Originally Posted by
Std Deviation
IMO, the whole DPE system is flawed. What does it indicate: On that day, in that aircraft, under those conditions, with that examiner, my private pilot applicant performed (or not) for a three hour period. I did 3000+ hours of primary instructing (gold seal CFI) and had applicants that I figured would pass with flying colors - nope, pink slips. Others that I did all I could for, and hoped for the best - passed without a glitch.
Then you have the financial incentive issue. In the 90's some examiners were making 100K a year just giving checkrides (when a private exam was $125). I don't know what the answer is. But after flying with someone for 40 hours you get a better glimpse at behavior and traits than a quick look see.
ACA had a pretty high failure rate on the CRJ at one point as well. To them it was the space shuttle. They gave us maintenance manuals at indoc. At JetBlue I get a lot of, "you don't need to know that." Which is great, but being old school, I'm still wondering what the answer is sometimes.
Agree 100%. Scientists and experienced flight instructors (career instructors) wince at the testing methodology used by the FAA to test pilots. A basic knowledge of statistics will tell you that a reasonable level of certainty, say 95%, cannot be achieved with less than about 30 samples taken from a purely random, consistent population. By contrast, the FAA has a pilot candidate taking a single checkride that will supposedly be representative of years of future aircraft operation, based on only one flight. No wonder airplanes crash.
A case study can be found in any college course- an instructor gives multiple closed book exams per semester, 4 or 5 usually, plus a final exam, plus they usually throw in a weighted score for homework grades, and after all that is calculated a final grade is established using a curve scatter because it is known that natural logarithms apply. Even with all that care and effort, students still complain "oh I got a B when I deserved an A". Imagine how random it would be if a final score were assigned using only one exam! At best you would get a ballpark idea such as, this candidate is alive or dead. Worst it is inadequate testing based on one (1) personal opinion on one (1) particular morning or afternoon.