DavoMan said:
Hmm..thats very interesting. I will look over these results again. Its just a lot of data and it's quite overwhelming for me.
Include in your quest for knowledge some reading on the horse known as
Clever Hans, a textbook case where an animal received unconscious cueing from a trainer and was able to seemingly perform math calculations. As long as his trainer knew the answer, that is.
Studies like this often suffer from a lack of knowledge of how people can be fooled, both statistically and otherwise. When results are obtained that seem impossible or unlikely, the correct step is to tighten controls -- make sure there is absolutely no contact between parties to the test, called double-blind. Typically when controls are tightened, the phenomena vanishes.
Then there is the method sometimes used in paranormal card tests. A testee is given a run to guess the next card out of ten, and if he desires and the run is unimpressive, that batch is discarded from the total results. While that might seem fair, it may distort random results into less than random ones. All the trials must be included, good bad.
The analogy with the dog test is that
all activity of the dog must be considered, not just when a cameraman gets a phone call to turn on the camera (that in itself may cue the dog that something is up). In my family, our dog was always waiting at the window for us to come home from work. Amazing! How did she know? But when I stayed home sick, the dog was at the window most of the day anyway, not very remarkable.
You cannot pick out the statistics you want and discard the rest unless you are ignorant or trying to prove something that doesn't exist.