Action- Use NOTECHS to Assess Crew Members in a Video Senario
Play a video scenario. (The ‘Brussels Engine Fire’ scenario from the JAR-TEL project
videos, if available, is often a good choice.) Ask the group to assess the non-technical
skills exhibited in the video. Compare and discuss the results.
The JAR-TEL project
The JAR-TEL project was funded by the EU with the objective of validating the NOTECHS
method for the evaluation of non-technical skills.
The hypothesis that JAR-TEL set out to validate were:
1) The NOTECHS method can be used to asses true variations of nontechnical
skills based in observed behaviour of flight crews.
2) The NOTECHS method is culturally robust with respect to variations in
nationality, organization and personal background
3) The NOTECHS method is practical so that it can easily be used by
training and check captains.
Only if these hypotheses can be confirmed can the NOTECHS method be reliably used in
the assessment of crews non-technical skills.
Participants and structure of the JAR-TEL project
The major sponsor was the EU. Sofréavia, a French research institute, was the project
leader and BA, NLR (Dutch Research Institute), DERA, DLR (German Research Institute),
Airbus, Alitalia, the University of Aberdeen and Imassa were associated as contractors or
partners to the project.
The study was formulated to encompass a representative sample of airlines from all
regions of Europe. A group of 100 instructor pilots was used to apply the proposed
method of evaluation to eight different scenarios filmed in a simulator. Each scenario was
around 10 minutes in duration and featured different degrees of NTS and various nontechnical
behaviours.
The robustness of the system was assessed according to three different measures.
1. Internal Consistency
Ratings at Category level should agree with Element level. A high level of consistency was
found.
2. Accuracy
Participants ratings should match the ‘correct answers’ formulated by the scenario
designers and expert raters. A high level of consistency was found, at the category level.
In the more ambiguous scenarios, the level of consistency was reduced. The largest
divergence from the expert results was that the experiment participants tended not to
use the rating not observed, even when the expert rating was not observed.
3. Inter-Rater agreement
The extent to which the different participants ratings agree. A high level of consistency
was found, at the category level.