By: E; Paulinyi May, T; Vassy, Z
Publication Name: J Altern Complement Med
Year: 2005
Objectives: The primary aim of this study was to conduct a replication, simplification, and extension of similar previous studies that claimed anomalous anticipatory skin conductance responses prior to various stimuli and to provide sufficient protocol and analysis details in order to foster additional replications. A secondary aim was to provide a testable model in order to understand the observed results. Design: We used standard skin conductance measures and techniques to search within 50 participants for responses prior to 1-second duration, 97-dB acoustic stimuli, compared to prior to silent controls. We used an interstimulus interval randomly and uniformly distributed between 30 and 50 seconds. Outcome measures: The dependent variable was the difference between proportions of 3.5-second prestimulus intervals prior to acoustic stimuli and prior to silent controls that contained a fully formed, nonspecific skin conductance response (ns-SCR). The null hypothesis was that the proportion difference should be zero. Results: We found a significant proportion difference of 0.032 (Z = 2.08; effect size = 0.077 +/- 0.037; p(1t) = 0.0018), which is a replication of earlier similar studies. Conclusions: We examined and ruled out a number of potential artifacts that might have accounted for this finding. To understand these results, we demonstrated, by Monte Carlo techniques, that a possible explanation is that experimenters may have used their own intuition to initiate experiment runs to somehow sort otherwise random nonspecific skin conductance responses into appropriate bins in order to mimic physiological responses. We found experimental evidence to support this idea as an operational mechanism. If this speculation is confirmed in prospective studies, then this intuition-based mimicking of effects may profoundly impact the interpretation of results from complementary and alternative medical studies that use statistical inference to assess outcomes.
Click purchase to access this publication. Many may be available at no cost.