This subject is not amenable to hearsay or personal convictions, period.
‘’There is an old and rigorous method of determining if there is an increase or decrease in safety levels over time; that method is called statistical process control. It was developed at least as early as WWI to assure the quality of safety sensitive products such as explosives and munitions and is based on straight forward and well understood mathematics.
‘’What is required is either an estimate or actual activity (eg. a9hours flown, number of flights). An estimate or actual numbers of accidents and/or incidents over time, in other words a history.
‘’From these numbers an average and standard deviation are calculated. We can then apply the normal distribution and calculate the three sigma limits for the distribution of incidents and accidents. This is the computed number within which 99.95%(im getting old) of all statistics will fall. That may be for example, 6 accidents per quarter plus or minus three. It is only after there have been 9 (6 + 3) accidents can we say Statistically that we can bev99% sure that something has changed.
‘’Of course there is more to it than this but you get my drift. I’m sure RAA know this and in any case ATSB can advise them on design and computation of the relevant data..
‘’In other words, we don’t know if four fatals are significant or not without mathematical analysis of our accident history; we just do not know.
‘’There is therefore no sense in getting knickers in a knot about the overall state of the pilot population just yet.
‘Having said that, if poor weather decision making appears to be a common factor in a relatively large number of incidents (statistics again) then it is possible to test that hypothesis (maths again) by developing a questionnaire, taking a sample of RAA, PPL and commercial pilots and testing their skills. The results of such an exercise then allows RAA and CASA, advised by ATSB, to change the syllabus and training on the basis of fact, not guesswork.
‘’Having done that - changed the training, you would then watch for a change in quarterly (or annual, or monthly) statistics (mean and standard deviation) to determine if your training was effective.
‘’Been there and done it, It ain’t that bloody hard. (at least when you are young).