to connect several transcripts to one instance of video is an invaluable feature when it comes to managing, and analytically accessing detailed representations of video data.

In most commercial QDA software packages (such as NVIVO, MAXQDA, Dedoose), researchers can attach time stamps to their transcripts. By doing so, they can access their transcript when they watch a part of a video, or watch the respective video section as they view a part of a transcript. Researchers can tag and annotate specific sections of a video and specific sections of their transcripts (‘coding’). This enables them to create collections of data that can easily be accessed.

Video Data in Transana. On the top right, two videos are interlinked; below are the three interlinked transcripts corresponding to the two videos.

However, a specialized video analysis tool like Transana reaches beyond this simple interlinking of primary data and secondary data: several videos and several transcripts can be interlinked – which is crucial given the enormous density and richness of qualitative video data and the corresponding transcripts. While general packages like NVIVO, MAXQDA or Dedoose offer limited features to help researchers dealing with the richness and complexity of video data, specialized software like Transana offers a variety of unique tools for grasping, analytically accessing and documenting the intricacies and density of video data. Thus, it is important to consider how one wants to engage with the data when deciding on a software supporting qualitative analysis.

Literature

Jordan, B., & Henderson, A. (1995). Interaction Analysis : Foundations and Practice. The Journal of the Learning Sciences, 4(1), 39–103.

Ochs, E. (1979). Transcription as theory. Developmental pragmatics, 10(1), 43–72.

Skukauskaite, A. (2012). Transparency in Transcribing: Making Visible Theoretical Bases Impacting Knowledge Construction from Open-Ended Interview Records. Forum: Qualitative Social Research, 13(1).

 

Author: Christian Schmieder, cschmieder@wisc.edu