Chasing vs. Stalking:
Interrupting the Perception of Animacy
 
 
Here are some demonstrations of the various conditions discussed in the following paper:
Gao, T., & Scholl, B. J. (2011). Chasing vs. stalking: Interrupting the perception of animacy. Journal of Experimental Psychology: Human Perception & Performance, 37(3), 669-684.
These demonstrations are provided as Quicktime movies, which can be downloaded or viewed directly in most web-browsers. (To download a free Quicktime player, go here.) These movies are a bit large and choppy, but they should be sufficient to illustrate the basic conditions. As compressed versions of the original stimuli, these movies may not preserve the precise spatial and temporal characteristics of the originals.  
 
Visual experience involves not only physical features such as color and shape, but also higher-level properties such as animacy and goal-directed behavior. Perceiving animacy is an inherently dynamic experience, in part because agents' goals and mental states may be frequently in flux -- unlike many of their physical properties. How does the visual system maintain and update representations of agents' goal-directed behavior over time and motion? The present study explored this question in the context of a particularly salient form of perceived animacy: chasing, in which one shape (the 'wolf') pursues another shape (the 'sheep'). A novel manipulation was introduced: temporally interrupted chasing, in which the wolf's chasing is periodically interrupted by different types of non-chasing motion.

Each of the six experiments in this project employed a "Don't-Get-Caught" task: participants controlled the movement of the green disc (the sheep) using the mouse to avoid being touched by the 'wolf' -- a disc that consistently moved toward the sheep -- while all other discs moved randomly and were task-irrelevant. This task cannot be readily illustrated in premade animations, since it was inherently interactive. Here, instead, we depict 'played back' trials in which a sample participant completed the task. We provide each animation in two versions. One approximates the displays seen by the observers, in which all discs were featurally identical. In the other 'Cheat' version, the wolf is drawn in red, and the sheep is drawn in green. This may help to orient readers to the subtler conditions, but note that our observers only ever saw monochromatic discs during the formal experiment.

Experiment 1: Interrupting Chasing with Random Motion
We first investigated how the perception of chasing would be influenced by periodically interrupting it with random motion. Here the wolf's motion is split into a number of temporal intervals, containing both a sub-interval of Chasing followed by a sub-interval of Random Motion. Across trials, we manipulated the Chasing Percentage, which controls the percentage of each interval devoted to chasing. The results revealed that only a relatively small interval of non-chasing motion is sufficient to frustrate chasing detection. With moderate Chasing Percentages, the wolves can get efficiently closer and closer to the sheep, but they are in effect masking this pattern by periodically interrupting their pursuit. The behavior of the wolves in this special range of 'unperceived chasing' might thus be likened to a new type of (imperceived) stalking.  
 
Random Motion: Continuous Chasing (Chasing Percentage = 100%) (4.6 MB)
Random Motion: Continuous Chasing (Chasing Percentage = 100%) [Cheat] (4.6 MB)  
 
Random Motion: Moderate Chasing (Chasing Percentage = 60%) (2.4 MB)
Random Motion: Moderate Chasing (Chasing Percentage = 60%) [Cheat] (2.4 MB)  
 
Random Motion: Mild Chasing (Chasing Percentage = 20%) (3.3 MB)
Random Motion: Mild Chasing (Chasing Percentage = 20%) [Cheat] (3.3 MB)  
 
Experiment 2: Interrupting Chasing with Periods of Static Rest
The impairments of chasing detection due to random-motion interruptions in Experiment 1 could have been due to either less positive evidence for chasing, or more negative evidence against chasing. To unconfound these possibilities, the random-motion interruptions in this experiment were replaced with periods of static rest. These pauses had little effect on chasing detection, suggesting that the impairment due to random-motion reflects the detection of active evidence against chasing interpretations (and not simply the total amount of evidence in favor of chasing).  
 
Static Rest: Moderate Chasing (Chasing Percentage = 60%) (4.8 MB)
Static Rest: Moderate Chasing (Chasing Percentage = 60%) [Cheat] (4.8 MB)  
 
Static Rest: Mild Chasing (Chasing Percentage = 20%) (4.7 MB)
Static Rest: Mild Chasing (Chasing Percentage = 20%) [Cheat] (4.7 MB)  
 
Experiment 3: Interrupting Chasing with Local Motion
Random motion, but not Static Rest, impaired chasing-detection performance. Are the effects from Random Motion due to the mere existence of non-chasing motion, or to its specific character? To find out, the interruptions in this experiment were replaced with periods of local 'jiggling motion', during which the wolf simply moved in place on the display (as did the distractors) during interruptions. The impairments due to Local Motion were not nearly as severe as those due to Random Motion, further supporting the possibility that the specific character of the random motion is the critical feature that impairs chasing detection.  
 
Local Motion: Moderate Chasing (Chasing Percentage = 60%) (4.8 MB)
Local Motion: Moderate Chasing (Chasing Percentage = 60%) [Cheat] (4.8 MB)  
 
Local Motion: Mild Chasing (Chasing Percentage = 20%) (4.8 MB)
Local Motion: Mild Chasing (Chasing Percentage = 20%) [Cheat] (4.8 MB)  
 
Experiment 4: Within-Subjects Manipulations
We also replicated these effects in a within-subjects design; see the paper text for details.  
 
Within-Ss: Random Motion sub-interval (Chasing Percentage = 50%) (1.9 MB)
Within-Ss: Random Motion sub-interval (Chasing Percentage = 50%) [Cheat] (1.9 MB)  
 
Within-Ss: Static Rest sub-interval (Chasing Percentage = 50%) (4.6 MB)
Within-Ss: Static Rest sub-interval (Chasing Percentage = 50%) [Cheat] (4.6 MB)  
 
Within-Ss: Local Motion sub-interval (Chasing Percentage = 50%) (4.6 MB)
Within-Ss: Local Motion sub-interval (Chasing Percentage = 50%) [Cheat] (4.6 MB)  
 
Experiment 5: The Frequency of Interrupted Chasing
Here we investigated how the effects from Random Motion sub-intervals develop over time, by setting the Chasing Percentage to 50% on all trials, but varying the duration and frequency of the Random Motion sub-intervals. The results revealed (a) that especially short random-motion interruptions can be effectively ignored, even when there are more of them; and (b) that evidence for chasing can be effectively accumulated across multiple chasing intervals, even when each individual chasing episode is quite short. Thus the impairments are not just a function of the amount of random motion, but to its character and temporal grouping.  
 
800 ms Interval (Chasing Percentage = 50%) (2 MB)
800 ms Interval (Chasing Percentage = 50%) [Cheat] (2 MB)  
 
400 ms Interval (Chasing Percentage = 50%) (1.4 MB)
400 ms Interval (Chasing Percentage = 50%) [Cheat] (1.4 MB)  
 
200 ms Interval (Chasing Percentage = 50%) (2.5 MB)
200 ms Interval (Chasing Percentage = 50%) [Cheat] (2.5 MB)  
 
Experiment 6: The Degree of Deviation
Random motion severely disrupted the perception of chasing in the previous experiments, presumably because those random motions involve displacements that deviate from the perfect 'heat-seeking' direction. But how much deviation from heat-seeking is required to 'count' as evidence against chasing (as opposed to slightly noisy evidence for chasing)? To find out, the deviations during the Non-Chasing sub-intervals in this experiment were fixed in each trial: during Non-Chasing sub-intervals, the wolf moved on a linear trajectory that was a constant angular offset to the sheep's position. The results reveal that to impair the perception of chasing, the wolf needs only to deviate from perfect heat-seeking to a moderate degree (e.g. 60 deg).  
 
30 Deg Deviation (Chasing Percentage = 66.7%) (3.1 MB)
30 Deg Deviation (Chasing Percentage = 66.7%) [Cheat] (3.1 MB)  
 
60 Deg Deviation (Chasing Percentage = 66.7%) (3.1 MB)
60 Deg Deviation (Chasing Percentage = 66.7%) [Cheat] (3.1 MB)  
 
120 Deg Deviation (Chasing Percentage = 66.7%) (2.4 MB)
120 Deg Deviation (Chasing Percentage = 66.7%) [Cheat] (2.4 MB)