RA3,

As long as data is collected and recorded by humans, there will be errors. As for outright lying, I don't know what the possible incentive could be, but I'm not going to debate that point. Punchcard returns never comport with creel census results. We have know for at least 30 years that punchcard data is not very accurate. That is one of the main reasons for doing creel surveys. Creel surveys, if they follow a reliable statistical sampling design, can provide higher quality estimates of catch and angler effort, and they can provide the information much sooner than waiting for the less accurate punchcard data. My main point is that no one should have been surprised that the punchcard results and creel census data for the EF Lewis provided disimilar estimates.

Sincerely,

Salmo g.