Retail survey 



The sample design for the retail survey vias worked out in coopera- 

 tion with the Bureau of Old Age and Survivors Insurance of the United 

 States Department of Health, Education, and Welfare as follows: 



1. The BOASI supplied Oregon State College with a list of all 

 retail stores in industry groups 51|1 (grocery stores) and 5U2 (meat and 

 fish markets) for California, Oregon, and V'ashington. The listing 

 included a stratification by county, industry groups, and nimber of 

 employees. The employee sizes were the same as those used in County 

 Business Patterns (United States Department of Commerce), and were the 

 number of employees for which social security funds were collected in 

 the first quarter of 1953. 



2. Upon receiving this listing, Oregon State College selected the 

 counties to be sampled (the same counties as used in the consumer sur- 

 vey), determined the sampling rates, and drew a systematic interval 

 sample, using a random start in each stratum. A total of UUO stores 

 were selected to give an allovjance for stores not handling fish. 



3. The sampling rate for each size classification was increased 

 with increasing size of store. A sampling rate of 1.7 percent was used 

 for stores with less than four employees. For stores with 1; to 19 

 employees the sampling rate was 2.2 percent, and for stores with 20 or 

 more employees, was 7.1 percent. Within each size- classification the 

 sample rate was constant across all strata. 



U. As part of the UUO selected stores did not handle fish, the 

 sample number was adjusted to yield the desired number of stores for all 

 strata by using a uniformly lower sampling rate. 



5. After the stores to be sampled vjere selected, they were sent to 

 the Bureau of Old Age and Survivors Insurance which in turn supplied 

 Oregon State College id-th a list of names and addresses for the sample 

 stores. 



A total of 260 retail stores was surveyed in the study. As differ- 

 ent sampling rates wei'e used for different sized stores, the field 

 records were duplicated at varying rates to give a self -weighting sample. 

 Thus in the analysis a total of 90U stores were used. 



The retail questionnaires xrere taken ty employees of Oregon State 

 College, except for the medium and small stores in the Los Angeles 

 metropolitan area. Mrs. Hazel Dugdale and Associates of Inglewood, 

 California, were hired to complete part of the retail survey in that 

 area. This saiae firm also worked on the consumer phase of the study. 



Wholesale survey 



The wholesalers selling fish to retailers and institutional outlets 

 in the three Pacific Coast states fell into txro general groups: (a) 

 wholesalers handling fresh, frozen and frozen packaged fish exclusively; 



IU6 



