I would like to reply to George Debono’s and Jim Wightman’s responses to my letter, which appeared in The Sunday Times of Malta (October 22). Again, they were very unfair in my regard.

In a letter in The Times on Sunday dated October 1, I corrected the misconception that road death or accident rates in Malta are climbing. I consider it my civic duty to correct inaccurate reporting of data in any form, especially when such is used to justify interventions which may have negative impacts on citizens, in this case road users.

My factual letter triggered very personal reactions, again last Sunday. I have been accused of misrepresenting facts and defying logic, of considering the lives of children as expendable and serious injuries as inconsequential. I will respond to such dehumanisation of my person with scientific argument.

If one looks at the official figures for road deaths in the years from 2012 to 2016 (11 in 2012, 18 in 2013, 11 each in 2014 and 2015, and 23 in 2016), a clear trend does not emerge. However, one may be tempted to interpret variations due to the small number effect as real effects. This is wrong, since the serious injuries on our roads (297, 265, 291, 306 and 293 year after year, in sequence) clearly do not follow any such trend. It is unreasonable to assume that any factor may influence fatalities but not road injuries, especially without any supporting evidence.

To put any alternative argument to rest, I calculated the standard deviation of both trends. The standard deviation is a mathematical measure of the natural variation due to chance. The standard deviation of the road death fatality figures above is approximately 5.5, while that of injuries is 15.3. As such, in the 2012 to 2016 period of observation, any road death rate in the range between four and 26, and any road fatality figure between 260 and 320 (rounded off), is within the range of variation to be expected, due simply to chance (i.e. within 1.96 standard deviations of the mean). The assumptions of the independence of such observations and the normality of the distribution do not strictly hold, but, if anything, a non-parametric test would give a wider confidence interval.

Road deaths and road injuries have not increased (or decreased) in the period of observation between 2012 and 2016, beyond the natural variation which could be expected. Stop shooting the messenger.

I will not enter into the other arguments presented. However, allow me to state that one cannot argue that the relatively higher death rate for cyclists in The Netherlands is due to people dying of natural causes, if one should believe that cycling is healthy and cyclists live longer than car drivers. One cannot argue that the road death rate in Malta is artificially low because we do not have motorways, when the road death rate on motorways is much lower than the corresponding urban statistic.

I would expect a formal and unreserved apology for this personal attack. I also encourage correspondents who believe that “one death on the road is one too many,” to encourage cyclists to wear all protective gear, abide by the law, and avoid the unnecessary use of major roads.

Editorial note: This correspondence is now closed.