Anticipatory awareness of emotionally charged targets by individuals with histories of emotional trauma

Journal of Parapsychology, The, Spring, 2004 by Theo K. de Graaf, Joop M. Houtkooper

Statistical Methods

For the final analysis of the results we concentrated in the first place upon the number of correspondences at these sensitive spots. This yielded the following displacement types:

A. p-1 ~ p-1: correspondence between response and Zener target (ZT), both preceding the location of an emotional target (ET).

B. p ~ p: correspondence between response and ET (= "direct hit").

C. p 1 ~ p 1: correspondence between response and ZT, both following the location of the ET.

D. p-1 ~ p: 1 forward displacement: response at location just preceding ET corresponds with ET.

E. p 1 ~ p: -1 backward displacement: response at location just following ET corresponds with ET.

F. p ~ p-1 : -1 backward displacement: response at location of ET corresponds with ZT preceding the ET.

G. p ~ p 1: 1 forward displacement: response at location of ET corresponds with ZT following ET.

H. p-1 ~ p 1: 2 forward displacement: response preceding location of ET corresponds with ZT following ET.

I. p 1 ~ p-1: -2 backward displacement: response following location of ET corresponds with ZT preceding ET.

J. p-2 ~ p: 2 forward displacement: response located two places before location of ET corresponds with ET

K. p 2 ~ p: -2 backward displacement: response at two places following ET corresponds with ET.

Table 1 gives an overview of the various position effects.

Global statistics. None of the participants had above chance hits for the exact location of the pictures on the target sheet and no correlation with the trauma scores was discovered. We then had to ask ourselves whether the various displacement scores could be considered to be stochastically independent from each other. (2) In view of possible response stereotypes, such as repetition avoidance (Crandall, 1988), combined with the observed departure from randomness of the ZTs, we decided that this was not a valid assumption. Thus, significance levels, e.g., in the form of a [chi square] statistic by adding up the squares of the z-scores, were deemed to be unreliable.

Therefore, in order to control for the influence of possible individual response biases and stereotypes, we decided to calculate for each participant and for each displacement type the number of displacements for "target" Zener cards (ETs) as a percentage of the total number of displacements of that type for all ZTs. In this way, any stacking effect was also regarded as being sufficiently compensated for, as this would have to show up in both numerator and denominator of the quotients involved.

The results are shown in Table 2. In the computation of the expected mean percentages and the z-scores, the actual target and call frequencies were accounted for with the help of Stevens' formula (Burdick & Kelly; 1977).

In this table nothing conspicuous strikes the eye, except perhaps for the fact that the percentage of direct hits on ETs (B-perc) is among the two highest of all 11 position effects and at the same time also shows the smallest variance. The meaning of this finding could be that in this sample "direct hitting" was less subject to individual differences than the various displacements.