Originally Posted by
Sunfish
However that argument fails statistically because you would need to compare AF to other GA flights operating in unfamiliar airports, which ATSB could not do. Furthermore that issue can be fixed by training and mentoring anyway. In any case the AF accidents were not caused by unfamiliar airports.
Then you raise “passenger carrying”, you would again need to compare AF with GA flights carrying passengers from A to B, not just tootling around the circuit on a sunny day or doing a local scenic flight.
You are suggesting that private GA has a higher accident rate for passenger carrying flights than non passenger carrying? It's certainly possible, the numbers would be interesting.
What you compare depends on the question you are trying to answer. It seems likely there was an expectation that the level of Angel Flight safety could be predicted by looking at the private GA accident rate. In that case you need to compare with the group you used to make the prediction.
Originally Posted by
Sunfish
However, even without those tests, the basic poisson and normal distribution tests for the accident rates can’t produce significant results or ATSB would have trumpeted them.
You really need to read the report. The data is there, even if it hasn't been trumpeted to your satisfaction.
Originally Posted by
Sunfish
What that means is that we DON’T KNOW if the two crashes are anything more than coincidence.
The same argument unfortunately applies to the RPT fatality rate in Australia. We cannot crow about how safe we are because the data is “lumpy”. You could go 40 years without an accident, then lose 500+ in a mid air over Sydney. If that ever happened, you can bet that the ATSB would be calling it an “isolated incident “ - statistically insignificant, random, etc. etc. They should afford the same latitude to Angel Flight.
One might be an isolated incident. 2 is more significant, if the expected rate is low. ATSB provide the numbers you can do the maths yourself if you want to. No need to speculate about hypothetical RPT accidents.
Don't get me wrong - I actually support the concept of Angel Flight. However, it appears they have a problem. One more similar accident and they are finished, I reckon. Denying the problem won't help them - they need to figure out how to stop it from happening again.