Standardizing the Display of IBR Data: Time and Date Variables
Frequencies and Rate of Missing Data for Time and Date Variables
Frequencies at a Glance
|Rate of Missing?||0% (required)||0% (required)||4% (also unknown)|
|Most Common?||August||1st of the Month||5 - 7 pm (not including 0)|
Frequencies of Time and Date
The 1999 NIBRS file contains 2,139,433 incidents from 18 states. Month and date are required, so there are no missing data for these elements. However, when a time is unknown, it is left blank. 4% of offenses are missing an entry for time of occurrence; however, it is impossible to differentiate between missing and unknown entries.
The frequencies for offenses committed each month are presented below.
For all offenses, the month of occurrence seems rather evenly spread out, with offenses more likely to occur in July and August than in February. When offenses are compared, the pattern is similar for violent (murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault) and nonviolent offenses, with the biggest differences occurring in December and January, when violent offenses increase and nonviolent offenses decrease.
When dates are compared, crimes are much more likely to occur on the 1st day of the month than on any other day. Perhaps this is due to the payment of assistance checks, which are usually received on the first of the month. Although it appears that offenses decrease at the end of the month, this may be due to the reduced number of months with 31 days rather than an actual decrease in reported offenses.
As the following graph shows, this pattern is consistent for both violent and nonviolent offenses.
As demonstrated in the Data Collection Guidelines and Data Quality section, the time when incidents occur shows a different picture. Although the peak at 0 is probably due to a coding error, offenses are much more likely to occur in the afternoon and early evening than in the morning hours. Peaks occur at 8 am and noon, periods right before school and during school lunch.
When violent and nonviolent offenses are compared, the pattern is slightly different. Violent offenses are more likely than non-violent offenses to occur during the evening and early morning hours (from 7 pm to 5 am). Note that due to the large spike at 0, this hour has been removed from the graph presented below to better display the differences between offenses.