The actual story of these reports begins on December 26 and 27, 1951 when Ed Ruppelt and Colonel S.H. Kirkland of the Air Technical Intelligence Center met with members of Battelle Memorial Institute, a think tank located in Columbus, Ohio. The idea was to analyze the UFO data "scientifically" and see if any sort of patterns could be found and to provide information so that better data could be gathered. Ruppelt had also complained that he needed experts to assist in the evaluation of UFO reports if any sort of meaningful conclusions were to be drawn.
Ruppelt, who used the fictional code name of Project Bear for Battelle later wrote, "The several hundred engineers and scientists who make up the team run from experts on soils to nuclear physicists… They would make these people available to me… they would do two studies for us; a study of how much a person can be expected to see and remember from a UFO sighting, and a statistical study of UFO reports."
According to a document dated January 8, 1952, the Battelle representatives agreed that there was enough material available to make the scientific study. The document said, "It is very reasonable to believe that some type of unusual object or phenomena is being observed as many of the sightings have been made by highly qualified sources."
There were five areas of study, or five requirements made of the Battelle Corporation, officially code named Project Stock, and supervised by William Reid. They were to "Provide a panel of consultants… Assist in improving the interrogation forms… Analyze existing sighting reports… Subscribe to a clipping service… Apprise the sponsor monthly of all work done."
The project began on March 31, 1952. They couldn't have picked a better time, or worse, depending on the point of view. The 1952 wave was just beginning, and by the end of 1952, the Air Force had collected an additional 1500 reports, of which, 303 remained as unidentified. In the Status Report #7, dated November 10, 1952, it was noted that sighting reports had improved in both numbers and quality.
In the second of the reports under the auspices of Stork, dated June 6, 1952, it was noted that the preliminary analysis of the existing reports had been completed. They had also developed a preliminary new questionnaire. They had developed a list of about thirty different characteristics of a report that could be keyed onto IBM punch cards for the statistical analysis. These included geographic location, duration of the sighting, color, speed, number, shape, and brightness of the craft and even the evaluation of both the sighting report and the observer.
Although the contracts called for the completion of the study by October 1, 1953, it is apparent that reports were gathered until the spring of 1955. The findings of Project Stork were then, apparently, incorporated into Blue Book Special Report #14.
They began with 4000 UFO sightings but eliminated nearly 800 for a variety of reasons including those too poorly documented to be useful for the study. This left 3201 UFO sightings, the majority of which were from the Project Blue Book files dating back to the original project Sign, with a few coming from other, outside sources.
First they wanted to gather the essential facts from the reports. They also worked to establish the credibility of the witnesses. They looked at the internal consistency of the report and the quality of it. Finally they attempted to make an identification of it. They tried to fit it into some category of natural or conventional identification.
The first run at identification was done by the individual who transcribed the report onto a worksheet they had developed. That would be passed to a member of the identification panel who would also evaluate the report without knowing what the first pass had revealed. If two of the evaluators arrived at the same conclusion, then that was accepted as a final identification. If there was a disagreement, then the report was passed on to the other panel members for more analysis.
If either or both suggested the case was unidentified, the whole panel studied it. The unidentified cases, or "unknowns" as they were labeled by the Stork scientists and researchers, were defined as "Those reports of sightings wherein the description of the objects and its maneuvers could not be fitted to the pattern of any known object or phenomenon."
Once all analysis of the data was completed, the cases were reduced to statistical fields so that some sort of conclusions could be drawn. When all those aspects of the report were completed, then the final draft was put together which included a summary of the data and conclusions drawn by the Battelle scientists.
When Special Report #14 was finally released to the public and the news media, it was the summary that was examined first. It suggested that better reporting, undoubtedly because of the improved questionnaires, and better investigative techniques developed by the Battelle scientists had reduced the percentages of "unknowns." For the final two years of the study, that is 1953 and 1954, the number of unknowns was only nine percent, and for the first few months of 1955, that had been reduced even further to three percent. Those figures, of course, overlooked 1952 when the number of unknowns was about twenty percent.
It was being suggested by the Battelle scientists and the Air Force investigators that the improved techniques were demonstrating that with enough information and with proper investigation, all sightings could eventually be explained. Again, as usual, the media seized on this idea included in the summary, and reported that the Air Force study had proven that flying saucers were just another myth.
Dr. Bruce Maccabee, who studied Special Report #14, related in a document published by CUFOS, that "Despite the use of information contained in the summary as part of the press release, it is clear that it was not written as a press release. Rather, it was written to supplement the information given in the main text. In other words, it was assumed that whoever read the summary would also read the main text."
I disagree with Dr. Maccabee on one point. I believe that those writing the summary assumed that those outside of the relatively few members of the military who were interested would ever read the main text. All had been around Washington long enough to know that journalists, congressmen, and government officials never read the whole document. It was the reason that executive summaries had been created. Pack them with information and that would be all that would be examined. Those who only read the conclusions would not be aware that the information in the main body of the text did not support the conclusions drawn in the summary. I'm convinced that those who wrote the summary counted on it being all that was read.
Ruppelt, who had with Kirkland, originated the idea of the report, criticized it on its release for what he called "weasel wording." For example, there is nowhere in the report where it is stated, positively, that UFOs don't exist. Instead, as Maccabee points out, they suggest in "probability arguments" that UFOs don't exist. They can't prove it, so it is reduced to UFOs probably don't exist, which by 1955, was the conclusion the Air Force wanted drawn. The original rationale for the report had long been lost.
Again, for example, one of these probability arguments was their attempt to "create" a "flying saucer model." Maccabee wrote, "Here they imply that because none of the sightings matched each other identically, the probability that any sighting was valid 'is concluded to be extremely small.'" Of course, this overlooks the fact that only one UFO sighting has to be valid for the whole argument to be eliminated.
The Battelle researchers had selected twelve cases that were unknowns and used them as a basis for constructing their "flying saucer model." They looked at the overall descriptions, the flight dynamics, and other observed data, as well as the reliability of the witnesses and the duration of the sighting.