I refer to the letter Ipoh police, we're fed-up with your statistics .

I personally have doubts over the statistics that the Polis Di Raja Malaysia (or any police force for that matter) releases in the press and whenever there is a conclusion that there is a rise or drop in crime rate without providing sufficient data.

Fine, to be fair, not all the data may be released at press conferences, but at least at their website ? After reading the above letter, I tried digging up some crime statistics from PDRM’s webpage, and it’s not there at all.

Whenever it is reported that the crime index has fallen, we must know what this ‘crime index’ is defined as. Is the data in this crime index solely dependent on reported cases only? Or does it cover unreported cases of which data can be attained by conducting a survey?

In the case of Malaysia, the concern of unreported cases is valid in view that the reputation of the police force is not good. Typically, trivial crimes would go unreported because it is not worth the hassle to go through all that bureaucracy.

However, it might be the case that increasingly, people are not reporting serious cases too. One of my family member’s warehouse was burnt down to the ground by unscrupulous thugs and that went unreported because these thugs were previously reported for illegal activities in that same area but resulted in police officers beating up the very same people who reported it.

In other words, the increasing occurrence of unreported crimes (of which data must be obtained for this statement to be true) would contribute to a drop in the crime index. Suppose, hypothetically, there is an actual increase of 1,000 crime cases from 2007 to 2008, yet there is also evidence that show an increase of 1,500 unreported crimes. The net effect is a drop in the crime index.

The other percentages that show a drop in violent crimes and crimes related to properties have to be dealt with carefully too. In statistics, it is analysing uncertainties. For the drop in percentage, it should be reported that there is a ‘significant’ drop, whereby a certain deviation is allowed from the previous year’s figure.

Meaning, if it is reported that there were 30% violent crimes in 2007 and suppose the standard error (which is the allowance for deviation mentioned earlier) is 3%, then, if it is reported in 2008 that violent crimes comprised 31%, then there is insufficient evidence to show that there is a rise in violent crimes – usually concluded as ‘it is possible there is no rise in violent crimes’.

However, if it is reported that violent crimes constituted 20% in 2008, then we can confidently conclude that there is a significant drop in violent crimes.

Nonetheless, base on my personal observation and experiences, I (and many other Malaysians) can confidently say that there is a need to think twice when going out in Malaysia.

Especially if I plan to walk or cycle to my destination. However, in Singapore, thinking twice is a waste of time and effort. The more important question there is to consider is which bus to take.