Gordon Mangan and the CIA's 1958 Psychometric Filter for ESP Assets

By Eleanor Voss · September 16, 2026

Kept secret for 43 years until 2001, this research into IQ and "social adjustment" provided the logic for an agency attempt to recruit psychic assets based on personality types.

The subjects were sorted into "sheep" and "goats," their minds dissected by ink blots and IQ tests to see if they possessed a power the government could weaponize. The cost of this sorting was the reduction of human consciousness to a set of actionable variables—intelligence, extraversion, and "social adjustment"—processed through a filter designed to find a needle in a haystack of academic ambiguity.

For decades, the CIA held a specific 1958 review in its files, a paper that sought to answer a cold, operational question: can you tell who a psychic is just by looking at their personality? The record is a review written by Gordon L. Mangan, an academic based in New Zealand for his work at Victoria University, a public research university in Wellington. Published by the Parapsychology Foundation, Inc., a New York-based publisher of paranormal research, the paper serves as a meta-analysis of how intelligence and temperament correlate with Extra Sensory Perception (ESP) scoring.

The Best Estimate Trick

The record begins with a failure. In 1947, Nash and Richards—researchers who conducted the first investigation into the relationship between intelligence and psychokinesis (PK) tests—found a "very small correlation (-.12)" between IQ and PK scores among 48 college subjects. Intelligence, it seemed, was irrelevant to the paranormal.

But the CIA's interest was not in failure; it was in the loophole. Mangan points to the work of Humphrey, a researcher who analyzed the correlation between intelligence and ESP scoring, who found a way to manufacture a stronger link. According to the record, Humphrey discovered that "when the 'best' estimate of scores is used as the ESP criterion," the correlation with intelligence increased. Rather than using a subject's average performance, which might be dragged down by "boredom and fatigue," Humphrey looked only at the highest peaks of achievement.

This is where the academic record ends and the agency's logic begins. The desk's reading is that the CIA's interest in this specific paper was the "best estimate" methodology. By discarding averages and focusing on the "best" results, an analyst can artificially inflate success rates and create the illusion of a reliable correlation. If the agency applied this logic to its own classified data, it could transform a series of random hits into a perceived pattern of ability, providing the justification for continued funding and the recruitment of "high-intelligence" assets who were, in reality, just lucky once or twice.

The Extravert's Edge

The search for a profile moved from the intellect to the ego. The record examines the Bernreuter Inventory, a tool used to separate introverts from extraverts, and finds a consistent trend: "extraversion was associated with higher ESP scores than introversion." In one study, 74 per cent of extraverts scored above chance, while 76 per cent of introverts scored at chance or below.

This finding is linked to the work of C. E. Stuart, a researcher who investigated the role of "affectability" and personality in ESP subjects. Stuart and Humphrey had previously collaborated on investigations involving a restricted 14-item scale to measure interest and personality. In Mangan's review, Stuart equates "affectability" with a broad range of interest, suggesting that the most successful ESP subjects were not the withdrawn mystics of pulp fiction, but people who were moderate in their interests and maintained a "temperate attitude towards their environment.

This transition from the "mystic" to the "extravert" is the pivot point of the file. The pattern suggests the CIA was not looking for the naturally gifted, but for a specific personality type: the social, the confident, and the adaptable. By targeting extraverts, the agency was seeking assets who could be integrated into operational environments without the psychological instability often associated with "paranormal" claimants. They weren't looking for psychics; they were looking for stable employees who happened to score high on a skewed test.

Sheep, Goats, and Ink Blots

The most aggressive sorting occurred in the division of "sheep" and "goats." In the record's terminology, sheep are those who believe in ESP and think they will succeed in an experiment; goats are those who reject the possibility. The record notes that while the sheep-goat criterion often separated high and low scorers, the results were inconsistent across different experimenters.

To sharpen the tool, the researchers turned to the Rorschach test. Using a checklist devised by Dr. Ruth Munroe to measure "overall adjustment," they found that the difference between believers and non-believers became "more marked for the well adjusted subjects." The record states that "poorly adjusted subjects scored at approximately the chance level," whereas well-adjusted sheep scored significantly higher.

This creates a precise, clinical target: the well-adjusted believer. The record describes the isolation of seven "signs" in Rorschach protocols that acted as "deterrents to ESP scoring," including a "cold, withdrawn, restricted attitude." The goal was to strip away the "poorly adjusted"—the erratic, the unstable, the truly fringe—and find the professionalized paranormal.

The Recruiting Profile

The document concludes with Mangan's academic honesty: "no more definite judgment can be made at this stage." He admits the variables remain in doubt and the conclusions are tentative. But the CIA did not keep this file for 43 years because they shared Mangan's hesitation. An intelligence agency does not archive a review of "undefined" relationships unless it is using that review to define its own requirements.

If the shape of this file is what it appears to be, it served as a feasibility study to create a psychometric screening profile for recruiting 'psi' assets. The agency wasn't interested in the existence of ESP—they were interested in the profile of the person who could be convinced they had it, and who was psychologically stable enough to be handled by a case officer. The focus on IQ, extraversion, and Rorschach-verified "adjustment" provides a blueprint for a recruitment filter designed to weed out the "goats" and the "poorly adjusted," leaving behind a pool of high-functioning, confident believers.

This published review served as a baseline or 'control' to be compared against the results of an undisclosed internal CIA experiment. The gap between Mangan's academic ambiguity and the agency's hoarding of the text is where the real story lives. The withheld portions of the file likely contain the agency's operational conclusion on whether these personality variables were actionable for intelligence gathering. The desk's reading is that the CIA attempted to systematize paranormal ability through personality typing, treating psychic power as a trait of the "well-adjusted extravert" rather than a scientific fact.

Ultimately, the record marks a failed attempt to turn the irrational into a bureaucracy. The agency tried to use psychometrics to find a weapon, but they only found a way to categorize people. The people who paid for this were the "subjects"—the college students and volunteers who were told their personalities were being studied, while the agency was actually calculating how to use their "social adjustment" for the state. The file is not a study of the paranormal; it is a study of how the government attempts to quantify the unquantifiable to ensure that when they hunt for ghosts, they only hire the ones who fit the org chart.

Gordon Mangan and the CIA's 1958 Psychometric Filter for ESP Assets

The Best Estimate Trick

The Extravert's Edge

Sheep, Goats, and Ink Blots

The Recruiting Profile

Sources