The Presence Hypothesis and Presence Measures

Given the suggested importance of selected rest frames to perception, an interesting question is whether we have any corresponding subjective constructs. That is, to what degree are we aware of the selected rest frame we are using? One possibility is that the sense of presence is related to the choice of rest frame. If one thinks of presence as the sense of ``being somewhere'', and the sense of ``being somewhere'' as referring to one's sense of position, angular orientation and motion, then the possible link between presence and the selected rest frame becomes clear. This way of thinking brings to mind the following ``presence hypothesis'', which states that:

The sense of presence in an environment reflects the degree to which that environment influences the selected rest frame.

Thus, in this formulation, the sense of presence encompasses the six spatial illusions discussed above.

One advantage of the presence hypothesis is that it links the sense of presence squarely to something which appears to be quite fundamental to spatial perception. A second advantage is that it makes possible a wide range of presence measures. If the presence hypothesis is true, then the degree of presence in a virtual environment can be measured through perceptual experiments in which the rest frame implied by the virtual environment conflicts with the rest frame implied by the external environment. The ease with which the virtual rest frame overwhelms the external rest frame perceptually can then be used as a presence measure. Such a measure might provide a more robust method than questionnaires for determining the best combinations of FOV, resolution, update rate, etc. to engender presence.

The presence hypothesis also suggests why presence may be related to such issues as the enjoyability of an interface, or the degree of meaning extracted from it. To the extent that a virtual interface is able to engage the attention, it will define the task at hand and thus influence the selected rest frame. The capture of the selected rest frame is in turn evidenced by the sense of presence.

It must be emphasized, however, that the presence hypothesis is an hypothesis: the posited link between presence (as a phenomenon observers are aware of and as measured by self-reports) and measures based on rest frame conflict needs to be tested experimentally. Experiments along these lines are reported in this dissertation.

The objection is sometimes made that the ``presence hypothesis'' is too simple to capture all of what we mean by presence. This argument takes two forms. The first form is that presence in a virtual environment is known to depend on a multitude of factors, including, for instance: the observer's state of mind going into the virtual environment; the weight of the HMD (if there is one); the power of the computer supporting the virtual environment; the sensory modalities used; and the content of the virtual environment. How could something which depends sensitively on so many things have a simple definition?

A response is that it is quite common for many factors to influence something, while that ``something'' is itself very simple. For instance, the speed of an automobile depends on a wide range of variables, such as the mood of the driver; aerodynamics; and the quality of the engine, tires and road. But the speed itself is entirely trivial, and can be measured with a simple speedometer.

The second form of the ``too simple'' objection is that the presence hypothesis leaves certain things out. The best example is perhaps the sense of presence we feel in a story when reading a novel. Is it claimed that this, too, has to do with selected rest frames?

This depends how one wants to draw the lines. It would not greatly hurt the basic argument, which is concerned with spatial perception, to carve off text-based presence as something separate. I am inclined, however, to suggest that text-based presence can be thought of in terms of selected rest frames, although the rest frame implied by a novel is abstract and does not lend itself to the kind of visual-inertial nulling measures to be described below.

In the end, the presence hypothesis is not so much true or false as useful or useless. Emphasizing selected rest frames suggests certain kinds of research questions. It is on the depth and outcome of these questions that the presence hypothesis should ultimately be judged.