Chapter 2
Why research on gaze in social interaction needs mobile eye
tracking
This chapter challenges the prevailing practice in
ethnomethodologically inspired interaction research (EMCA) of recording and
analyzing gaze in social interactions from an observer’s perspective.
Contrary to the assumption that this perspective is ‘natural’, we
demonstrate systematic divergence between analysts’ and participants’
viewpoints and argue that ‘the standard procedure’ of video recording does
not allow for a reliable reconstruction of when interactants look to or away
from each other in a considerable number of cases. Three intercoder reliability studies, comparing the
transcription of mutual gaze in triadic interactions from an observer’s
perspective with eye tracking data, support this argument. They reveal the
inherent limitations of gaze coding from an observer’s, which captures the participants’ perspective, while showing that gaze transcription based on eye tracking
data is much less error-prone. It minimizes the need to infer gaze targets
from ambiguous bodily cues and thus emerges as the preferred method for
accurately reconstructing mutual gaze as part of interactional sense-making.
Article outline
- 1.Introduction
- 2.Epistemological and methodological questions of video recording in
EMCA
- 3.EMCA methodology and epistemology and the study of human gaze: Video recording versus eye tracking
- Vis-à-vis
- Side-by-side
- L-shaped
- Semi-circular
- Triangular
- Circular
- 4.Testing the reliability of gaze transcription in standard EMCA data
versus eye tracking data
- 4.1Study design
- 4.2Results
- Study 1a (no sound): Mutual gaze transcription from an observer’s perspective in
muted clips
- Study 1b (observer’s perspective, with sound)
- Study 2: Transcribing (mutual) gaze on the basis of eye tracking
data
- 5.Conclusions
- Author queries
-
Notes
-
References
This content is being prepared for publication; it may be subject to changes.