To examine the relationship between the 'seat' variable and the attitude items, we performed T-tests on all of the attitude items looking to see if the means for captains and first-officers were significantly different. The tests showed significant effects with respect to three of the attitude items.
On the ``company pressure'' probe, both captains and first officers tended to disagree with the statement. Captains gave it an average response of 2.32 and first officers had an average response of 2.49. The difference is significant (p< .05).
On average, our pilots disagreed with the item ``not reduce workload''; grand mean 2.7. First officers disagreed significantly more than captains did. (2.56 to 2.85, p<.002.)
Overall, our pilots agree with the ``adequate training'' item (mean response is 3.5). First officers agree with this item significantly more than captains do (3.6 to 3.3, p< .003). It is difficult to be sure what this means, however, because the probe is in the form of a comparison and we have no knowledge of the other training experiences to which the subjects are comparing their 757/767 training. (But see the correlation analysis of the ratio total/auto hours to this item below).
The simple T-test analysis also indicated a significant relationship between seat and the probe, ``It is easier to bust an altitude in an automated airplane than in other airplanes'' (p<.04). However, because this relationship is not strong and because T-tests can be problematic when the distributions being compared are not nearly normal, we also conducted a non-parametric test of all of these relationships. The Wilcoxon (rank sums) test indicated a significant relationship for the ``company pressure'', ``not reduce workload'', and ``adequate training'' probes, but not for the ``altitude bust'' probe or any of the other probes.