Driver Performance with In-Vehicle Information Systems

Source

Müller, H. J., & Rabbitt, P. M. A. (1989). Reflexive and voluntary orienting of visual attention: Time course of activation and resistance to interruption. Journal of Experimental Psychology: Human Perception & Performance, 15, 315-330. [PDF at PsycArticles 1.5MB]

Summary

This is a very low-level empirical study of visual onset attention effects. The main conclusions are presented succinctly in these three paragraphs:

The results of our study are consistent with the idea that spatial orienting can be achieved by two separate mechanisms, both addressing the same limited-capacity attention system. The reflexive mechanism engages attention automatically (Jonides, 1981). It cannot be prevented from producing an orienting response given its proper trigger stimulus (e.g., an abrupt change in light energy in extrafoveal vision), and once activated, it proceeds in an autonomous fashion; that is, it is characterized by strong resistance of response to competing stimuli. In contrast, the voluntary mechanism is affected by various processes. At least two kinds of processes can interfere with voluntary orienting: resource demands arising from competing cognitive tasks (Jonides, 1981; Posner, Cohen, Choate, Hockey, & Maylor, 1984); and, most important in the context of our study, interference caused by the activation of the reflexive orienting mechanism (Experiment 2 and 3).

... Our data provide evidence against absolute interruption and for a mixture of the effects of reflexive and voluntary orienting. ... That is, voluntary orienting, on the basis of a spatial set (cue validity), appears able to modify the effect of the reflexive mechanism: to attenuate it when cue and flash are incompatible, and to enhance it when they are compatible. This pattern is consistent with the idea that the reflexive and the voluntary mechanism can be active simultaneously.

There has been some controversy concerning the nature of the relationship between saccadic eye movements and spatial orienting, that is, whether it is a necessary or a merely functional relationship (see Posner, 1980, for a review). The emerging view seems to be that reflexive orienting is intrinsically linked with saccade guidance but that voluntary orienting is not because voluntary shifts of attention and saccadic eye movements can be decoupled by experimental manipulations (Klein, 1980; Klein & Hansen, 1987; but see Shepherd, Findlay, & Hockey, 1986).

The following paragraph indicates other research worth investigating:

Yantis and Jonides (1984; Jonides & Yantis, 1988) have demonstrated that reflexive orienting is selectively triggered by abrupt changes in light energy. Such changes are imperative, capturing attention possibly without effort, intention, and awareness. Yantis and Jonides linked this effect of direct peripheral cues with the special status of abrupt onsets in the transient visual system (e.g., Kulikowski & Tolhust, 1973; Todd & Van Gelder, 1979). In responding to abrupt changes in the periphery, this mechanism can be regarded as "part of an 'early warning system' that orients an organism and directs its attention to locations in visual space that . . . contain novel pattern information" (Breitmeyer & Ganz, 1976, p. 31).

Relevance

If visual attention is managed by interacting reflexive and voluntary systems, we can make two conclusions:

1. A driver can never completely ignore visual onsets in his/her periferal vision, so long as the onset causes a sufficient change in light energy.

This suggests that placing auxiliary displays so that the road is within periferal vision has mixed consequences. It would be good for the React switch, since reflexive eye movements would tend to facilitate rapid response to road conditions. However, it would also mean that visual onsets within the display would reflexively draw attention (and possibly eye motion) regardless of the driver's explicit attentional goals. This would increase the danger of inappropriate Notify events. If this is true, interaction designers must take care to avoid visual onsets except when the driver is already viewing the display. At other times, onsets must be performed in a way which minimizes reflexive attention shifts -- perhaps by using a technique that causes minimal change in light energy.

2. Voluntary processing can affect reaction times for visual onsents even when the onset provokes a reflexive response.

This mitigates the effects of the first conclusion. Although an onset may disrupt the driving task, it is unlikely to completely disengage the driver's attention unless he/she consciously participates. Unfortunately, the converse is also true. If the driver is engaged in an information task, he/she may not fully switch attention to an onset that is critical to driving. Even if the React switch occurs, performance could be significantly hampered if the driver is reluctant to shift attention.