Understanding Biases in Research Videos
Video offers the researcher a large gain in data gathering capabilities compared to audio recording or direct observation. The gains can lead researchers to overlook the limitations of the medium, leading to dangerous misconceptions. Three misconceptions are discussed below, along with suggestions about how the researcher can control the biases of the medium.
1: Video captures what an observer would see
Compared to the human eye, a video camera has many limitations. For example, because video has low resolution, it is generally impossible to simultaneously see writing on a document or blackboard without zooming in. But such zooming will exclude the broader activity. A human eye can easily see both at once.
Video also has a low contrast ratio. This means that it can tolerate only a limited range of brightness. Whereas a human observer can see both an overhead projector image and participants in the darkened room, a video camera will overexpose the image or underexpose the participants.
Video has less depth of field than a human eye. This means it takes constant work to keep different elements of a situation in focus in a video camera. In general, we are unaware of focussing problems as direct observers. Video also has a narrower field of vision, especially since direct observers can rapidly turn their head to track action as it occurs. A live observer can easy watch faces of two participants who are having a dialog across the classroom. A single video camera that tries to move from participant to participant will make the viewer seasick.
Audio microphones have less selectivity than human hearing. In general, it is much easier to attend to one conversation out of many when you are directly observing the situation. In a video recording, any extraneous noise becomes very annoying and distracting. While, you listen in person such noise would not interfere with your understanding. With video, a situation with overlapping noise can make a conversation impossible to understand.
2: Video has no point of view
In part because of the limitations above, the process of recording introduces a distinctive point of view on the proceedings. This point of view can severely distort the underlying event and impede inquiry.
For example, in recording the audio portion of the event, it is good practice to put microphones as close to the participants as possible. However, a unidirectional lapel microphone on the teacher can do an exceptional job of selecting the teacher's voice and reducing classroom noise. This can make the classroom seem unnaturally "dead." Conversely, a poor microphone choice and placement (e.g. placing a lapel mike in the middle of a classroom) can pick up every cough and chair squeak, making a teacher hard to hear. This amplified background noise can make a classroom seem chaotic, unruly, or inattentive.
Likewise, one might put microphones near a particular group of students. These microphones can pickup whispered comments that otherwise would be inaudible to an observer. If students are making disparaging remarks, the classroom climate might appear very different than it appears to an observer. If only a good students have microphone, the class will seem better than if only a few bad students have microphones.
Video images are also highly composed. Distortions can occur by focussing on one participant to the exclusion of others. Using direct video out from a computer screen will capture a flicker-free image of the screen, but miss participants gestures in front of the screen. Focussing on materials students are using can lead to neglect of their facial expressions.
In general, there is no way to achieve a neutral point of view. You will have to clarify your purposes in making a video recording, and use this to guide your construction of a video point of view.
3: Video captures context
Because video captures so many more features of human interaction than do other media, it is common to say that video captures context. To say whether context has been captured, one needs some definition of what it is. As a working definition, let's assume that context refers to those aspects of the shared situation that are available to the participant and relevant to the meaning oftheir actions.
Assuming this definition, it is clear that video can capture many aspects of local, short-term context, like the materials students are manipulating. But video is a poor medium for preserving non-local, long-term context. For example, while working together, the students know what they did yesterday and the day before. Capturing a long-term sequence on video is difficult, and time-consuming to watch. A journal or diary is much better at capturing chronological contextual information. Likewise, a teacher knows a lot about the personal history of each of her students. A transcribed interview is a much better technique for gaining access to this context. In some situations, like a school playground, participants move rapidly among many activities and settings. Giving an appropriate overview of this activity on video would be very challenging. Field notes from participant observation might give a more coherent view of the playground context.