APPENDIX A. 

 General Explanation 



Page 



TAKING THE CENSUS A 1 



COMPUTER PROCESSING A-1 



OFFICE PROCESSING A 2 



DEFINITIONS AND EXPLANATIONS A 2 



FARMS CLASSIFIED BY SPECIFIED CHARACTERISTICS A-4 



ENUMERATION PERIODS A 5 



USE OF SAMPLING AND STATISTICAL ADJUSTMENT A-5 



NONSAMPLING ERRORS A-6 



CENSUS COVERAGE A-6 



RELIABILITY OF CENSUS ESTIMATES A-6 



TAKING THE CENSUS 

 Report Form 



One report form was used in the 1982 Census of 

 Agriculture for Puerto Rico. Except for a few items, the 

 1982 report form was very similar to the one used in 

 1978. Most of the changes have resulted in simplifica- 

 tion of the questions asked. Also, the two sections con- 

 cerning interplanting of crops and selected crops under 

 cultivation were omitted. These modifications were made 

 to improve the quality of the data and to provide data more 

 useful to the user. 



The "products for home consumption" section was 

 changed to acquire the amount of products consumed or 

 given away during the last week before the enumeration 

 day. In the 1978 census, the data were collected for 

 products consumed or given away during the previous 1 2 

 months before the enumeration day. This should be noted 

 in comparing these figures with that of previous censuses. 



Training Program 



Supervisors, crew leaders, and enumerators employed 

 for the census in Puerto Rico received special training for 

 the job in accordance with instructions prepared by the 

 U.S. Bureau of the Census. The training included practice 

 in interviewing, filling out the report form, and a detailed 

 discussion of the enumerator's instructions. 



Method of Enumeration 



The 1982 Census of Agriculture for Puerto Rico was 

 conducted largely by enumerators assigned to specific 

 geographic areas. In rural areas, they were instructed to 

 visit every household and/or dwelling (living) unit. In 

 urbanized areas, they were instructed to make inquiries 

 of knowledgeable people to find persons conducting 

 agricultural operations. To assist in obtaining as complete 

 an enumeration as possible, enumerators were provided 

 with detailed maps of their assigned areas. Also, they 

 were given a Listing Book for each enumeration district 

 (ED) or areas to be canvassed. The book was used to 



record names, addresses, and additional data for each 

 location visited, and whether or not a report form was 

 filled out. Also, it provided the enumerator with a series 

 of questions. By asking these questions, the enumerator 

 could determine whether a housing unit located on a place 

 with agricultural operations qualified as a farm according 

 to the census definition. 



When a place was located which qualified as a farm, 

 the enumerator noted it in the Listing Book and completed 

 a census report. When a place was located which required 

 a census report, but on which the operator did not live, 

 the enumerator entered the appropriate information for the 

 place on a special form. This information was transferred 

 to the enumerator assigned to the area where the operator 

 lived, who was then responsible for obtaining a complete 

 report for the operation. 



Because of an interruption in census planning operations 

 following the publishing of the 1 978 Census of Agriculture 

 for Puerto Rico, time constraints barred the conducting 

 of a farm identification survey in advance of the next cen- 

 sus as was done in 1 978. Instead, a list of large or special 

 farms was prepared from 1978 records. Farm operators 

 appearing on this list were mailed a report form in June 

 with a request to complete the form and hold it until the 

 enumerator arrived during enumeration. 



COMPUTER PROCESSING 



For the first time, a microcomputer system was estab- 

 lished and used in Puerto Rico in connection with a cen- 

 sus of agriculture. This system was located in the Bureau 

 of the Census San Juan Central Processing Office. The 

 microcomputer system allowed the data to be proc- 

 essed in a more timely and precise manner. 



The specific tasks accomplished with the microcom- 

 puter were: 



1 . Check-in of large farms — This system allowed for a 

 tally of all mailed cases in the census. It replaced the 

 time consuming manual check-in. 



2. Data keying — All forms were data keyed using this 

 system. This reduced the shipping cost and the 

 possibility of losing forms during shipping by omit- 

 ting the keying stage in the Census Bureau proc- 

 essing center in Jeffersonville, Ind. 



3. Consistency edit — An edit provided analysts with 

 printouts of the individual cases with potential 

 problems. These cases were examined and the prob- 

 lems solved either by correcting the data using other 

 data supplied on the report form, or by followup 

 activity of the field enumerators through either a 



1982 CENSUS OF AGRICULTURE 



APPENDIX A A-1 



