Cluster Analysis of Features Associated with Unidentified Anomalous Phenomena Described in 216 Select Reports from 1947-2016
- Scientific Coalition of UAP Studies
- Aug 5
- 1 min read
Updated: Aug 15
By Stephen Bruehl, Sarah Little, and Robert Powell
in press, World Futures, Aug 2025
Patterns in reported characteristics of unidentified anomalous phenomena (UAP) in historical databases can be statistically examined to gain insights into distinct types of reported UAP. We applied cluster analysis to the UAP report database created by Powell et al. (2023) that applied strict selection criteria to witness reports in several historical databases. They identified n = 301 UAP reports likely to represent true anomalous phenomena and manually grouped them into classes based on shape. Here we examined a subsample of n=216 of these reports containing complete data necessary to conduct cluster analysis. We targeted five reported UAP features described in that previous work: shape, estimated size, ability to hover, presence of electromagnetic (EM) effects, and presence of sound. The two-step clustering algorithm identified seven distinct clusters, with this model of good statistical quality (silhouette value = 0.6). Object shape was the strongest driver of clustering results although other features contributed. For example, two distinct clusters were characterized primarily by absence of hovering (potentially reflecting unique reporting sources) and two distinct clusters comprised objects all exhibiting either: 1) EM effects (disc/sphere or ovoid) or 2) sound (unassociated with shape). Of the objects exhibiting EM effects, 97% were viewed at estimated distances of ≤ 2,000 feet from the observer. Similarities to and differences from manual classification results of Powell et al. were noted. Findings suggest that statistical pattern recognition techniques like cluster analysis can extract meaningful information from high quality witness report data that could improve understanding of the nature of UAP.
Keywords: Unidentified Anomalous Phenomena; UAP; Pattern Recognition; Cluster Analysis; Typology; Characteristics
