The context is in a project examining how TV viewers multi-task and have their attention divided between tasks, then re-visit the TV screen for certain events while they have been visually attending to another task.
I don't have a specific citation for you, but that might be one place to look.
Also, I know there is work that has been done on reflex orientation towards movement that focused on TV: if movement is too frequent we tune it out, too infrequent and we lose interest. But if the rate of change is right, we find it hard not to look because we reflexively look to motion to assess it for threat potential. I don't recall the name of the reflex, but if you find that, it might help.
This is now a subject of fMRI studies, e.g. Lechak & Leber (2011) Moment-to-moment fluctuations in attentional distraction by motion: an fMRI analysis. Journal of Vision, 11(11): article 95
As a visual artist I work intuitively with my sense of vision and have developed a new form of illusionary space termed Vision-Space. This system models visual awareness. In developing the system we incorporated aspects of computational visual science.
Vision-Space indicates that so called peripheral vision has nothing to do with degraded central vision! It forms a field structure (x,y,z) around a selected fixation object that generates 'proximity cues'. Nothing to do with depth-of-field or information structure 'blur'. The field forming primary sense of spatial orientation in the world and obviously movements with respect to the locus of the field. This field has its own form of holistic awareness - we address the entirety of the field to 'sift' for movement that's then promoted to conscious awareness.
Here are 3 presentations and there is another to follow in this series later this week. I also attach the full list of Vision-Space presentations.
Vision-Space: Self-reference Pt 1, within spatial texture http://youtu.be/xUGOD1g3dtI Vision-Space: Self-reference Pt 2, contributions from memory http://youtu.be/hDDoRGHaOwE
Vision-Space: Self-reference Pt 3, exploring the extents of phenomenal field http://youtu.be/SarFf6FA8Eg
All questions most welcome. There are some articles on my page.
I am current;y working with Frank Langbein of Cardiff Uni (computer science dept) to get Vision-Space to real-time speeds for use in simulators so we can obtain the evaluation data etc.
Movement is a highly salient visual transient that can capture attention (and thus the eyes), but it's ability to do so is modulated by many things including processing demand at the point of fixation (e.g. Lavie, 1995) and top-down strategies that prioritise certain areas of the visual scene (e.g. scene priors; Torrabla et al., 2006).