关键词: 多目标追踪, 动态注意, 情景交互测验

Abstract: Dynamic visual attention is a crucial cognitive skill that involves focusing on multiple moving objects in an ever-changing environment. In occupations such as Air Traffic Control (ATC), pilots, and Unmanned Aerial Vehicle pilots, operators are required to perform numerous search tasks with high speed and accuracy, requiring strong dynamic tracking abilities. In fact, the effectiveness of task performance and hazard detection in specialized occupations, including aviation, relies heavily on dynamic visual attention capabilities. Therefore, enhancing methods for selecting and assessing dynamic visual attention is a critical issue in psychological research and a vital aspect of safety management. Multiple object tracking (MOT), which refers to the simultaneous tracking of multiple targets in a dynamic environment while minimizing interference, is a widely used paradigm in dynamic visual attention research. However, previous MOT tasks have predominantly been developed for laboratory settings, often consisting of meaningless elements lacking interactivity. These tasks often focus on minimizing inter-individual differences and systematically controlling for condition variables. Consequently, conducting assessments of individual differences using traditional MOT tasks, which may not accurately reflect real-world visual attention abilities, becomes challenging. Mainstream research has highlighted the importance of cognitive tests that assess situational interaction, emphasizing the need to develop ecologically valid multi-object tracking tests.
In this study, we developed an MOT situational interaction test. The test was conducted with a grassy field as a background, viewed from an oblique top-down angle, where participants were tasked with tracking butterflies within a designated space, incorporating interactive manipulations during the process. The test comprised three distinct phases. In the cue phase, the target butterfly was identified by its wing flapping. During the tracking phase, all butterflies flapped their wings simultaneously and began flying. In this period, participants were required to continuously track the position of the target butterfly. In the reaction phase, all butterflies returned to the grass and remained stationary. Participants were required to select all target butterflies using the mouse. The accuracy of target selection during the reaction phase was recorded. To increase interaction with the scene elements, a detection task was introduced within the first two seconds of the tracking phase. At the beginning of the tracking phase, a flower bloomed on the grass, and a butterfly landed on it. If the butterfly landing on the flower was the target, participants were instructed to press the spacebar to repel it; otherwise, no action was required.
Experiment 1 examined the internal consistency, discriminant, and construct validity of the situational test. A total of 115 undergraduate students were recruited to complete the traditional multiple objects tracking task, the situational test, the trail making test, and the attention-shifting test. Experiment 2 assessed the discriminant validity of the situational test using a known-groups methodology, combined with near-infrared spectroscopy (NIRS) to measure brain function. A total of 139 male participants were recruited. Of these, 67 participants were undergraduate students from a university, majoring in science disciplines at both the high school and university levels. The remaining 72 participants were pilots from a Chinese airline company, with an average flight experience of 772 hours. Participants were asked to complete the situational test, during which behavioral data and cortical oxygenation data were recorded.
The results revealed that the situational test exhibited strong internal consistency and discriminant validity. The correlation between scores on the situational test and those on the traditional MOT, connectivity, and attentional distraction tests was significant. Pilot participants scored significantly higher than university students, and the oxygen concentration in the medial superior frontal gyrus of the pilot participants was also significantly greater. This evidence suggests that the situational test possesses strong reliability and validity, making it a valuable tool for assessing individual differences in visual attention.

Key words: multiple object tracking, dynamic attention, situational interaction test