PROCEDURE AND SAMPLE DESIGN 

 General Procedure 



After initial orientation, personnel assigned to this study visited 

 numerous fish wholesalers in Oregon, VJashington, and California for the 

 purpose of getting background information and to inform the wholesalers 

 about the study. Later, formal meetings were held with official indus- 

 try groups in Seattle, Portland, and San Francisco. At the suggestion 

 of the College, each of these groups set up an advisory committee to 

 work with the College and to encourage industry support for the study. 



Oregon State College personnel met with the members of the advisory 

 committees, as well as other wholesalers suggested by coiranittee members, 

 to learn more specifically of the problems confronting the industry and 

 to get suggestions on questions the industry desired ansvjered. 



Following the meeting irith the industry groups, preliminary ques- 

 tionnaires were developed (consumer, retail, and wholesale). These 

 questionnaires were then checked with members of the industry advisory 

 committees for corrections and suggestions. The questionnaires were 

 then sent to the Fish and Wildlife Service and the Bureau of the Budget 

 for approval. Permission to pretest was granted and the questionnaires 

 were pretested, revised, and returned to the Bureau of the Budget (via 

 the Fish and Wildlife Service) for final approval. Pinal approval for 

 conducting the field surveys was granted and the field xrork started 

 about February 1, 1956. The field work was completed about August 1, 

 1556. 



Sample Design 

 Household survey 



The primary sampling units and blocks for the househo3.d survey 'i^.ve 

 selected by the Market Research Corporation of America ac< follows: 



1, The tliree West Coast states ( Washington j Oregon, and Cali- 

 fornia) were stratified by city size and geographic area. The trxree 

 city sizes \-jere rural (\ander 2,500), urban (2, 500-50,000) and Standard 

 Metropolitan Area (SIIA) which included all the area in the counties 

 designated by the Bui'eau of the Census. The sample areas are shoxm in 

 appendix table 1. 



lljO 



