We tested the hypothesis that a tolerance for logical incongruity characterizes hypnotic responding and is related to reports of duality experiences during age regression and hidden observer responding during suggested analgesia. Contrary to the logical incongruity hypothesis, (a) highly susceptible hypnotic (n = 15) and imagination control (n = 15) subjects (collectively called "reals") failed to differ significantly on any task, (b) measures of logical incongruity did not differentiate consistently between reals and unsusceptible subjects (n = 15) instructed to fake hypnosis, and (c) different measures of logical incongruity failed to correlate significantly with duality reports or hidden observer responding. The differences in responding that did emerge between reals and simulators were accounted for by the different task demands to which subjects were exposed.

Two experiments tested predictions derived from the logical incongruity and differential demands hypotheses of trance logic responding. In Experiment 1, subjects that were highly susceptible to hypnosis showed higher levels of responding on three trance logic indexes (i.e., transparent hallucinating, duality, incongruous writing in age regression) than did subjects low in susceptibility to hypnosis who were instructed to fake hypnosis (i.e., simulators). In line with the differential demands hypothesis, hypnotic "reals" were less likely than simulators to report believing in the reality of the suggested situations and were less likely to report fine details in their hallucinations. Also consistent with the differential demands hypothesis (but inconsistent with the logical incongruity hypothesis), rate of trance logic responding correlated negatively with the degree to which hypnotic reals rated themselves as subjectively experiencing suggested effects, and as becoming absorbed in suggestions. Experiment 2 found that highly susceptible hypnotic and nonhypnotic subjects (collectively called "reals") responded equivalently on all suggestions. High-and low-susceptible simulators also performed equivalently on all suggestions. Consistent with the differential demands hypothesis (but not the logical incongruity hypothesis), (a) trance logic indexes differentiated reals from simulators when these indexes also measured incomplete subjective responding, and (b) trance logic indexes that failed to measure incomplete responding also failed to differentiate reals from simulators. The findings of both experiments are more consistent with social psychological views than with special processes views of hypnotic responding.In an influential early study, Orae (1959) compared the response to a visual hallucination suggestion in highly susceptible hypnotic subjects (i.e., "reals") and subjects low in susceptibility to hypnosis who had been instructed preexperimentally to fake deep hypnosis (i.e., simulators). During their experimental session, all subjects were administered the same hypnotic procedures, suggestions, and interviews. During their hypnotic session, all subjects sat facing a coexperimenter. While the subjects' eyes were closed, the coexperimenter rose from his chair and moved from the subjects' field of view. Subjects were then instructed to open their eyes and to "see" (i.e., hallucinate) the coexperimenter still sitting in the chair. After the subjects described their hallucinated image, the hypnotist pointed to the actual coexperimenter and asked the subjects what they saw. Ome (1959) reported that some of the hypnotic reals but none of the simulators spontaneously indicated that they could simultaneously see the coexperimenter and see through him (i.e., transparency response). Relatedly, when faced with the actual coexperimentei; hypnotic reals were much more likely than simulators to report seeing two images of the coexperimenter at the same time (i.e., double hallucination response). According

Council, Kirsch, and Hafner (1986) obtained empirical support for the hypothesis that significant correlations between questionnaire measures of absorption and hypnotic susceptibility are an artifact of subjects' beliefs about their own hypnotizability. We tested this hypothesis in a two-session experiment. During Session 1, subjects completed questionnaire measures of absorption, mystical experience, daydreaming frequency, and paranormal beliefs. During Session 2, subjects were tested for hypnotic susceptibility. Subjects were also exposed to one of three information manipulations: They were told about hypnotic testing either before or after filling out the questionnaires or were not told about hypnotic testing. The information manipulation moderated the prediction of susceptibility by the questionnaire measures for women, but not for men. For women, scores on the absorption questionnaire predicted susceptibility only when subjects were informed about hypnotic testing. In the told-after condition, this effect generalized to all of the remaining questionnaire measures. For men, none of the questionnaires was a reliable predictor of susceptibility.

We factor analyzed the responses of 784 subjects to a seventy-five-item paranormal belief questionnaire. Principle components analysis extracted four components which were subjected to alpha factoring with oblique and orthogonal rotations. An oblique solution was judged as most appropriate based on the factor correlation matrix. A higher order factor analysis yielded one common underlying factor. The results are discussed with respect to previous factor analyses of paranormal belief questionnaires by Jones, Russell, and Nickel [1], and Tobacyk and Milford [2].

Subjects who had been tested in the previous 3 months for hypnotizability completed questionnaires that assessed trait indexes of absorption in imaginings, dissociation, and temporal lobe dysfunction. Half of the subjects completed the questionnaires in a context unrelated to their earlier hypnotizability testing while the remainder were explicitly informed of the experimenter's interest in the relationship between their questionnaire responses and hypnotizability. Correlations between each attribute variable and all indexes of hypnotizability were never significant in the out-of-context condition and always significant in their context condition. In addition, regression analyses indicated that context added significantly to the prediction of subjective hypnotizability scores by each of the attribute variables. Implications are discussed.