NCJ Number
19425
Date Published
1969
Length
88 pages
Annotation
WEATHER VARIABLES WERE STUDIED AS PREDICTORS OF POLICE INCIDENTS IN THE CITY OF FORT WORTH (TX) FOR A SIX MONTH PERIOD IN 1964.
Abstract
FREQUENCIES OF 31 CATEGORIES OF CALLS TO THE POLICE DEPARTMENT AND MEANS OF 13 WEATHER VARIABLES, AS WELL AS THE DAY OF THE YEAR, WERE COMPUTED BY SIX-HOUR INTERVALS. CORRELATIONS AMONG ALL VARIABLES WERE COMPUTED FOR EACH OF THE FOUR QUARTERS OF THE DAY OVER 178 DAYS, AND ALSO ACROSS ALL 712 QUARTER-DAY PERIODS; NINE CATEGORIES OF POLICE CALLS WERE SELECTED FOR FURTHER ANALYSIS. EACH OF THE FIVE MATRICES OF CORRELATIONS AMONG THE 14 PREDICTOR VARIABLES WAS FACTORED BY THE CRITERION FACTORIZATION METHOD USING SUCCESSIVELY EACH OF THE 9 SELECTED POLICE VARIABLES AS CRITERION. OF THE 45 MULTIPLE CORRELATIONS, 13 WERE SIGNIFICANT AT THE .05 LEVEL AND 10 AT THE .01 LEVEL. FOLLOWING VARIMAX ROTATION, THE FACTORS WERE MATCHED ACROSS TIME PERIODS FOR EACH CRITERION VARIABLE. FROM 2 TO 4 FACTORS WERE FOUND TO MATCH ACROSS AT LEAST TWO TIME PERIODS FOR EACH CRITERION. TWO FACTORS, TEMPERATURE-PRESSURE AND PRECIPITATION-FOG, WERE FOUND FOR ALL CRITERIA; A PRESSURE-WIND FACTOR OCCURRED FOR 8 OF THE 9 CRITERION VARIABLES, AND A PRECIPITATION FACTOR OCCURRED FOR THREE CRITERIA. METHODOLOGICAL PROBLEMS SUGGEST CAUTION IN INTERPRETATION, ALTHOUGH BOTH THE RESULTS AND THE METHODS USED APPEAR PROMISING FOR FURTHER RESEARCH.