1983 02 00 Perceptual and Motor Skills - Vol 56 No 1 - Persinger

Magazine Overview

This document is a scientific article titled "GEOPHYSICAL VARIABLES AND HUMAN BEHAVIOR: VIII. SPECIFIC PREDICTION OF UFO REPORTS WITHIN THE NEW MADRID STATES BY SOLAR-GEOMAGNETIC AND SEISMIC MEASURES" by M. A. Persinger, affiliated with Laurentian University. It was published in the journal Perceptual and Motor Skills in 1983, Volume 56, pages 243-249.

Summary of the Article

The article presents a study that aimed to predict the number of UFO reports in the six states surrounding the New Madrid area in the central U.S.A. using seismic and solar-geomagnetic variables. The research utilized data from 1952 and 1969, with 6-month intervals, to develop optimal prediction equations. These equations, which involved multiple lagged variables and achieved high correlation coefficients (rs larger than 0.83), were able to predict major "UFO flaps" and other minor increases in UFO reports that were not immediately evident in the observed measures. The study also presents predicted UFO variations for the years 1970 to 1982.

Hypothesis and Previous Findings

The central hypothesis posits that UFOs are natural luminous phenomena linked to tectonic strain within the Earth's crust. This hypothesis has been empirically supported by previous research by the author. Bivariate analyses had previously shown that UFO reports clustered in earthquake-prone regions and increased in the six months preceding increased numbers of low-intensity earthquakes. Subsequent multivariate analyses indicated that a combination of seismic and solar-geomagnetic variables could explain over 80% of the variance in UFO reports over a 17-year period.

Methodology and Data Sources

To test the predictive capacity of the hypothesis, a third series of analyses were conducted using data not previously involved in the initial computations. The study area, temporal increment, and data source were specifically chosen to verify the intrinsic consistency and general applicability of the concept. Monthly UFO reports from 1951 to 1977 were obtained from the UFOCAT file, covering Illinois, Indiana, Kentucky, Tennessee, Arkansas, and Missouri. Seismic data, including different intensity levels (MM, Modified Mercalli), were collected for these six states and the surrounding 16 states. Geophysical data, including yearly U.S. Earthquake publications, Journal of Geophysical Research for geomagnetic and solar data, and preliminary earthquake data from the U.S. Geological Survey, were also utilized. Solar-geomagnetic data were calculated from the Preliminary Report and Forecast of Solar-Geophysical Data (NOAA) and Solar Bulletin.

Variable Construction and Analysis

Six-month analysis intervals (January-June and July-December) were selected based on previous findings. Various variables were constructed, including six-month totals of seismic events of different intensities within the New Madrid states (MALOW, MVI, MVII) and surrounding states (EXLOW, EXVI, EXVII). Percent changes in these seismic variables were also calculated (MALOCH, EXLOCH, ALLOCH). UFO report numbers were represented by absolute totals per 6-mo. interval (UFOSIX) and its square root (SQCUF). Geomagnetic activity was measured using monthly Ap indices, from which variables like APAVE (mean of six monthly averages), APMAX (mean of the single most intense daily Ap index), AAVSD (simple variation), APACH (relative variation), and MAGPOW (mean-weighted variation) were derived. Sunspot numbers (SUNSQ) and their percent change (SQSCH) were also included. All calculations were performed using SPSS software.

Statistical Modeling and Results

Step-wise multiple regression (REGRESSION) was the primary analysis procedure. For each dependent variable (CUFSIX or SQCUF), various combinations of lagged variables were entered, including seismic measures from the six states only, surrounding states only, combinations of both, and all of the above plus geomagnetic and solar variables. A cut-off criterion of a greater than 5% change per variable or a maximum of seven variables was used for selection. A second analysis was performed using only these selected variables.

Table 2 presents four optimal sample equations for predicting UFO report measures within the New Madrid states. These equations involved different combinations of lagged variables (e.g., APAVE4, TVII3, AAVSD, MAGPOW2) and achieved high multiple r values (MR) ranging from 0.84 to 0.88. The adjusted multiple r² (AMRS) values were also high, indicating that these variables explained a significant portion of the variance in UFO reports.

Figure 1 visually compares the predicted and observed percent variations of mean UFO report numbers for the years 1965 to 1982. The study notes that the equations generally predicted the beginning of the 1973 UFO flap and the subsequent reduction in reports until 1977. The square root equation (QSQ4) demonstrated the best fit for both the data used in its calculation and later observations. Notably, one equation (RSQ2), calculated without data after 1965, accurately predicted elevations in reports in 1966-1967 and the 1969 and early 1973 flaps. This suggests that the elevation in UFOCAT report numbers might reflect processes beyond simple data sampling.

Predictions and Future Implications

The study consistently predicted elevated UFO reports in the New Madrid states for intervals in 1981 and the first half of 1982. The authors suggest that with a lag of at least two increments and the contemporary variable AAVSD, these equations could be useful for future predictions, potentially allowing for UFO report prediction up to a year in advance by assuming AAVSD values.

Notable Patterns and Conclusions

A frequently emerging pattern involved equations with multiple Rs exceeding 0.80 that included first entries of MVII3, MVII7, or TVII3. These variables represent seismic events of intensity VII or higher. The study concludes that, in a conservative sense, the equations are predicting variations in UFO reports collected by CUFOS. If the CUFOS sample is a reliable measure of the total population of reports and if these reports accurately reflect actual physical events, then further analysis of the independent variables could help isolate the underlying mechanism. At the very least, these unusual responses can be predicted using patterns of geophysical stimuli.

Recurring Themes and Editorial Stance

The article aligns with a scientific and empirical approach to understanding UFO phenomena, treating them as potentially natural events linked to geophysical processes. The editorial stance of Perceptual and Motor Skills appears to be open to publishing research that investigates unusual phenomena through rigorous statistical and scientific methodologies, even if the subject matter is unconventional. The recurring theme is the exploration of correlations between geophysical variables and human behavior, specifically UFO sightings, aiming for predictive modeling and the isolation of underlying mechanisms.