Viewing page 62 of 134

This transcription has been completed. Contact us with corrections.

- 52 -

An Evaluation of the Interview

An attempt was made to obtain a rough estimate of the degree to which the interviewers in this survey reliably recorded information given to them by the airline pilot interviewees. It will be recalled that for the purpose of training the interviewers, supervisors were asked to conduct several interviews in which one interviewer observed another and recorded the information independently. These interviews provided the data for this evaluation. Only 15 such interviews were obtained. The recorded information obtained from those question asking pilots for examples of critical incidents provided the data upon which was based the comparison of the records of the interviewer and his observer in each of the interviews. The 15 interviewers reported 45 incidents. There were five different kinds of information which interviewers were requested to obtain about each incident: (1) type of plane, (2) time of incident, (3) flight reference (instrument, contact or hooded), (4) phase of flight, and (5) behavior of the pilot. For the 45 incidents, this made a total of 225 items upon which to base a comparison of the reporting of each interviewer and his observer. There were a few incidents which analysis showed were really double incidents. Although these were reported by the interviewers as only one incident, each contained more than five items, i.e., two examples of pilot behavior. These double incidents raised the total number of possible items to 240. The results of the comparison are as follows:

EXTENT TO WHICH THE INTERVIEWERS WERE ABLE TO GET THE DESIRED INFORMATION

[[3-columned table]]
|   | No. | % |
| --- | --- | --- |

| Items which both interviewer and observer reported | 157 | (65.4%) |
| Items which neither interviewer nor observer reported | 54| (22.5%) |
| Items which one reported and the other didn't | 29 | (12.1%) |

EXTENT TO WHICH INTERVIEWERS AGREED ON INFORMATION REPORTED BY BOTH

[[3-columned table]]
|   | No. | % |
| --- | --- | --- |

| Items which both interviewer and observer reported | 157 | (100%) |
| With complete agreement | 155 | (98.7%) |
| With disagreement | 2 | (1.3%) |

From these results it would appear that interviewers agreed on items which they reported, but that one occasionally failed to record items which another did record. Furthermore, it would appear that the interviewers were only 65% effective in obtaining from the pilots all of the information desired about an incident. This may not be solely the function of the interviewers' techniques or the interview method itself, inasmuch as it was found that some pilots were reluctant to relate incidents which were brought on by some act of their own.

Transcription Notes:
7/26 -- corrected table formatting Reopened!! Took out indentations.