[Rudism]

From the interview, I recall that Dr. Odgen-Avrutik mentioned that she only wished to be involved in the dog experiment if animal communication could also be worked in to the design. It seems to me that a much simpler approach (that would also have the benefit of not muddying the dog experiment) would be to test her animal communication separately.

The anecdote that she gave which convinced her that her ability was real is a really good starting point--she said that her pet cat told another communicator the same thing that the cat had earlier said to her, thus confirming that the information really was coming from the cat. It's not a very huge leap to come up with double-blind experiment designs that could test that under controlled conditions. I would say these designs could even possibly be much simpler and less time consuming than the dog-at-the-window ones.

Just a quick example off the top of my head that only involves one communicator and one dog: flash a random symbol from a pre-defined set of symbols at the dog in a closed-off room, then have someone (who doesn't know which symbol was flashed) bring the dog to the communicator and record which symbol the dog says was shown. Rinse and repeat for N trials and see if you get results above chance.

Instead of the communicator expressing the importance of going to the window when they sense their owner coming home (as I seem to recall was suggested during the interview), the communicator could express the importance of remembering which symbol was shown and identifying it to her after each trial.

The ultimate claim that should be focused on here is that the animal can convey information to a person that the person couldn't otherwise know. Another possible design would be to pass messages between two communicators using one animal, but I think a single-communicator design would be better since that precludes any possible conspiracies between communicators who have their careers at stake.