Hunting activity record-cards and the accuracy of survey estimates

Sending hunting activity record-cards prior to a hunting season has been recommended as a methodology for reducing digit preference (heaps on numbers ending in 0 and 5) and improving the accuracy of participation and harvest estimates derived from survey responses. This article examines the extent to which record-cards influence the number of 0–5 responses and contributes to changes in mean days of duck hunting and harvest. Data were obtained from the 1999–2000 Illinois Waterfowl Hunter Survey. Approximately half of the respondents (n = 1, 717, response rate = 72%) were mailed the preseason record-card; the other half of the sample (n = 1, 430, response rate = 61%) were not mailed the record-card. Record-card recipients reported less 0–5 heaping, but excessive 0–5 heaping explained only 20% or less of the change in means. Changes in mean estimates were primarily the result of more reporting of low days of participation and harvest by record-card recipients. Sending out record-cards prompted more individuals in a sample with low days of participation and low harvest to become respondents. Given that record-cards are relatively inexpensive, the authors encourage state fish and wildlife agencies to adopt this methodology.