J 



2. Saniple Design 



Requirements for a sajnple design . It was recognized that a sample 

 design for a study of insurance experience of commercial fishing vessels should 

 fulfill the following requirements. 



First, the design should be flexible enough to allow continuous 

 revision of plans in order to meet effectively unforeseen difficulties and 

 keep costs to a minimira without seriously impairing reliability. 



Second, the sample-unit should be the commercial fishing vessel. 

 Information about insurance experience should be collected by intervie^rfjig 

 the vessel oimer and studying the insxirance record of the sampled commercial 

 fishing vessels in each specified region. Whether a vessel carried insurance 

 or not was necessarily a part of the stucfy. 



Third, for satisfactory results on claim experience the vessel 

 population should include all active and lost vessels for the years 1950-195U. 



Fourth, the need for a qualitative analysis required emphasis on 

 depth investigation of cases (sample-ujiits) while statistical measurements 

 of precision should be considered to have limited applicability. 



The final sample design . In vievj of the above requirements, a 

 sample was designed which provided for area-imiltiple-stage stratified 

 sampling of commercial fishing vessels in order to study insurance experience 

 for a five-year period, from 1950-through 195U. Randomization was applied 

 in the selection of sample-units in so far as feasible and investigation of 

 the New England situation was to be used as a quasi-pilot study. The new 

 sample design was recommended to the Fish and Wildlife Service of the U. S. 

 Department of the Interior and subsequently approved by the Bureau of the 

 Budget. 



Sampling on this basis involved the folloiving major steps: 



(a) Selection of the statistical universe . The sample-unit of 

 the population was defined as a vessel vjhich (1) vjas of $ net 

 tons or more; (2) was active or lost during the period from 

 January 1, 19^0 to January 1, 1955; (3) was engaged in catching 

 fish or shellfish for commercial purposes; (ii) was registered in 

 any port located in the specified regions; and (5) whose owner- 

 ship did not change during the five-year period. 



From the annual listing of Merchant Vessels of the United States , 

 Bureau of Customs, U. S. Treasury Department, vessels of 5 net 

 tons or more which were active as of January 1, 1955, which had 

 as port of registration any port located in the six regions, and 

 whole service was fishing, cod fishing, or oystering were selected. 

 Vessels belonging to the sixth region, Inland Waters, were merged 

 with the Middle Atlantic Area. Then, in order to meet the above- 

 mentioned definitional characteristics the initially available 

 universe underwent a number of amendations. First, the universe 

 was augmented with vessels which were lost daring the period 

 from January 1, 1950, to January 1, 1955. Second, the vessel 



318 



