Electrodermal Activity — State-Of-The-Art Measurement And Techniques For Parapsychological Purposes - EDA

Journal of Parapsychology, The, June, 2000 by Stefan Schmidt, Harald Walach

This analysis yields an interesting picture. In the early work up to 1993, researchers looked for the DMILS effect within the fast-changing phasic component of the EDA. But seven of the eight studies published since 1997 have used only the tonic component. This shift in the experimental setup was never reported or discussed or even mentioned due to incomplete descriptions of the technical setup until the PA Convention 1999 when an earlier version of this analysis was presented. And one wonders if even the researchers themselves have noticed this change.

The surprising result of our analysis is that there is not even one description of a state-of-the-art EDA methodology within the whole DMILS/Remote Staring literature. Although researchers have been reporting irregularities within EDA data that can be related to experimental conditions for more than 20 years, the nature of these irregularities was never addressed. Most of the researchers interpret these findings as a reflection of changing activity of the autonomic nervous system. But this conclusion cannot be maintained as the applied techniques allow a set of different interpretations. Irregularities within DMILS/Remote Staring data may be as well caused by:

Irregularities within DMILS/Remote Staring data may as well be caused by:

* irregular breathing. Irregular breathing causes a large EDA response.

* drifts. Drifts within the data can be caused by inappropriate electrode gel, insufficient time lag between electrode attachment and data recording or changes in the overall arousal of the participant. Balanced randomization should avoid those drifts, but not all studies were randomly balanced. Any drift starting or stopping within the experiment may create artifacts, although the trials are randomized in a balanced order.

* movement of the hand.

* inadequate filtering process. Potential artifacts can be caused by the reconstruction of the EDA curve by digital recorded data that have not been filtered according to Shannon's sampling theorem (see above). Data may also contain electronic noise as the result of inappropriate filtering.

* Ebbecke waves enforced by inadequate attachment of electrodes by velcro bands or by strain of electrode cable.

On the other hand, it can be argued that DMILS/Remote Staring effects were not detected because the researchers employed inappropriate technology. Existing effects can be covered up by:

* the use of inadequate electrode paste. Hypertonic gels may change the conductivity of the skin in a large scale and remove minor changes caused by the autonomic nervous system.

* inappropriate data recording. Large effects may decrease or even disappear if the sampling rate is below 10 Hz for the phasic component.

* inappropriate measurement principle. If the effects are within the fast-changing phasic component they will not be detected by methods which only score tonic parameters. On the other hand, effects within the overall arousal will be filtered out if the data are only checked for changes in the phasic component.