RussR
En-Route
This was on a Facebook post (CFI Discussion Group), but I thought it would be an interesting discussion here too. The image was from an Avemco webinar.
Does anybody have any more information on this statistic?
This was my response to the FB post:
This is why I actually enjoyed Stats class.
Does anybody have any more information on this statistic?
This was my response to the FB post:
This is an interesting statistic, but also one that REALLY needs the background data in order to make sense and be useful. However, like any statistic, it's being used to prove a point, and undoubtedly Avemco's is that more frequent training is beneficial, and probably that pilots should get training at least every year. Which of course I believe in too, but the presentation of the statistic is still flawed.
- IF (and it's a big if) the pilot is getting the required "B"FR's, then the maximum number for the data set is 730.
- If the pilot crashes on a training flight, the number is zero.
- It's "landing accidents". Most of us do far more landings in rapid succession during training than at any other time. So the results are skewed toward zero.
- Even if the crashes were randomly distributed through the 24-month period, the average would be 365.
- So you could make the argument that everything is good because the pilots went longer between accidents than a random distribution would expect. (I wouldn't make that argument, but it could be made).
This is why I actually enjoyed Stats class.