Ten pilots assess and rate of a new landing assistance system under differing stressed situations. What confidence level can be postulated that the cockpit experience mimics the real world conditions adequately well? I would drill down on the validity of having 10 (small sample size, experimental design) and defining confidence levels p-tests given the samples).

A longitudinal data study. Experienced pilots report on X landing scenarios at Munich airport. Their opinions are garnered via i) a structured qualitative study and a ii) free-form 15 minute discussion. The dataset was derived from instrumenting the same scenarios as the pilots used. Oscars work and the questionnaire produced.

An initial research question is “What interpretations can be found from a machine produced dataset?”. Follow-up questions are Can the quantitive data predict the pilots and indeed the planes’ responses?

Data-driven forecasting. Pilots state that they strive to be ‘ahead of the plane’. For events that are not themselves pilot induced, can any corrective measures in response to an unstable landing be?