This is the fourth in a series of posts concerning selected papers presented at the July, 2014, GEIPAN workshop held in Paris. This post summarises my take on a paper titled "Qualitative Expert Evaluation and Quantitative Characterization of Official Reports on Alleged Unidentified Aerial Phenomena in France (1970-1979)," by Jean-Pierre Rospars (click here.)
Rospars sets out to analyse a set of about one thousand reports from the Gendarmerie Nationale for the decade 1970-1979. This analysis aimed to explore:
1. Are the "interesting" sightings of little or not understood phenomena?
2. Is there "Statistically significant quantitative difference between "interesting" and "uninteresting" sightings?"
Reliability of report classification:
Twenty eight engineers from CNES Toulouse classified the reports into four categories. These were:
(A) Fully identified phenomenon.
(B) Phenomenon likely assignable to a known phenomenon.
(C) Unidentified phenomenon but the report is of little value.
(D) Unidentified phenomenon and report is of sufficient interest to deserve a subsequent analysis.
The author also personally categorised the set of reports.
He concludes that "A classification will be considered reliable if and only if the same report evaluated by two (or more) experts is put in the same category."
1. "Is the identified/unidentified...correlated to the distance between the observer and the reported phenomenon?" The answer was yes, "72% of the 'close' ones are in category D."
2. "How are the reported events related to the population of potential observers?" This result suggested that reports of category D occur preferentially in the least densely populated communes."
3. "How does the frequency of reports vary as a function of the time of the day?" The answer was, "...the expression in percent...shows a clear excess of D cases over AB cases from 9pm to 3am..."
1. "Call for a better definition of the "surely explained" A and "probably explained " B categories.
2. It would appear that "explained" and "unexplained" reports have separate characteristics.
3. "More generally, may the contribution of expert evaluation and statistical comparisons lead to a more objective appraisal of the global significance of reports."
The full English version of the paper is available, click here.