Have a personal or library account? Click to login
Statistical methods for detecting and comparing periodic data and their application to the nycthemeral rhythm of bodily harm: A population based study Cover

Statistical methods for detecting and comparing periodic data and their application to the nycthemeral rhythm of bodily harm: A population based study

Open Access
|Nov 2010

Figures & Tables

1740-3391-8-10-1.jpg
Figure 1

Normalized crime rates and its Fourier approximations. Black dots show the relative frequency of 11402 crimes of bodily harm committed in the years 2002 to 2005 in Middle Franconia, Bavaria/Germany during the 168 hours of a week, starting Sunday at 9:00 a.m. Nycthemeral rhythms are visible. Solid line show the Fourier approximation of relative number of crimes versus cumulative time in hours, starting at 9:00 a.m. Light gray line shows the Fourier approximations of normalized crime rates for d1 to d5 (Sunday 9:00 a.m. to Friday 9:00 a.m.); the dark gray line for d6 and d7 (Friday 9:00 a.m. to Sunday 9:00 a.m.). A difference of these two rhythms is a shift of the maxima from 10:00 p.m. to 1:00 a.m. Furthermore the maxima of the second rhythm are higher than those of the first.

1740-3391-8-10-2.jpg
Figure 2

Distribution of crimes of bodily harm on the seven days of a week. Distribution of the 11402 crimes of bodily harm committed in the years 2002 to 2005 in Middle Franconia, Bavaria/Germany on the seven „days“ of a week, with 95% confidence intervals. d1 is the 24 hour timespan starting at Sunday 9:00 a.m. and ending at Monday 9:00 a.m. and so on.

1740-3391-8-10-3.jpg
Figure 3

Periodogram of incidents per hour. The periodogram is applied to the 35064 hours of the four year sampling period. The period of 168 hours (one week), 24 hours (one day) and their corresponding harmonic frequencies are tagged with circles. All high peaks of the spectral density coincide with these frequencies. The other shown periods have relatively small density.

1740-3391-8-10-4.jpg
Figure 4

Quantil-quantil-plot of residuals. Quantil-quantil-plot of residuals of d6 and d7 against standard normal quantiles (black cross). The gray line joins the first and the third quartile. The absence of large deviations between the black crosses and the gray line implies a normal distribution of the residuals.

Language: English
Published on: Nov 8, 2010
Published by: Ubiquity Press
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year

© 2010 Armin M Stroebel, Matthias Bergner, Udo Reulbach, Teresa Biermann, Teja W Groemer, Ingo Klein, Johannes Kornhuber, published by Ubiquity Press
This work is licensed under the Creative Commons Attribution 4.0 License.