"It is critical to realize that data produced from a survey are inextricably tied to the methods used, and that changes in methods are inevitable.

"This project was the result of an accidental discovery which occurred when comparing the comments on original, handwritten data sheets from earlier surveys to experiences in the field during modern surveys.  Our analysis showed that significant differences in catches between surveys are most likely due to differences in bottom trawling methods and may not accurately represent true trends in abundance.  These observed trends, however, were accepted as true trends in abundance, and most likely would have persisted, especially with the impending transfer of the survey from the AFSC to the NWFSC.  The reason this problem remained undiscovered is that we did not formally ask questions of our own survey; we concentrated on dragging the bottom trawl on the ocean floor and accurately recording the area swept, without recognizing that our increasing success in accomplishing these goals changed what we were catching.  We provided our methods and data summaries to stock assessment scientists without analyses of the changes in methods, and assumed that any critical analysis of our data would be performed by the stock assessment scientists.  On the other hand, it appears that most stock assessment scientists relied on the field biologists to perform any critical analysis of methods wich might affect abundance trends, and did not understand the implications for the changes in methods that we reported."