Remote viewing · 12 MIN READ · 8 SOURCE QUOTES
Feedback and evaluating a remote-viewing report
In remote viewing, feedback is the moment the viewer is shown the real target after the session, and evaluation is the separate step of comparing the report with that target. The DIA protocol for the Stargate unit defines both. A project officer picks the target, tasks the viewers, and after the reports are recorded gives each viewer the feedback and a written evaluation. Joe McMoneagle says that without some form of feedback you cannot tell whether you got something valid or invented it, and Paul Smith says a viewer's edge slips without it.
Judging has taken several forms. At SRI, McMoneagle says, each session was ranked in one of four places against the targets, with first place meaning an unmistakable match. The DIA protocol's evaluation form asked project officers and customers for a numeric score for accuracy and another for value, plus a narrative. The Monroe Institute's first remote viewing practicum, reported in the Fall 2002 Hemi-Sync Journal, used teams of viewer, monitor and judge, sealed pictures and a four-photo lineup, then tested the results against chance: 13 first-place matches in 24 trials, where about six would be expected.
The argument is about what those scores prove. A 1995 review the CIA commissioned from the American Institutes for Research accepted that a statistically significant laboratory effect had been demonstrated, but found the operational reports vague and ambiguous, never used to guide an intelligence operation, and recommended against continuing. McMoneagle, replying to the point that remote viewing was never used as a standalone source, says no agency acts on a single source. Inside the field, viewers also argue over whether in-session cues help or contaminate.
Feedback shows a viewer the real target after the session; judging scores the report against it. How that scoring works is the heart of the debate.
Key points
Feedback and evaluation are two steps
The DIA protocol defines feedback as information about the intended target, shown to the viewer by the project officer once a proficiency project is conducted and the information is secured. Evaluation is a written assessment of what the viewer reported, done by the project officer for training work and by the customer for operations. Feedback waits until every session is complete and every report is recorded, and viewers may not discuss their findings with each other until then.
Why viewers say it is essential
Paul Smith, who worked operational targets at Fort Meade, says real-world tasks often never produced feedback, and without a comparison a viewer's edge can slip unnoticed. The unit therefore mixed in practice targets with known answers, which he calls utility assessment targets: real sites whose ground truth was already known, such as a Soviet antiballistic missile radar. Comparing the result showed where to adjust the process or confirmed the viewer was still on the signal.
Ranking against decoys at SRI
McMoneagle describes the early laboratory yardstick as four places. First place meant no doubt that this was the target, and he says his detailed drawings made his sessions easy to match; second held valuable information but not enough to name the target; third, around 20 percent similarity, he calls chance; fourth meant no similarity. He says five of his six SRI sessions were first-place matches and the sixth could have been. Atwater, recalling the period to him in 2012, said the test used blind differential discrimination and that he thought McMoneagle got six out of six.
A worked example: the 2002 practicum
The Fall 2002 Hemi-Sync Journal reports twenty-four double-blind, independently judged trials at the Institute's first remote viewing practicum, led by Skip Atwater and Carol Sabick. Twenty-three participants and the trainer formed eight teams of three, rotating through viewer, monitor and judge. Each target was a pictured location sealed in an opaque envelope. The judge took the team's descriptions and sketches to Atwater for a sheet of four photos, one real and three shams, and ranked them first to fourth. Thirteen of 24 were first-place matches, about one chance in 480 by the article's calculation.
How the DIA scored operational work
The protocol required the project officer or customer to complete a Project Evaluation Feedback Form giving a numeric score to both the accuracy and the value of what viewers reported, plus a narrative. The stated purpose was to judge how useful the information would be in an actual operation. The 1995 CIA-commissioned review later summarized forty viewings from 1994 and 1995 rated by end users: accuracy averaged 3.0 on a six-point scale where 1 is high, and value averaged 3.5 on a five-point scale where 1 is major significance.
Accuracy means two different things
McMoneagle wrote in 1996 that scientists compare what a viewer got right with everything that could have been said about the target, producing a figure of merit. For collection, one new fact that completes a puzzle can be immeasurably valuable even if most of the rest is wrong, which he called the paradox of great value with little accuracy and the reason remote viewing information should not stand alone.
The 1995 review: laboratory effect, operational doubt
The American Institutes for Research reviewed the program for the CIA in 1995, with outside reviews by statistician Jessica Utts and psychologist Ray Hyman. It found a statistically significant laboratory effect but no clear evidence that paranormal phenomena caused the hits, and on operations found the information inconsistent, inaccurate in specifics and in need of substantial subjective interpretation. It concluded that continued use in intelligence gathering was not warranted. Utts called psychic functioning well established; Hyman agreed the effect sizes probably could not be dismissed as chance, said statistical effects alone do not show psychic functioning, and objected to a single judge for all the experiments.
The training cue debate
Feedback also shapes training. Lyn Buchanan says Ingo Swann gave in-session feedback, saying correct for a right answer and nothing for a wrong one, and that trainees were later given targets with none so they would stop depending on the monitor. McMoneagle objected to the Army's improvised training after Swann left, in which the trainer sat with the target and confirmed right answers. He compared it to being told you are getting warmer or colder, a cue that steers the viewer.
In their own words
Quotes are checked word for word against the recordings and documents they come from. Bracketed words correct transcription errors; “…” marks an omission.
“Feedback is information about an intended target. This is displayed to a remote viewer by the Project Officer after a Proficiency Enhancement Project is conducted and the information is secured. Feedback in an Operational Project may not be immediate due to the fact that the customer may have to do research or act upon the information.”
“The Project Officer and/or customer will be required to complete the Activity's Project Evaluation Feedback Form (see Appendix D). This form will allow the Project Officer and/or customer to assign a numeric score both to the accuracy and the value of the information reported by the remote viewers in response to taskings. The Project Officer and/or customer are also required to provide a narrative response as well to the information received by the remote viewers.”
“They're like UFO targets. I don't like UFO targets because how am I going to know whether I got something valid or invented it? You can't tell unless you have some form of feedback.”
“Well, there's four places. First place through fourth place. Fourth place is no similarity to the target at all. Third place, maybe 20% similarities, which is chance.”
“Feedback in a remote viewing session is essential so you can learn and so you can keep your skills tuned up. If you don't—if you can't find out what the target was for, for months and years even sometimes, then your edge can slip and you don't even know it, because you don't have anything to compare to.”
“The big surprise was [Ingo] always gave in-session feedback. So if you did something right, he'd say 'correct.' Okay? And if you did something wrong, he wouldn't say anything. No reward. No negative rewards. The silence was a negative reward when he didn't say anything.”
“One of the photos matched the target; the other three were shams. The judges, who were not otherwise involved with the remote-viewing process, were to use the 'police lineup' method to compare qualitative descriptions and sketches provided by the viewers and monitors with the four locations pictured. The judges were asked to score first-, second-, third-, and fourth-place matches.”
“Statistically, the very best world-class remote viewers under laboratory conditions tend to run 50/50 as to whether they are on or off target. Anywhere from 60 to 85 percent accuracy is expected. In eighteen years I have yet to meet a remote viewer who knows when he/she is on target (with a couple of exceptions based on specifics of targets).”
Important things to consider
Keep the original report intact
The 1995 review's finding that viewers considered it appropriate to make written reports consistent with known characteristics of the target shows why order matters. A revised report is no longer the original result. The protocol's order is the safer model: record the report first, hand over feedback only after all sessions are finished, and keep viewers from comparing notes until then.
Judge blind, with decoys
A judge who sees only the viewer's material and a set of similar pictures, one of them the target, gives a result you can test against chance: with four photos, a first-place match happens one time in four by luck. The 1995 review describes laboratory judging as a blind judge ranking a response against five photos, where chance puts the average rank at three.
The judge can be part of the puzzle
The Institute's own write-up added a caution: the practicum's significant results suggest psychic functioning but may not represent remote viewing as classically defined, because the positive matches were made by independent judges who may have used their own form of psychic functioning. In other words, a match may reflect the judge as much as the viewer.
Partial credit is not the same as usefulness
The 1983 Army assessment of the Gateway process reported, on the word of an Institute trainer, that experiments asking people to read a series of ten computer-generated numbers in a university laboratory on the other coast while out of body typically produced enough digits to suggest presence, but never all ten. It named information distortion as a major concern limiting practical use.
Does the viewer perceive the target or the feedback?
Some theorists argue that viewers perceive the feedback they will later receive. In one interview the host cited Eric Wargo's time-loop idea. Buchanan answered that his training pushes viewers to describe the target whatever the feedback shows: he describes a summer lake with ducks when the feedback photo showed it frozen, and counts that as a hit. Utts's review said feedback of the correct answer may or may not be necessary.
Operational feedback is slow or absent
The protocol notes that operational feedback may not be immediate, and Smith says many real targets never got any. When a source says it worked, ask what the comparison was, who judged it and whether the original notes were kept.
What is and is not established
The scoring procedures, the DIA forms and the 1995 ratings are documented. Individual hit stories and the accuracy rates viewers quote are their own reports. McMoneagle replies to the point that remote viewing was never used as a standalone source by saying no agency acts on a single source, and the two outside reviewers read the same data and disagreed about what it shows. The sources do not settle whether, or how, remote viewing works.
Common questions
What is feedback in remote viewing?
Feedback is the viewer being shown the real target after the session so the report can be compared with it. The DIA protocol defines it as information about the intended target, displayed to the viewer by the project officer once a proficiency project is conducted and the information is secured. It is separate from evaluation, the written assessment of what the viewer reported. In operations it may be delayed, because the customer may have to do research or act on the information, and viewers often never get it. Some teachers also give in-session feedback, a monitor saying correct for right answers.
Why do remote viewers need feedback?
They say it is how they learn and how they know whether the method is working. Paul Smith, a former Army remote viewer, says feedback is essential to learn and keep skills tuned up, and that without a comparison a viewer's edge can slip without the viewer knowing, which is why the unit mixed in practice targets with known answers. Joe McMoneagle makes the point about targets with no checkable answer: on UFO targets he cannot tell whether he got something valid or invented it.
How were remote-viewing sessions scored or judged?
In several ways. In SRI's early laboratory work, Joe McMoneagle says, each session was ranked in one of four places: first was an unmistakable match, third about chance, fourth no similarity. The Monroe Institute's 2002 practicum used a similar lineup, with judges ranking a viewer's description and sketches against four photos, one real and three decoys, and a binomial test against chance. The DIA's operational protocol asked the project officer or customer for a numeric score for accuracy and another for value, plus a narrative.
How accurate is remote viewing?
It depends on who is measuring and how. Joe McMoneagle wrote in 1996 that the best laboratory viewers run about 50/50 on or off target and that 60 to 85 percent accuracy is expected. In the 1995 CIA review, psychologist Ray Hyman cited SAIC researcher Ed May's estimate that about 20 percent of what a viewer supplies is accurate. For operations, end users in 1994 and 1995 gave an average accuracy rating of 3.0 on a six-point scale where 1 is high, and a value rating of 3.5 on a five-point scale where 1 is major significance.
What did the CIA's 1995 evaluation of remote viewing conclude?
The American Institutes for Research reported to the CIA on September 29, 1995. It accepted that a statistically significant laboratory effect had been demonstrated, but said the evidence had not clearly shown paranormal causes. On operations, it found the information inconsistent, inaccurate in specifics and in need of substantial subjective interpretation, said it had never been used to guide an intelligence operation, and concluded that continued use in intelligence gathering was not warranted. Its outside reviewers split: statistician Jessica Utts said psychic functioning was well established, while psychologist Ray Hyman agreed the effect sizes probably could not be dismissed as chance but said statistical effects alone do not show psychic functioning. McMoneagle replies to the point that the information was never used as a standalone source by saying no agency relies on a single source.
Can remote viewing work without feedback?
The question is described as open. Army remote viewer Lyn Buchanan says his training deliberately moved viewers from in-session feedback to targets with none, and finally to targets with no feedback ever, so they learn to describe the target whatever the answer sheet shows. He says most people view their feedback while learning and never get that training. The 1995 review's statistician, Jessica Utts, wrote that feedback of the correct answer may or may not be necessary. Others argue, in Eric Wargo's time-loop theory, that viewers perceive the feedback they will later receive.
How can I fairly check my own remote-viewing results?
Borrow the structure the Institute's practicum used. Have a partner choose a target picture, add three decoys, and seal the target in an opaque envelope so you cannot know it. Write your description and sketches first. A judge who knows nothing about your session then ranks the four pictures against your notes, and only after that is the envelope opened. Chance gives a first-place match one time in four, so keep a running tally across many sessions and compare it with that baseline. Keep the original notes unchanged.
Keep exploring
Prepared answers open instantly in the chat; anything else you ask gets a fresh answer.
Connected ideas
Remote viewing
Why blinding matters in remote viewing
Double-blind means no one in the room knows the target; frontloading means hinting at it in advance. Remote-viewing protocols exist to prevent both.
Read the summary →12 min · 6 quotes · 6 answersRemote viewing
Analytical overlay: impression or interpretation?
Analytical overlay is the mind naming what it perceives; viewers learn to record the raw data first and treat every label as unproven.
Read the summary →12 min · 8 quotes · 6 answersRemote viewing
Stargate & remote viewing
A secret U.S. Army and DIA remote-viewing program, its ties to Monroe's lab, and what the 1995 evaluation found.
Read the summary →12 min · 7 quotes · 6 answersPeople & places
Joe McMoneagle
Remote Viewer 001 of the Army's Stargate unit, who spent months in Bob Monroe's lab and now teaches remote viewing at the Monroe Institute.
Read the summary →12 min · 8 quotes · 6 answersPeople & places
Skip Atwater
The Army officer who ran training for the remote-viewing unit, brought its viewers to Bob Monroe, and led research at the Institute for two decades.
Read the summary →12 min · 8 quotes · 6 answersRemote viewing
Coordinate Remote Viewing
The staged, number-cued method Ingo Swann developed with SRI so Army remote viewers could be trained and run under tight controls.
Read the summary →12 min · 7 quotes · 6 answers