A Replication Approach to Controlled Selection in an Intercept Survey of Marine Recreational Fishing Trips

Intercept surveys of marine recreational anglers at fishing access sites can be costly and present logistical challenges.  Limited resources and increasing demands for data can put survey designers and managers in the difficult position of balancing the need for interviewing productivity against adherence to probability-based designs.  The U.S. National Marine Fisheries Service (NMFS) was in this position with its own intercept survey, conducted along Atlantic and Gulf of Mexico coasts since 1982.  Under the original design, sample assignments were often rescheduled for several reasons including availability of field samplers.  This flexibility generally did improve the operational performance of the survey at the cost of deviations from probability sampling and potential for bias.  In 2013, NMFS implemented a redesigned Access Point Angler Intercept Survey (APAIS) to address deficiencies identified by the U.S. National Research Council.  Under this design, many issues were eliminated by requiring strict adherence to the sample draw.  However, strict protocols meant that solutions to logistical constraints must be implemented during sample selection to produce a draw that was logistically feasible and a sample that did not violate probability-based sampling principles.  This paper describes the sample selection method developed for the 2013 APAIS which incorporates logistical constraints through replication.