The psychic pets thing sounds like people are remembering the times it works and forgetting the times it doesn’t.
I read two of Sheldrake’s papers on his site. The first was a paper reporting the results of a telephone survey. I thought that the questions were pretty poor. The first question was something like, ‘Have you or has anyone in your family ever noticed your pet anticipating the arrival of a family member?’ To which about 38% of dog and cat owners said yes. Well, if we think about what our pets have ever done, they do weird things all the time. If I remember one time that Fluffy ran to the door ten minutes before my wife came home, I answer “Yes” to the question. The fact that Fluffy didn’t run to the door 99.9% of the time doesn’t come across in the survey.
Sheldrake does admit that the survey is potentially unreliable. This, however, did not stop him from conducting at least four of them.
The other was the paper about Jaytee, the dog which “knew” its owner was returning. According to Sheldrake, the dog’s behavior has been documented over four years or so. At first by the owner’s parents, and later in a series of experiments by Sheldrake. Apparently, the dog would exhibit a particular behavior at nearly (plus or minus 10 minutes) the same time as its owner set out for home. Where there is a greater discrepancy, Sheldrake explains this away as saying that the dog may have reacted to when its owner started thinking about going home. Sheldrake’s experiments involved random times for the owner to start back, multiple ways of getting home (e.g. taxis, bikes, walking, etc.) all in an effort to isolate the dog’s reaction.
Sheldrake claims that an skeptical and independent researcher replicated the results, but still refused to believe them.
One problem that I see is that the experiments were not done in a “controllable” environment, like a lab, but rather in the dog’s home. If the ability were really there, the dog should be able to perform it anywhere.
Again, Sheldrake says that these observations don’t prove anything, and that more research needs to be done.
The phone surveys seem very useless and poorly designed to me. The observations with Jaytee are interesting, but they would need to be replicated independently, in a controlled environment, with a different dog, before I would believe there was anything to them.