Security measures in an automated ganzfeld system

Journal of Parapsychology, The, June, 1996 by Kathy S. Dalton, Robert L. Morris, Deborah L. Delanoy, Dean I. Radin, Robin Taylor, Richard Wiseman

At the end of this timed period, the computer signals the experimenter to begin the sending period. The experimenter then fades down the sender's relaxation audio tape, signals the computer to begin the sending period, starts the mentation tape recorder, and prepares to take down the receiver's mentation.

At completion of the sending/mentation period, the computer signals the experimenter to fade out the white noise to the receiver and to review the session mentation with him or her. After review, the receiver then takes off the eye shields and prepares to review the four target possibilities. After reviewing the four possible targets, the receiver ranks and rates them according to the correspondence of his or her imagery to each target. When the judging sequence is completed, the computer saves the data and then instructs the sender to return to the receiver's room and reveal the target. Session data are then sent to the printer for multiple printout, and the experimenter is prompted to close out the session.

SECURITY MEASURES

The automated ganzfeld procedure developed at PRL is widely recognized as one of the soundest methodologies in parapsychology. However, it has not been without its criticisms. Naturally, any replication attempt of complex studies, such as those carried out at the PRL laboratories, must take into account the advantages and disadvantages encountered in those studies and, while capitalizing on the former, must attempt to eliminate or minimize the latter. We have attempted to evaluate these criticisms in our own work at the Koestler Chair, and we will address them here. The main criticisms of the earlier automated ganzfeld work (e.g., Morris et al., 1993; Wiseman, Smith, & Kornbrot, 1994) were that there might have been: (1) possible subliminal sound leakage to the receiver through inadequate electronic component isolation; (2) repeated playing of the target tape during sending, which might alter it physically such as to provide a subtle cue; (3) sounds from the VCR, which might provide cues to the experimenter about which clip was being played as target; (4) sound leakage from the sender's room to the experimenter, which might provide cues if senders are noisy; (5) the possibility of a complex electronic signalling system between sender and receiver; and (6) deliberate experimenter bias. Our responses are as follows:

Criticism 1: There might be subliminal sound leakage to the receiver. The audio systems as well as the video systems are electronically isolated from each other. The only direct connection between the sender's and the receiver's audio or video systems is the output of the audio mixer into the output of the sender's audio mixer [ILLUSTRATION FOR FIGURE 2 OMITTED]. The technicians from the Electronics/Audio-Visual Department at the University of Edinburgh have electronically checked all such connections, following recommended procedures (all sound levels at upper limit), and have verified electronically that no such leakage exists in our facility. Checks should be conducted prior to the beginning of ganzfeld studies and again at approximately the midway point to verify continued security. In general, it is important to ensure that any electronic system that links various components within an environment is in fact functioning as it should. Faulty connections, inadequately shielded adjacent cables, inadequately isolated electronic components, components that drift outside of specified parameters, and other such problems can produce bias of information in the system. Even if this leakage is so minimal that it would be extremely unlikely to be having an effect, that remote possibility can still be enough to raise concerns, especially from those who regard genuine psi effects as even more unlikely (e.g., Wiseman et al., 1994; Humphrey, 1995). Thorough testing of the system, ideally during the course of the study as well as initially, can thus prevent considerable debate at later stages.