ESP and the Flaws of Psychological Research

The ability to see into the future is often thought to be nothing but the fantasies of fringe people, and so it was of great surprise that Extrasensory perception was the subject of a paper published in the Journal of Personality and Social Psychology at the beginning of the year. The article was written by Daryl Bem, a Professor Emeritus of Psychology at Cornell University. His paper describes a series of experiments performed with thousands of students that seem to support the existence of some sort of ESP phenomena, but the academic world of psychology has largely scoffed in response.

The nine experiments that Bem conducted vary in detail and procedure, but the structure is the same conceptually for them all. In psychology there is something known as priming, which is to show a stimuli, such as a word, that has a negative or positive connotation, followed by a different stimuli, such as a picture, that also has a negative or positive connotation. The subject is asked to judge whether the picture is negative or positive, and the time it takes for them to make this judgment is recorded. For subjects who see a word that matches the picture in terms of connotation, this reaction time is significantly quicker (for instance, someone who sees the word ‘bad’ will judge a picture of a thunderstorm to be negative much quicker than someone who received contrasting stimuli). What is intriguing about Bem’s study is that he has reversed this process- the subject judges the connotation of a picture and his reaction time is measured, and afterward he sees the priming word- and yet the primacy effect is seen despite the fact the priming stimuli was only presented after the fact.

The publishing of the paper has left the academic world of psychologists aghast. Popular criticisms include scrutiny of the statistical analysis Bem used. His analysis was done using what is called a one-tailed T-Test, a statistical procedure to demonstrate that there is a statistically significant difference amongst a group of results compared to the rest of the data, either greater or less than the norm. To picture this, imagine a bell-curve on a graph, with the tailed ends of the curve (this is indeed where the test gets its name) colored in. A one-tailed T-Test will declare results to be significant if there is significance located in one of the two colored tails of the curve; whether the test is looking for significance in the tail lesser or greater than the rest of the curve is determined by the experimenter in advance. The problem is that this method allows for researchers to ignore possibilities at the other tail of the spectrum, or forces them to backtrack and redo their analysis when it turns out their hypothesis was false (for instance, a researcher may do a second analysis that looks for significance at the opposite tail of the curve, even though his study was initially meant to look for significance at the originally chosen tail alone). A two-tailed T-Test is favored by many experimenters because it demands for significance at both ends of the spectrum. Some have accused Bem of using a one-tailed T-Test to deliberately fit his needs and eschew the need to find significance at both ends of the spectrum. The problem here is that this argument evades what is really represented by this bell-curve spectrum: ESP causing a greater chance of success in a particular task. This greater chance of success would be represented by the high-end of a bell-curve. Because no one would argue for ESP to interfere with the success of foresight- indeed, it goes against the whole idea- the need to test for significance below the norm is not needed. There may be deeper problems in Bem’s analysis that have yet to be pointed out, but this is not one of them.

What is shocking is not that a respectable psychologist wrote a paper that supports the existence of ESP, but that a touted scientific journal agreed to publish it. To get an article into one of these journals, the paper must be read and approved by an anonymous trio of fellow researchers. A study generally has a better chance of being published if it takes an aspect of an existing study in print and attempts to further its aims from a derived perspective. Because of this, and because of the specialized fashion in which research students in fields such as psychology are molded, academic journals are full of studies that are derivations of previous studies. Very rarely are entirely new lines of thought accepted, because they represent an unproven tract. For three anonymous reviewers to approve of Bem’s work, their reservations only coming in suggestions, is shocking.

Others have tried replicating Bem’s work and have failed to do so, although their experiments were done over the internet and thus debatable in procedure. Others will continue to poke holes in Bem’s statistics or methodology, and undoubtedly valid complaints will be raised. The problem is that many of these complaints could be brought upon existing research papers that are lauded.

The problem is the dirty little secret of psychological research. Papers are routinely published with analysis techniques skewed to favor their wanted results, with methodologies that often seem arbitrary. Of course, to begin with the agreed ratio for statistical significance was basically made up without rationale or logic. Essentially, what Bem has done is that he has beaten psychologists at their own game. How can they criticize him for the same sins that so many of them commit?

Ultimately, attacks on the paper’s technical components are superficial. What is truly appalling to psychologists is that Bem asserts a fundamental assumption of the universe, the linear flow of time and causation, can be broken. The boat has been rocked, and in a field where the current research is primarily the umpteenth derivative of previous research, no one wants to start looking at their tiny slices of the field in an entirely new light. For all the talk of independent thought, scientists- especially psychologists- are generally traditionalists. After all, in science it is often the outsider who rejects the small portions of empirical study and presents the grand new theory to view things. Darwin did it for biology; Einstein for physics. To put Bem in such a lofty category would of course be ludicrous. There is a very good chance his results may turn out to be easily explained by previously known means; this is the case with most papers. But the fact scientists are so unwilling to listen to new paradigms, even ones presented in conventional means of experimental study, is a sign of the times. Science has gotten small. Science has gotten safe.

Carl Jung believed in synchronicity, that events of mind and matter often occur simultaneously to form a greater meaning. Of course, despite their immense contributions both Jung and his famed colleague Sigmund Freud are largely dismissed in American psychology and are reduced to the butt of jokes. Not surprising. For psychologists, the only thing that matters is the statistical significance. Even if it’s just a made up number, it makes them feel safe.

Comments are closed.