Analysis 101: Time Series Analysis Answers
From your first chart you should have easily
identified that the riskiest time of day to
travel the roads in your area was between 0600
and 0900. You should probably recommend that
staff avoid travel during this period at least
until you have a better idea of what is going on.
From our second chart we can see a cluster of
incidents on Fridays (with one very early
Saturday incident). So essentially we have two
distinct groupings. The first group is comprised
of weekday morning attacks while the second group
are the ‘Friday’ attacks. There are a couple of
obvious hypothesis; either two separate groups
with differing modus operandi are responsible for
the incidents or the clusters represent two
different sets of intended targets.
At this point we should probably start looking
for more collateral information. What potential
targets are on the road during these periods?
What other events occur at these times of the day
or week? We also need to look at exactly where
these events occurred but that’ll have to wait
for a future instalment of Analysis 101.
For those who are wondering this scenario is
based on a real problem from my past. The details
have been changed just enough to keep me out of
trouble.