The interviewer was instructed to call on the sample 

 households and any other household (not shown on the list) found 

 between the sample household and the next listed, and to inter- 

 view the meal planner in all these households. Where no one was 

 at home, or the meal planner was not at home on first call, a 

 second call was made on a different evening. 



In rural areas, the field worker who did the listing 

 had a map of the cluster on which he entered the location of every 

 dwelling within the cluster by the use of a small square (D) and 

 identified each by a sketch number. On the listing sheet, the 

 sketch number than served as the primary identification of the 

 dwelling in addition to a description of the house and its location. 

 After the s ample units for the survey were designated on the list- 

 ing sheets, the interviewer could easily find them by reference to 

 the sketch numbers and the sketch map. 



Accounting for the Non-Interviews 



There was a total of 3,198 occupied sample households. 

 These were accounted for as follows: 



Total sample households . 3*198 



Less ineligible households^/ 35 



Eligible sample households 3,163 100. p£ 



Refusals lk$ U.6/> 



Not-at-home after 2 calls U?5 lg.Off 



Total non- interviews 620 19.6% 



Completed interviews 2,$U3 80. 1$ 



The not-at-homes were accounted for by a special weighting 

 procedure based on the "nights-at-home" record of respondents found 

 and interviewed. This technique accounts for all individuals who 

 were, in fact, at home on any of six evenings - including and just 

 preceding the evening of the interview. 



However, the weighting procedure makes its own contribution 

 to sample tolerances. To reduce this contribution and at the same 

 time retain the benefits of not-at-home weighting, this survey em- 

 ployed a weighting plan based on a combination of the weighting with 

 one call-back on each respondent not found at home on the interviewer's 



1/ These are households in which the main meal was not taken at 

 home once a week or more. No interviews were to be obtained 

 in such households. 



302 



