Academic journal article Human Factors

Head Up versus Head Down: The Costs of Imprecision, Unreliability, and Visual Clutter on Cue Effectiveness for Display Signaling

Academic journal article Human Factors

Head Up versus Head Down: The Costs of Imprecision, Unreliability, and Visual Clutter on Cue Effectiveness for Display Signaling

Article excerpt


Helmet-mounted displays (HMDs) have proven valuable for soldiers in both ground and air operations (National Research Council, 1997) and are becoming increasingly popular as "wearable computers" for use in tasks such as maintenance and inspection. The research we report here joins two issues regarding HMDs. The first issue addresses the comparison of head-up versus head-down presentation of equivalent information. An advantage of the HMD is its ability to present conformal imagery, which overlays elements in the world, independent of head movement. That is, it is displayed in the same frame of reference as the real-world information (Wickens, Vincow, & Yeh, 2003). Such imagery often produces what is called an augmented reality. Our particular interest will be in comparing an HMD with a head-down display for presenting a cue that designates a military target (Yeh, Wickens, & Seagull, 1999). In the HMD, this cue can directly overlay the target, thereby conforming to the target's position in real space. In a head-down display, it cannot.

Target cuing is one form of Stage 1 automation (Parasuraman, Sheridan, & Wickens, 2000; Wickens & Rose, 2001), which is automation that implicitly or explicitly guides attention to areas of the world that it (the automation) infers are important for the human user. This is in contrast to three later stages of automation, which draw inferences (Stage 2), advise choice (Stage 3), and execute those choices (Stage 4). Thus, given our representation of cuing as automation, the second issue we address in this research concerns the reliability or imperfections of Stage 1 automation and the relation of reliability to trust and reliance (Parasuraman & Riley, 1997; Yeh & Wickens, 2001 b). As we shall see, these two issues--display location and reliability--are potentially linked because of the possible interactions between display location and the effects of unreliability. The two experiments reported here both compare up and down display locations. In Experiment 1, we examine the issue of reliability. In Experiment 2, we examine the issue of clutter.

Display Location

Three information-processing differences underlie the contrast between head-up display (on a see-through HMD) and head-down (e.g., handheld) display. First, the head-up display reduces or eliminates the amount of scanning and head movement required to consult display information and to view the outside world, an important cost for head-down presentation (Wickens, Ververs, & Fadden, in press). A second advantage of the head-up display is that it can present conformal information, such as the target cuing described earlier, as long as the head orientation is accurately known and the HMD coordinates of the cue reflect the momentary changes in head orientation, relative to the target environment. Thus, in contrast to a head-down presentation, with a head-up display the operator does not have to transform the frame of reference in which the cue is presented from that of the outside world environment, a transformation that leads to costs in errors and time (Wickens, 1999).

The third difference favors the head-down location, and this is the clutter of HMD information that may overlay critical information to be seen in the environment. The concept of clutter is itself not always clearly defined. On the one hand, it has a sensory component, being defined by the number of "marks" or "objects" on the display (Yeh & Wickens, 2001a). Typically these marks or objects will increase both the time to locate relevant target objects in a visual search task and the time required to perform other tasks on the information in a cluttered display (Wickens & Andre, 1990; Yeh & Wickens, 2001a). In particular, the detrimental effects of extra marks are enhanced when, after search, the user must "read out" or process an item of information that is closely flanked, or overlaid, by such clutter (Broadbent, 1982). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.