# Statistically deciding time intervals for independent events

I am working with camera trapping data and am trying to decide on a time interval to classify independent events. I have many images of wild dogs that were captured on 60 camera traps that are spatially independent of one another, but I am unsure of how to ensure events are temporally independent. In most camera trapping studies the amount of time that has to elapse between images in order for events to be independent is usually chosen based on ideas of the animals behavior, but I am after something that statistically justifies my chosen time interval. I have begun by calculating the number of independent events that occur based on different time intervals (1 -15 minutes). As can be seen in the graph I have attached, the number of events does change based on what time interval is used ...

The number of independent events at different time intervals

My question is how would I statistically determine which time interval is best? Would looking at the difference in slope between time intervals work? Or is something much more complicated needed.

• Without a detailed understanding of activity and movement patterns of the target species, it is difficult to determine an ecologically justified threshold for temporal independence. This excerpted from cam capture, but probably best to have an expert's advice. And @TimAssal would probably have some thoughts. – Chris Apr 20 at 2:47