While it was expected that some of the probes would convey positive attitudes and some negative attitudes toward automation, it was not known that this positive/negative split would be so pronounced or that it would map so cleanly onto the first principal component of the attitude cross-correlation matrix.
The second principal component of the attitude cross-correlation matrix was a surprise. It appears that in their responses to the probes, the pilots treat probes that tap representations of their authority similarly. The distinction between hand-flying the airplane and managing automation is a salient one in the industry, but we did not expect to see it emerge from the analysis of the structure of the attitude probes. The interpretation of this dimension is complicated because it involves issues of authority, direct versus mediated control, and the context of automation usage.