RESULTS 



TEO-tvJICAL CONSIDERATIONS IN INTERPRETING DATA 



Species groups 



The number of observations for a species was an important consideration in determining 

 the selection of species groups for the data tables. The catch of a single species is reported 

 only if the number of observations exceeded a certain threshold. In some cases it was possible 

 to combine several closely related species and report the catch for these species as a group. 

 Species with fewer than the threshold number of observations which could not be combined 

 with closely related species were put into the species group "Other Fish." Exceptions to these 

 procedures were made in cases of economically important species such as striped bass. A list 

 of species included in each species group is included after each year's data. Care should be 

 exercised when comparing the tables with previous years as the species composition of some 

 groups may have changed slightly. However, all species estimates are maintained individually 

 in the MRFSS data base. 



Cooperative data collection 



In 1986 the basic MRFSS effort was enhanced through the cooperative participation of 

 State and Federal agencies. The States of Florida, Georgia, Louisiana, New York, and Virginia 

 either reimbursed NMFS for additional interviewing effort or used State personnel to collect 

 an increased number of intercepts. These cooperative efforts provided the participants with 

 needed data while taking advantage of the cost savings and methodology of a major survey. In 

 addition, Texas Parks and Wildlife Department collected and provided data to NMFS in lieu of 

 NMFS conducting the MRFSS in Texas. Interviews that would have been conducted by the 

 MRFSS in Texas were reallocated to the remaining Gulf States. 



Statistical considerations 



The procedures used in the surveys to estimate the various parameters of interest (e.g., 

 number of fish caught, number of participants) used ratio estimators. The reliability of ratio 

 estimators is directly related to sample size; as sample size decreases, the amount of bias in 

 the ratio estimate increases. No appreciable bias was apparent for any State in 1986. 



11 



