
Times of day
Crime and disorder records contain time information relating to when the crime
occurred. For personal types of crime or disorder events this is usually a specific
time. This time can be aggregated into one hour intervals to reveal times in the
day when these events peak.
It is important to consider the time and date range that may exist in certain crime
records. This is particularly the case with property crimes where there is often
a reported ‘committed from time’, committed from date’, committed to time’, and a
‘committed to date’.
Time of day range analysis.
A suitable method that incorporates the time and date range in an analysis of hotspots
in times of the day is referred to as the weighted method. This is not a simple averaging
method between the ‘from’ and ‘to’ times, but instead assigns probabilities to each
one hour time interval within the range that relate to when the crime may have occurred.
These probabilities calculated for all crime records are then summed for each one
hour time interval to reveal trends in the change of crime over a typical day.
Consider the following crime record examples.
|
|
Committed
from time
|
Committed
from date
|
Committed
to time
|
Committed
to date
|
|
Residential
burglary 1
|
20.15
|
12/01/2001
|
00.25
|
13/01/2001
|
|
Residential
burglary 2
|
09.30
|
13/01/2001
|
10.45
|
22/01/2001
|
|
Residential
burglary 3
|
16.10
|
14/01/2001
|
16.50
|
14/01/2001
|
Residential burglary record 1 spans 2 days (Friday 12th January to Saturday 13th
January) and a five hour interval range. Record 2 has a range of 10 days. The probability
for this date and time range is calculated by working out how many one hour intervals
have spanned this 10 day range. Residential burglary 3 occurred on the 14/01/2001
between 16.00–17.00. Probabilities that relate to the times of day when the burglary
occurred can be calculated from these ranges. These probabilities are shown in
the linked table.
The sum of the individual weights for all crime records is then calculated to reveal
how crime changes over the duration of a day.
Partnerships have often simply used the committed
from time or the committed to time to describe times of day when property crimes
occur. This can create misleading results as these times usually reflect human
behaviour patterns (e.g. when a person left their house to go to work and when
they returned). Analysis of residential burglary in the London Borough of
Croydon revealed that if the committed from time was used, peaks in residential
burglary tended to occur between 8.00-10.00 and 12.00–13.00.

If the committed to time was used,
then the main peak that was between 17.00-20.00.

The weighted time method uses probabilities to
explain peaks in the times of day when crimes occur where a time range exists.
Using this method, patterns during the day when residential burglaries occurred
could more accurately be identified (see graph 3). This information can then
contribute to helping in the design of residential burglary reduction
initiatives and the deployment of patrols over shifts.

Click here for an enlarged version
|