Tuesday, June 1, 2010
[ Read the first part ]
The Design and Reported Results
In September 2006, Swedish tabloid Expressen published an article titled "Scientist proves tricky telephone classic" (2006). Reuters news agency reported that Rupert Sheldrake claimed he had evidence of telepathic ability in conjunction with e-mails and phone calls. In tests with both e-mail and telephone, test participants had scored 40%, which is far better than the 25% which is to be expected by chance. The odds that the results were caused by chance was 1 in 1 000 billion, reported Reuters. That the research had been received with some suspicion was due to the fact that only 63 people had participated in the telephone study and 50 in the e-mail study.
The background of these news is several studies on telepathy that Sheldrake conducted between 1999 and 2004. His interest in telepathy derives from the belief that this phenomena confirms the theory of morphic resonance and its application in morphogenetic fields to which members of a social group are connected. These fields cannot be measured as such, but only by the effects they have and one effect is telepathy, suggests Sheldrake (2006). Another effect is the sense of being stared at, which Sheldrake claims is due to vision not being a one-way process. The image that is created by consciousness during visual perception also radiates from the eyes and can be sensed by the person being stared at (Blackmore, 2005).
The following review, however, deals with one of the studies on telepathy which, according to Sheldrake, constitutes evidence of the existence of telepathy: Experimental Tests For Telephone Telepathy (Sheldrake & Smart, 2003a). The review only covers experimental design and tests - Sheldrake's statistical analysis is not commented on.
Sheldrake's basic experimental design consisted of one person (the receiver) being called by one person (caller) randomly chosen from a group of four persons. In some of the trials, all the callers in a group were known and nominated by the receiver, in others at least two were known and the rest unknown and chosen by the experimenter. If Sheldrake's hypothesis was correct, people from the same social group would have telepathic contact, but people who were not members of the same social group would not. If the receiver was able to name the right caller to a greater extent than what was expected by chance, 25%, telepathy was considered to be the cause. In these tests, Sheldrake reported results that were slightly above 40% concerning callers known by the receiver and 25% concerning callers not known by the receiver, i.e. an obvious support for Sheldrake's hypothesis.
Sheldrake used a convenience sample for the study. Participants were recruited through newspaper advertisments which read: “Do you know who is ringing before you pick up the phone? Good pay for fun and simple experiments as part of psychic research project.” Additional recruiting was done through a recruitment website. Those who responded (receivers) was sent a more detailed description of how the experiment was to be conducted, and was also asked to nominate four people in their circle of acquaintances who also were willing to participate (callers). Thus, callers were recruited through snowball sampling.
Faithful to Sheldrake's approach, the experiments were done in the homes of the participants. The receiver sat in his or her home, the callers in their respective homes, and experimenters Sheldrake and Smart at yet another location. The experiments consisted of a preliminary experiment and two real experimental series, in which the following methods were used:
Two callers from the group of four known callers were chosen randomly by throw of a die where the numbers five and six were thrown again. If the same number came up twice, that caller got to call twice. The time for the call was also selected randomly but kept within the stipulated 60 minute test period, which in turn was divided in six segments. An experimenter called the caller one or two hours before the chosen calling time and notified the caller when to make his or her call. The caller was also asked to think about the receiver a minute before making the call. The callers who weren't chosen were also notified that they were not selected for the current trial. A couple of minutes after the test call, the experimenter called the receiver and asked what he or she had guessed. Sometimes, the caller was asked as well. The experimenter then made notes of the result, date, time, receiver, caller, and guess. This method was used in a preliminary experiment and with the first 17 receivers in the first real experimental series, in total 198 trials.
The random choice of time for the call was changed to a fixed schedule, for instance 10.15 and 10.45. Otherwise, it was exactly like method 1 and used for the remaining five receivers in the first experimental series and the first three in the second experimental series, in total 87 trials.
Only one call was made during the test. The experimenter chose a caller less than 15 minutes before the chosen time and the caller was notified, at latest, 10 minutes before the call. This method was used for 37 of the receivers in the second experimental series.
This method was similar to 3a but the callers who were not chosen were not automatically notified. Instead, they were told that if they hadn't been notified at least five minutes before the calling time, they had not been chosen. This enabled more tests during a shorter time period, in general with a 15 minute interval. This method was used for the remaining 34 receivers in the second experimental series, in total 268 trials.
Sheldrake's & Smart's null hypothesis was that the receivers would make a right guess in 25% of the calls, which is to be expected by chance. The alternative hypothesis was that the receivers would guess right in more than 25% of the cases, which would then be explained by telepathic ability. For hypothesis testing, an exact binomial test was employed. To combine the results of different test trials, Stouffer was used. For comparison between results from known and unknown callers and first and second trials, Fisher's Exact Test, an alternative to Chi2, was used. A 95% confidence limit was calculated when analyzing the probability of right guesses.
In a preliminary experiment reported by Sheldrake & Smart, Smart was the receiver and two sisters, her mom, and Sheldrake were callers. Sheldrake also acted as experimenter. Smart's result was 43%, i.e. significantly above 25%. Smart's best result (67%) was achieved when Sheldrake was calling.
In the first real experiment series, 9 receivers carried the stipulated number of trials (10) through and all but one guessed right in 40% of the calls. This series had a mortality - people who dropped out during the test - of 12 receivers. The most common mortality cause was said to be an inability to get all four callers to participate at the same time.
During the second and last experiment series, yet another hypothesis was introduced: phone calls from callers known by the receiver could be sensed but in calls from unknown callers, the result was what would be expected by chance. 16 receivers carried the stipulated number of trials (10) through. They guessed right in 54% of the calls from known callers and in 24% of the calls from unknown callers, results that lends support for the new hypothesis. This series had a mortality of 21 receivers.
All in all, Sheldrake's & Smart's experiments supported both the original alternative hypothesis and the one introduced in the second series; it seems confirmed that you can sense, telepathically, who is calling and that this ability is dependent on the caller being someone you know.
Are Sheldrake's & Smart's results to be trusted? Are the experiments well designed and conducted, or do they have weaknesses that threats the conclusion? This will be discussed in my next post. Stay tuned.
[ Read the third part ]
Blackmore, S., (2005). Confusion Worse Confounded. Commentary on Sheldrake. Journal of Consciousness Studies, 12, (6), 64–66. URL http://www.susanblackmore.co.uk/Articles/jcs2005.htm.
Forskare bevisar lurig telefonklassiker. (2006, 7 September). Expressen, URL http://expressen.se/index.jsp?a=676876.
Sheldrake, R., (2006). In Conversation on Abc Radio National – Rupert Sheldrake [www document]. URL http://www.abc.net.au/rn/inconversation ... 754367.htm.
Sheldrake, R., & Smart, P., (2003a). Experimental Tests For Telephone Telepathy. Journal of the Society for Psychological Research, 67, 184–199. URL http://www.sheldrake.org/Articles&Paper ... _tests.pdf