Review of: "CAN ANIMALS DETECT WHEN THEIR OWNERS ARE RETURNING HOME? AN EXPERIMENTAL

[Im a Hedge]

I have just read the Wiseman et al article for the British Journal of Psychology, "CAN ANIMALS DETECT WHEN THEIR OWNERS ARE RETURNING HOME? AN EXPERIMENTAL TEST OF THE ‘PSYCHIC PET’ PHENOMENON", wherein the authors attempt to replicate Rupert Sheldrake's and Pam Smart's "Dogs that know" experiments. I was not able to access the actual journal article, but did find a manuscript at:
http://www.google.com/url?sa=t&sourc...Ja2wnqhj4JodZg

This appears to be a pre-publication manuscript, and I have not been able to confirm if any changes were made between this version and the published version.

As an overdue caveat, I want to mention that I am very familiar with reading and evaluating scientific papers, but not in the field of psychology. As such, some of my criticisms of this work, and others in the area of parapsychology, may actually be applicable to the field of psychology as a whole, and not specific to the studies in question.

There are many problems with this study.

No negative controls.
While it is the case that this experiment is trying to replicate previous work, and makes some effort to follow the same protocol, they did make several changes to the protocol. This is to be expected, as one of the purposes of replication is to tighten up the protocols. It is difficult to understand why a reasonable negative control was not included. Had the results been positive, the question of the validity of the phenomenon would not have been substantially addressed because of this failure of the experimental design.

Changing the protocol during the study.
Wiseman et al improved on the previous protocols by having a more objective definition of success. This involves a blind rater scoring Jaytee's visits to the window and indicating the apparent reason for each visit. The first visit that happened for "no obvious reason" was scored as the "signalling" event. However, after the first of the four experiments, they changed this criterion. For the remaining three experiments, signalling is the first unexplained visit that lasts more than two minutes. This was done in response to Pam Smart's post hoc rationalization of Jaytee's failure on the first experiment. For those without scientific training, let me assure you, changing your protocol during an experiment because you don't like the results you are getting is a bad thing to do. There were multiple other changes to the protocol. This is so prevalent that each of the four experiments essentially has it's own protocol. This is not four replications of the same experiment.

Small sample size.
Four trials isn't enough to establish anything. What is the point of conducting study that cannot possibly answer the question at hand? If all four trials had been successful would Richard Wiseman now be a proponent of animal telepathy? I doubt it. It wouldn't convince me.

General Comments.
The paper seems insignificant. It does not produce any meaningful results, and does not seem to be designed in a manner that would make it possible to have produced any meaningful results. The initial experimental design is sloppy, the manner in which the experiment was carried out was sloppy, and the presentation (at least in the manuscript I had access to) is sloppy. I am not sure why the British Journal of Psychology found it worthy of publication. I am led to question to thoroughness of their peer-review process (although, in fairness, bad papers appear from time to time in the best of journals). I fail to see what this paper contributes to the total body of knowledge. While I have not been convinced by what I have seen so far of Sheldrake's work on the topic, I don't consider this study to be meaningful as a refutation (or, as Sheldrake claims, as support).


I am a Hedge

[jacob]

Thanks for the review, I am Hedge.

Looks like you're doing a fine job in reviewing the various publications.
My hope you'll continue your efforts for our mutual expansion of knowledge.