A Comparison of Longitudinal and Cross-Sectional Mail Survey Designs for Estimating Annual Participation in Marine Recreational Fishing

Surveys of recreational fisheries have used both cross-sectional and longitudinal designs to collect data to estimate the number of fishing participants (participation).  A cross-sectional survey design contacts respondents once at one specific point of time.  For annual estimates of participation, a cross-sectional survey typically collects data from respondents on their fishing activities during the past 12 months.  One of the major concerns of this type of data collection is the potential recall bias.  Because a recall period that is considerably shorter than 12 months is believed to be more accurate, a longitudinal survey design that re-contacts respondents to report their activities for a number of shorter recall periods is therefore considered superior.  The potential drawbacks of a longitudinal survey are an increase in survey complexity and cost because respondents need to be tracked and re-contacted multiple times.  In this study, we compared longitudinal and cross-sectional survey designs for estimation of annual marine recreational fishing participation.  The results indicated that the cross-sectional survey estimates based on 12-month recall do not significantly differ from the longitudinal survey estimates based on 4-month recall.  This suggests that a cross-sectional survey design is a cost-effective alternative for estimating the annual participants in marine recreational fishing.