O’Keeffe and Wiseman asked five mediums to give 60-minute readings for five volunteer sitters, so there were 25 readings in all. The mediums and sitters never met and were placed in separate rooms. The sitters never actually heard their readings which were recorded and transcribed.
Afterwards, the sitters read through all the statements taken from all 25 readings, and they indicated how much each statement was appropriate to them. They had no way of knowing which statements came from their readings and which were taken from the other participants’ readings. The readings couldn’t be identified by any references to the day of the week either because all the sitters received one reading a day for five days.
If the mediums’ powers were genuine, the participants should have rated the statements from their own readings as more appropriate to them than the other statements. But in fact, with only one exception, statements from a given reading tended to be rated as less appropriate by the person for whom they were actually intended, than by the other participants. And among the statements, there were those that were rated highly, regardless of who they were intended for – these tended to be general (e.g. “Yes, a relative. Is it a man?”), and it’s thought such statements could make readings seem convincing. It’s known, for example, that there are statements (e.g. “I have scar on my left knee”) that appear highly specific but which the majority of people actually agree with.
“The present study found no evidence to support the notion that the professional mediums involved in the research were, under controlled conditions, able to demonstrate paranormal or mediumistic ability”, the authors said.
O’Keeffe, C. & Wiseman, R. (2005). Testing alleged mediumship: methods and results. British Journal of Psychology, 96, 215-231.
You have read this article with the title Testing mediums. You can bookmark this page URL http://psychiatryfun.blogspot.com/2005/05/testing-mediums.html. Thanks!