In addition to the potential pitfalls of choosing a measuring method, researchers must also account for the great degree of variability of physiological changes, which Radin does not do convincingly. Skin resistance, like other physiological measures, varies greatly from person to person and over short and long periods. It also reacts to many aspects of the test subjects' internal and external environment, which is why investigators use a number of adjustments to remove unwanted variation so they can focus only on the changes in which they are interested.
While to his credit, Radin does try to reduce some unwanted variability, his efforts seem to be indirect and arbitrary at best, especially when the process can be very tricky. Radin measured the change in physiological states by subtracting the very first sample of skin resistance on each trial from all the remaining samples of skin resistance for that trial. The evidence for presentiment, he says, is the fact that the averages of the change in skin resistance are larger before viewing emotional pictures than before calm pictures. That seems to make sense. But the first samples of skin readings taken in the trial set the baseline against which the results of all future trials will be compared and computed. So if, for some reason, the first samples of the trials involving emotional targets happen to have a somewhat lower value of skin resistance than the first samples of the calm trials, this alone would yield, perhaps falsely, a bigger "change score" for the emotional trials.
To see this problem in action, assume that the average raw score level for both calm and emotional trials is 20. If the baseline for the calm trials is 15, then the change score for the calm trials would be 20-15=5. If the baseline for the emotional trials is 10, then the change score for the emotional trials would he 20-10=10. Thus, the scoring procedure produces a bigger change in the emotional trials, all because of the differences in baselines. As you can see, the simple choice of method can greatly influence the findings. The devil, as they say, is in the details.
So how do we know which measure to trust when each gives a different outcome? The solution is clear only when we have a detailed theory of the underlying process being studied. But that's a key problem here: There is a general lack of a positive theory of ESP or psi. What kind of a process is it? How does it behave? Indeed, as many parapsychologists recognize, ESP is, at present, defined negatively, in terms of what it is not; the experimenter claims she has found psi when she has eliminated all normal scientific explanations for the outcome. Given this hazy state of affairs, there is no principled way to state what the correct measurement procedure should he. This would not be too serious if different methods produced the same outcome. But we do not know if this is the case in Radin's experiments.
If I am investigating changes in the firings of nerve impulses in the optic nerve, for example, we have both extensive theory and data to inform us of the appropriate measures to use: We know the underlying distribution of such firings and we know how to appropriately transform them so that the measures of change make sense in terms of what we know about nerves and nerve impulses, in the case of the presentiment hypothesis, however, we do not have a detailed theory and sufficient data to know what sorts of transformations and measures of change make sense. So before we can believe that the physiological changes show that the subject is anticipating the emotional picture, we need to show, at the least, that different ways of measuring the physiological changes will yield the same outcome. We also would need to collect the physiological measures under more varied circumstances and over longer time periods.
The history of attempts to investigate scientifically psychical phenomena goes back 150 years, and is replete with examples of psychical researchers claiming they finally proved the existence of the paranormal. In each instance, subsequent generations of parapsychologists have had to discard as badly flawed what had seemed to the previous generation to be irrefutable proof of psi, or psychic phenomena.
A case in point is the study cited by Radin of the "significant correlations in brain waves between isolated identical twins." This study was reported in the journal Science in 1965 by Duane and Behrendt. These investigators took advantage of the fact that alpha brain waves can be induced by simply closing one's eyes. The researchers put two members of a pair of twins in separate rooms and connected them to electrodes to measure their brain waves. They instructed one of the twins to shut her eyes at predetermined times. This produced the expected alpha rhythms in her brain waves, and supposedly caused the other twin's brain waves to show alpha rhythms at the same time. If this is indeed what had happened, it would be evidence for ESP. But there were many methodological problems. First, the isolation of the twins was not very convincing since they were in adjacent rooms. Second, the evidence for the correlation of the brain waves was based solely on subjective, visual inspection of the brain wave recordings. As psychologists know, people are very poor at determining correlations subjectively, which is why experimenters trust only correlations tabulated by computers.
Tags:
brain,
bullets,
fiancee,
hunch,
hunches,
intuition,
parapsychology,
psychic,
ray hyman,
research psychologist,
sixth sense,
university of oregon
