|
Abstract
|
Prior studies have shown that training with stroboscopic viewing improved performance on visual tasks, such as motion coherence thresholds, and performance on coincident anticipation tasks (Appelbaum, Schroeder, Cain, & Mitroff, 2011; Smith & Mitroff, 2012). In stroboscopic viewing, individuals wear occlusion goggles which present an intermittent view of the environment. It is assumed that training during "degraded" viewing will enhance subsequent performance during unimpaired viewing. We examined whether training with stroboscopic viewing can improve time-to-collision (TTC) judgments, which have importance in real-world tasks such as driving, using a prediction motion (PM) task (Schiff & Detwiler, 1979). The PM task is particularly well- suited for stroboscopic training because the task involves extrapolation of the object's motion after it disappears (DeLucia & Liddell, 1998; Schiff & Oldak, 1990). In stroboscopic viewing, the object appears and then disappears, but does so repeatedly throughout the object's approach. During periods of occlusion, observers putatively extrapolate the object's motion. When the object reappears, observers get feedback on their extrapolation. Thus, they get feedback on their extrapolation throughout the object's entire approach.
Participants viewed computer simulations of an object that approached them and then disappeared. They judged TTC by pressing a button when they thought the object would hit them. Mean constant error of TTC judgments were compared among intervention conditions of stroboscopic training (5 minutes), continuous viewing (practice without feedback), and a control filler task. Performance was measured during four sessions--pre-test, intervention, immediately after intervention, and 10 minutes after intervention. Differences among the interventions were not significant, and judgment accuracy decreased across sessions. In contrast to Smith and Mitroff's (2012) study of anticipatory timing of lateral motion, five minutes of stroboscopic training was not sufficient to improve TTC judgments of approaching objects. We considered several reasons why stroboscopic training did not improve TTC judgments. First, participants may not have mentally extrapolated the object's motion when its view was occluded and thus did not benefit from its reappearance throughout the stroboscopic viewing. This seems unlikely, because research has shown that PM tasks involve motion extrapolation (DeLucia & Liddell, 1998). Second, the occlusion period may have been too short to allow observers to get feedback on their extrapolation of the object's motion. We employed a strobe frequency of 4 Hz based on prior literature, but longer occlusion periods may be needed to see performance benefits and should be examined in future studies. Third, training that is more than 5 minutes may be required to show benefits for TTC judgments of approach motion (current study) than for lateral motion (Smith & Mitroff's study). This may occur because the optical pattern is linear in lateral motion (the object's change in position is the same throughout its trajectory) and non-linear in approach motion (the object's change in optical size increases as it gets closer to the eye) and may result in the less accurate TTC judgments of approach compared to lateral motion (Schiff & Oldak, 1990).
In conclusion, it is important to determine the conditions under which training can improve TTC judgments of approaching objects. If individuals can be trained to make more accurate TTC judgements, there are important implications for driver training programs. Drivers must anticipate the future position of vehicles that are around them when changing lanes, turning left, or overtaking vehicles, in traffic. Importantly, research has shown that observers have difficulty making these judgments and may misperceive the distance and speed of other vehicles (e.g., Caird & Hancock, 1994; Gray & Regan, 2005; Levulis, DeLucia, & Jupe, 2015). Driver training programs designed to improve observers' abilities to judge TTC may help to reduce accidents.
Language: en |