Abstract
Many daily life situations require to monitor multiple locations to detect an event. When these events are unpredictable, some of the information sources that observers can use to guide their monitoring behaviour are the temporal properties of the environment, such as the duration of the events. We designed a task in which participants had to detect an event that could appear randomly at any of two locations. Participant could only observe one location at a time while the other was occluded, but they could switch between them whenever they wanted by pressing a key and waiting for a short delay without fully visualizing any of the locations. The duration of the event was different on each location, being three times longer in one location than in the other. We manipulated the duration of the delay when participants switched between locations (500, 750 or 1000 ms) and the duration of the events (500, 750 or 1000 ms for the short duration and 1500, 2250 or 3000 ms for the long duration respectively). We registered the monitoring pattern of each participant for every combination of variables by calculating the average looking time at each location before switching to the other. This pattern was then compared with an optimal observer model. Results show that participants are good at determining which location to prioritize, since they dedicated more looking time at the shorter duration event location and as short time as possible to the long duration event location. However, they adjusted this pattern suboptimally to changes of other variables, such as the relative duration of the events and the switch time. In conclusion, participants are able to integrate multiple duration intervals to plan an action pattern, but they seem to fail at incorporating the switch cost to their temporal monitoring pattern.