Included in this definition 2/ are: 



Restaurants 



Cafeterias 



Hotel dining roons and cafeterias 



Dru2 and proprietaiy stores with fountain seivice 



Lxmch counters 



Refreshnent stands 



Drinking places 



Excluded by this definition are: 



Soda fountains (not serving food) 

 Ice cream par].ors (not serving food) 

 Roadstands (not serving food) 



Also excluded are eating places operated by institutions and open 

 only to initiates, members or staff personriel. These would include establish- 

 ments such as restaurants or bars operated by cou:itry clubs, school cafe- 

 terias, hospitals, in-plant cafeterias, and establishments operated by the 

 Federal Governraent on milittiry posts. 



The respondent vjho answered the questionnaire used, which is shown 

 in Appendix A was any proprietor, manager, chef or any other qualified 

 representative of the establishment who was responsible for the pvirchase or 

 serving of food. The establislments which were covered in the suivey were 

 distributed in various areas of the United States and were selected on a 

 stratified, random (probability) sample basis. The 230 primary sampling 

 units or areas from which the establisl'iments were taken are shown graphi- 

 cally in figure 2, Further details about the nature of the sample design 

 for this survey are given in Appendix B. 



Reliability of the Itesu3.ts of Survey 



Quantitative data presented in this report such as purchases of 

 fish and shellfish are based on reported activities during a one week period 

 in May 195S» Any expansion of such figures to annual values is of necessity 

 of questionable reliability. In addition, all estimates shown in this re- 

 port are subject tc sampling and non-sampling errors of various magnitudes. 

 Although thei'e is no method for actually calculating the latter every effort 

 has been made to hold such errors to a minirmim. 



The sampling errors for the various statistics sho;«i in the sum- 

 maries can be calculated and are knoivm. Table 2 presents: information on the 

 magnitude of the sampling errors for estimates made from the national and 

 regional saniples, by type of establishiaent detail. Additional tables pre- 

 senting coefficients of variation for (1) aggregates for quantitative data, 

 (2) nmiber of establishments and (3) percentage of establishments reported 

 in tables throughout this report will be found in Appendix C, 



3/ Definitions of each type of establislunent included in the study will 

 be found in Appendix A. 



