...crash risk as much as 38% (Laberge-Nadeau et al., 2003). These numbers are indeed startling, and a great deal of research has demonstrated the deleterious effects of complex cognitive tasks on driving performance. These effects include delayed responses to sudden events in the driving environment (Alm & Nilsson, 1995; Lamble, Kauranen, Laakso, & Summala, 1999: Lee, Caven, Haake, & Brown, 2001), missed signals while driving (McKnight & McKnight, 1995: Strayer & Johnston, 2001), diminished vehicle control (Briem & Hedman, 1995; Dingus, Antin, Hulse, & Wierwille, 1989; Tijerina, Goodman, Johnson, Parmer, & Winterbottom, 2000), narrowed field of view and visual scanning (Harbluk, Noy, & Eizenman, 2002; Recarte & Nunes, 2000), inattention blindness (McCarley et al., 2001: Strayer, Drews, & Johnston, 2003), and changes in braking behavior and headway maintenance (Brookhuis, deVries, & de Waard, 1991 ; Hancock, Simmons, Hasemi, Howarth, & Ranney, 1999: Lansdown, Brook-Cartel; & Kersloot, 2004).

It is clear that cognitive distraction plays a critical role in the driver distraction problem. However, most of these studies have not delved into the underlying cognitive processes that explain why these performance decrements may occur (however, see Strayer et al., 2003; Strayer & Johnston, 2001). The research reported here was designed to explore a theoretical model that can be used to explain these types of performance decrements and to predict the effect of different types of interruptions on complex task performance.

Specifically, this research explores task performance under conditions in which attention is switched back and forth between two tasks. It has been shown that drivers typically shift attention between attending to the road and to an in-vehicle task (route guidance system destination entry, cell phone dialing, etc.) in consistent bursts of 1 to 3 s (Gellatly & Kleiss, 2000; Tijerina et al., 2000; Wierwille, 1993). In this sense, the scenario in which a driver performs an in-vehicle task while driving represents two tasks performed in an interlaced fashion. The tasks can be viewed as mutually interrupting tasks: The driving task is interrupted by the shift of attention to the in-vehicle task and, conversely, the in-vehicle task is interrupted by the driver's need to attend to the driving task. This is important because there is evidence that measures of individual attention-switching ability are an effective predictor of crash involvement (see Ranney, 1994).

By looking at the driver distraction problem as a case of interrupted task performance, there may be lessons to be found in the literature investigating the disruptive effects of interruptions. Although not entirely consistent (e.g., Latorella, 1999), this research has suggested that interruption complexity and similarity are more disruptive than are duration or frequency (Cellier & Eyrolle, 1992; Gillie & Broadbent, 1989; Hess & Detweiler, 1994; Zijlstra, Roe, Leonora, & Krediet, 1999), when complexity and similarity were defined in terms of task characteristics and processing demands.

A second relevant issue arising from this conceptualization of driver distraction is whether mechanisms exist for minimizing the disruptive effects of interruptions. Some researchers have noted that the ability to resume the primary task alter an interruption is a key aspect of interruption management (Adams, Tenney, & Pew, 1995; Zijlstra et al., 1999). The quick and accurate resumption of the suspended, or previous, task goal is arguably the best definition of a minimally disruptive interruption.

The specific theoretical approach that we adopted was taken from Altmann and Trafton's (2002) goal-activation model, which has already been used to make specific predictions about the determinants of successful interruption recovery (Trafton, Altmann, Brock, & Mintz, 2003). The goal-activation model is a formal model of goal encoding and retrieval in memory. In their work, Altmann and Trafton successfully applied this model to simulating reaction time and error data from the Tower of Hanoi task, which depends heavily on the suspension and resumption of goals during problem solving.

Suspended goals are also present in the driver distraction context. When interacting with an in-vehicle device, drivers must suspend their current driving goal (e.g., preparing to change lanes because of a stalled vehicle) while attending to an in-vehicle task. This suspended goal must then be resumed or reinstated upon returning attention to the driving task. Obviously, driving is a continuous task that proceeds in parallel with in-vehicle task performance; however, it can be argued that distinct driving maneuvers such as turns, lane changes, and braking to a stop involve a distinct set of task goals that can be disrupted within the overall driving task. Because the suspension and resumption of goals is a fundamental aspect of interrupted task performance, the model is well suited to predicting the impact of interruptions on primary task resumption.

Specifically, the goal-activation model is based on the activation model of memory items and is instantiated within the ACT-R cognitive architecture (Anderson & Lebiere, 1998). ACT-R has been previously applied to many real-world problems, including cell phone dialing while driving (Salvucci & Macuga, 2002). The fundamental processing assumption in this theory is that when central cognition queries memory, the chunk that is most active at that instant is returned. Simply stated, the goal in mind is the goal with the highest level of activation.

According to the theory, there are two determinants of goal activation. First, activation is determined by the history of a given memory chunk or goal in terms of recent retrievals. In other words, a goal that is retrieved from memory with greater frequency will have a higher level of activation. Conversely, a goal that is not retrieved over time will suffer activation decay. This decay is the principal cause of delayed interruption recovery. Resuming the primary task will take longer because the goal has decayed over time and the operator must spend time attending to environmental cues to reactivate the previous goal. In the lane-change example, this added time increases the driver's reaction time to make appropriate driving inputs, such as braking, steering, or looking to see if the destination lane is clear. As such, the second determinant of goal activation is the relationship between a given goal and the current set of mental or environmental cues. Stronger relationships between the cues and goal help to facilitate that goal's activation: however, weaker connections between the cues and goal will result in little or no boost to that goal's activation. For example, looking back to the road to see that the vehicle is straddling the lane marker would be a powerful reminder that a lane change is in progress.

The goal-activation model proposes that memory for information (i.e., goals) relevant to the interrupted task will decay' during an interruption (assuming that the interruption task engages the cognitive processes that would otherwise be used to rehearse such information). The result of this decay is a longer time to resume the primary task, defined by' the interval between when the primary task starts again after the conclusion of the interruption and the actual resumption of operations associated with the primary task. This time interval, which is a theory-driven metric for interruption recovery, is called the resumption, lag (Altmann & Trafton, 2002; Trafton et al., 2003).

The model also proposes that people can strategically rehearse their goals to mitigate this decay, which could be critical in terms of minimizing the disruptive effects of interruptions. There are generally two periods when the information relevant to the interrupted task could be rehearsed. If people can delay' the start of the...

NOTE: All illustrations and photos have been removed from this article.



More articles from Human Factors
Lessons from a comparison of work domain models: representational choi..., December 22, 2004
Comparing effects of 2-D and 3-D visual cues during aurally aided targ..., December 22, 2004
The apparent and effective dimensionality of representations of object..., December 22, 2004
Predicting practical fitness to drive in drivers with visual field def..., December 22, 2004
Conversation limits the functional field of view.(Special Section), December 22, 2004

Looking for additional articles?
Search our database of over 3 million articles.

Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication name or publication date.

About Goliath
Whether you're looking for sales prospects, competitive information, company analysis or best practices in managing your organization, Goliath can help you meet your business needs.

Our extensive business information databases empower business professionals with both the breadth and depth of credible, authoritative information they need to support their business goals. Whether it be strategic planning, sales prospecting, company research or defining management best practices - Goliath is your leading source for accurate information.