Participants reporting any unusual phenomena in the 'haunted' location within the Palace were asked to mark the locations of their experiences on a floorplan. These floor plans were then divided into several equally-sized areas. The mean strength and variance of the magnetic field within each of the areas was measured and subsequently correlated with the number of unusual phenomena reported by participants within that area. It was predicted that the sample mean of the resulting correlations would be significantly different from zero.
These are some parts taken from the original paper. They did find a few spots which have a low electromagnetic field which correlate with the hypothesis made by Persinger.