Agricultural Sample Enumeration, Land Use 2001-2002 (1994 E.C) - Ethiopia

Abstract
---------------------------

From agricultural point of view, land is an indispensable factor for production of crops, raising of livestock and other ancillary agricultural activities. The proper utilization of land holdings under different components will contribute to the development of the nation's agricultural products. In order to scrutinize this development as well as farmers' attitude towards land use practices, a timely and comprehensive land use data has paramount importance for governmental, non-governmental and private sector data users. According to the international recommendations for Agricultural Census, the total land use is categorized into five main land use types, which are land under temporary crops, land under permanent crops, grazing land, fallow land, woodland and land for other purposes. Based on these major land use categories, the 2001/02 (1994 E.C.) Ethiopian Agricultural Sample Enumeration (EASE) provides quantitative information on land utilization at Regional, Zonal and Wereda levels.

Geographic coverage
---------------------------

The 2001-2002 (1994 E.C) Agricultural Sample Enumeration was designed to cover the rural and urban parts of all districts (weredas) in the country on a large-scale sample basis excluding the pastoralist areas of the Afar and Somali regional states.

Analysis unit
---------------------------

Household, holder, type of land use

Universe
---------------------------

The 2001/02 (1994 E.C.) EASE covered all land holdings i.e., rural and urban holdings. In urban areas, size of land holdings was restricted to only cropland area and urban agricultural households were required to have a minimum size of 250 square meters in order to be covered in the census. On the other hand, in the rural areas land use data were collected without any size limitation of land holdings.

The coverage of land use data items in rural private holdings included all the categories of land used. However, in urban private holdings, the coverage of land use data item is limited to cropland area, which includes land under temporary and permanent crops.

Kind of data
---------------------------

Census/enumeration data [cen]

Sampling procedure
---------------------------

Sampling Frame
The list of enumeration areas for each wereda was compiled from the 1994 Ethiopian Population and Housing Census cartographic work and was used a frame for the selection of the Primary Sampling Units (PSU). The 1994 Population and Housing Census enumeration area maps of the region for the selected sample EA's were updated, and the EA boundaries and descriptions were further clarified to reflect the current physical situation. The sampling frame used for the selection of ultimate sampling units (agricultural households) was a fresh list of households, which was prepared by the enumerator assigned in the sampled EA's using a prescribed listing instruction at the beginning of the launching of the census enumeration.

Sample Design
In order to meet the objectives and requirements of the EASE, a stratified two-stage cluster sample design was used for the selection of ultimate sampling units. Thus, in the regions each wereda was treated as stratum for which major findings of the sample census are reported. The primary sampling units are the enumeration areas and the agricultural households are secondary (ultimate) sampling units. Finally, after the selection of the sample agricultural households, the various census forms were administered to all agricultural holders within the sampled agricultural households.

For the private peasant holdings in the rural areas a fixed number (25) of sample EA's in each wereda and 30 agricultural households in each EA were randomly selected (determined). In urban areas, weredas with urban EA's of less than or equal to 25, all the EA's were covered. However, for weredas with greater than 25 urban EA's, sample size of 25 EA's was selected. In each sampled urban EA, 30 agricultural households were randomly selected for the census. The sampled size determination in each wereda and thereby in each EA was based upon the required precision level of the major estimates and the cost consideration. The pilot survey and the previous year annual agricultural sample survey results were used to determine the required sample sizes per wereda.

Sample Selection of Primary Sampling Units
Within each wereda (stratum) in the region, the selection of EAs was carried out using probability proportional to size systematic sampling. In this case, size being total number of agricultural households in each EA obtained from the listing exercise undertaken in the 1994 Ethiopian Population and Housing Census of the region.

Listing of Households and Selection of Agricultural Households
In each sampled enumeration area of the region, a complete and fresh listing of households was carried out by canvassing the households in the EA. After a complete listing of the households and screening of the agricultural households during the listing operation in the selected EA, the agricultural households were serially numbered. From this list, a total of 30 agricultural households were selected systematically using a random start from the pre-assigned column table of random numbers. The sampling interval for each EA was determined by dividing the total number of agricultural households by 30. For crop cutting exercise purposes (rural domain) a total of 20 agricultural households were randomly selected from the 30 sampled agricultural households. The systematical random sampling technique was employed in this case, because its application is simple and flexible, and it can easily yield a proportionate sample.

Mode of data collection
---------------------------

Face-to-face [f2f]

Research instrument
---------------------------

Although the planning, organization and execution of the census were the responsibilities that rested within the CSA, development of the census forms was a tedious task that involved the formation of a working group composed of members of government and non-governmental organizations who are major users of agricultural data. Members of the working group were given the opportunity to identify their data requirements, define the needs of others and determine the specific questions that the forms should contain. The working group included the staff of the organizations that are involved in agricultural planning, collection of agricultural statistics and the use of data within the agricultural sector. The working group designed different forms for the various data items on crop area, production, and other variables of interest to meet the needs of current data users and also considered the future expectations. Attempt was made to make the content of the forms of acceptable length by distributing the variables to be collected in the different census forms.

The rural census questionnaires/forms included:
- Forms 94/0 and 94/1 that are used to record all households in the enumeration area, identify the agricultural households and select the units to be covered by the census.
- Form 94/2 is developed to list all the members of the sampled agricultural households and record the demographic and economic characteristics of each of the members.
- Forms 94/3A, 94/3B, 94/3C and 94/3D are prepared to enumerate crop data through interview and objective measurement.
- Form 94/5 is designed to record crop area data via the physical or objective measurement of crop fields.
- Form 94/6 is used to list all the fields under crop and select a crop field for each type of crop randomly for crop cutting exercise.
- Forms 94/7A, 94/7B, and 94/7C are developed for recording yield data on cereals, oil seeds, pulses, vegetables root crops and permanent crops by weighing their yields obtained from sub-plots and/or trees selected for crop-cuttings.
- Form 94/8 is prepared to enumerate livestock, poultry and beehives data by type, age, sex and purpose including products through interview (subjective approach).
- Forms 94/9, 94/10 and 94/11 are used to collect data on crop and livestock product usage; miscellaneous items and farm tools, implements, draught animals and storage facilities, in that order, by interviewing the sample holders.

"Belg" season questionnaires identified as:
- Form 94/12A and 94/12B that are used to record data on farm management practices of the "Belg" season.
- Form 94/4 was the questionnaire used for collecting data on crop production forecast for 2001-2002 and the data collected using this form was published in December 2001 subjectively, while 94/12C is for recording "Belg" season crop area through objective measurement and volume of production through interview approach.

On the other hand, the census questionnaires/forms used in the urban areas include:
- Form U-94/1 which used to record all households in the EA, identify the agricultural households and select the units to be covered by the census.
- Form U-94/2 is developed to list all the members of the sampled agricultural household and record the demographic and economic characteristics of each of the members.
- From U-94/3 is prepared to enumerate crop data through interview method.
- Form U-94/4 is prepared to enumerate livestock, poultry and beehives data by type, sex, age and purpose including products through interview (subjective approach).
- Form U-94/5 is used to collect data on crop and livestock usage.

Cleaning operations
---------------------------

Editing, Coding and Verification:
In the 2001-2002 Ethiopian Agricultural Sample Enumeration (EASE), the filled-in forms that were retrieved from 47 Branch Statistical Offices were primarily received and systematically registered at the documentation unit of the CSA head quarters in Addis Ababa. Before launching the actual editing and coding activities, the Natural Resources and Agricultural Statistics Department staff gave adequate training to the 157 editors and coders. These editors and coders carried out the manual editing, coding and verification of the filled-in EASE questionnaires in two shifts. At the outset, the editing and coding activities for the filled-in forms on area and agricultural practices took place; this was followed up by the editing and coding of the forms on the production of temporary crops (cereals, pulses, oil seeds, vegetables and root crops), livestock, farm implements, permanent crops, "Belg" and miscellaneous questionnaires region by region. For the filled-in forms on area and agricultural practices, verification was carried out on 100% basis for the first five weeks from the launching of the activity and then considering the quality performance of editor-coders the activity was dropped to 66% of the forms gradually. On the other hand, the verification activity has been carried out on 100% basis for the filled-in forms on production of the temporary and permanent crops, livestock, farm implements and all other completed forms. For the total country, the editing, coding and verification of the filled-in forms in general took about 330.6 working days. That is, the editing, coding and verification of the filled-in forms for area, agricultural practice, the production of the temporary and permanent crops, and livestock took about 198.5 working days, while that of the filled-in forms on farm implements, demographic characteristics, Belg season and the urban forms took around 132.1 working days.

Data Entry, Cleaning and Tabulation:
About 144 data encoders were assigned to undertake the data entry activity of 2001/02 EASE and it has been carried out on two-shift basis. Before the starting of the data entry operation data encoders were trained for about 5 days using computer programs developed by the Data Processing Department staff. The Programmers prepared the data entry programs using CENTRY, which is a data entry module of IMPS (Integrated Microcomputer Processing System). The data entry exercise has been carried out using 76 personal computers (PC's). Like that of the manual editing and coding activity, the filled-in forms on area and agricultural practices were entered first and this was followed by entry of the filled-in forms on the production of temporary crops, livestock, farm implements, permanent crops, "Belg" and miscellaneous questionnaires region by region till all the census data entry operations are completed. In order to ensure the quality of the data entry work, verification exercise was carried out. The entry of the filled-in forms on area and agricultural practices were verified on 100 % basis. Then the verification exercise was dropped to 66 % from the 6th week of the launching of the operation and was further reduced to 50% from the 10th week onwards by observing and assessing the magnitude of the percentage of errors. Later on verification process was carried out on 100% basis for the filled-in forms on the production of temporary and permanent crops, livestock, farm implements and all other completed forms. The verification activity was carried out through the process of re-entering the data. For the total country, the whole data entry process of the filled-in forms on area, agricultural practice, the production of the temporary and permanent crops, and livestock took around 253.1 working days, while that of the filled-in forms on farm implements, demographic characteristics, Belg season and the urban forms took about 257.9 working days. Data entered into the computer needs to be checked for completeness, consistency and validity. For this purpose computer edit programs were prepared by programmers using CONCOR, which is the editing module of IMPS. Using print-outs from these programs and referring to the filled-in census forms, corrections were made by nine trained manual data cleaning technicians. Moreover, nine other data-cleaning computer operators were involved in making the actual corrections of the data on the computer. Additionally, an intermediate set of instructions or programs were made available and applied on the data to prepare information suitable for tabulation. These programs were prepared using CSPro and IMPS software. Like IMPS Software, CSPro is used as a tool for entering, editing and tabulating data. CSA used the CSPro software for data editing and calculation of CVs. Data made ready for tabulation through the process of cleaning and intermediate programs was finally used to generate the required tables. This was done using tabulation programs developed by the senior programmers of the Data processing Department. The CENTS software, a tabulation component of IMPS, was used in producing the 2001-2002 EASE results.

Sampling error estimates
---------------------------

Estimated procedure of parameters of interest like total, yield and ratio and their sampling errors is presented in Appendix I of the reports which are attached with this metadata. Standard errors and coefficients of variations of estimates for selected variables are also given as an annex at the end of each report.

摘要
---------------------------

从农业的角度来看，土地是农作物生产、畜牧业饲养及其他辅助农业活动不可或缺的生产要素。在不同组成部分下合理利用土地，将有助于国家农业产品的开发。为了审视这一发展以及农民对土地利用实践的看法，及时且全面的土地利用数据对于政府、非政府及私营部门的数据使用者来说至关重要。根据国际农业普查的建议，总土地利用被分为五种主要土地利用类型，包括临时作物用地、永久作物用地、牧场、休耕地、林地及其他用途用地。基于这些主要土地利用类别，2001/02年（1994 E.C.）的埃塞俄比亚农业抽样调查（EASE）提供了区域、州和Wereda级别的土地利用定量信息。

地理覆盖范围
---------------------------

2001-2002年（1994 E.C.）农业抽样调查旨在以大规模抽样为基础，涵盖该国所有区（weredas）的农村和城市部分，不包括阿法尔和索马里州游牧地区。

分析单位
---------------------------

家庭、持有者、土地利用类型

总体
---------------------------

2001/02年（1994 E.C.）EASE涵盖了所有土地持有，即农村和城市土地持有。在城市地区，土地持有规模仅限于耕地面积，城市农业家庭必须拥有至少250平方米的面积，才能被纳入普查。另一方面，在农村地区，土地使用数据收集时对土地持有规模没有限制。

农村私人持有土地的土地使用数据覆盖了所有土地使用类别。然而，在城市私人持有土地中，土地使用数据项的覆盖范围仅限于耕地面积，包括临时和永久作物用地。

数据类型
---------------------------

普查/枚举数据 [cen]

抽样程序
---------------------------

抽样框架
...
（以下内容省略，因为翻译长度限制，完整内容请参照原始文档）...

数据收集方式
---------------------------

面对面 [f2f]

研究工具
---------------------------

尽管普查的计划、组织和执行责任落在CSA身上，但普查表格的开发是一项繁琐的任务，涉及由政府和非政府组织成员组成的工作组，他们是农业数据的主要使用者。工作组成员有机会确定他们的数据需求，定义他人的需求，并确定表格应包含的具体问题。工作组包括参与农业规划、农业统计收集以及农业部门数据使用的组织的工作人员。工作组为作物面积、产量以及其他对当前数据使用者感兴趣的可变因素设计了不同的表格，以满足当前数据使用者的需求，并考虑了未来的期望。通过将收集的变量分配到不同的普查表格中，试图使表格的内容达到可接受的长度。

农村普查问卷/表格包括：
...
（以下内容省略，因为翻译长度限制，完整内容请参照原始文档）...

城市地区使用的普查问卷/表格包括：
...
（以下内容省略，因为翻译长度限制，完整内容请参照原始文档）...

数据清理操作
---------------------------

编辑、编码和验证：
...
（以下内容省略，因为翻译长度限制，完整内容请参照原始文档）...

数据录入、清理和制表：
...
（以下内容省略，因为翻译长度限制，完整内容请参照原始文档）...

抽样误差估计
---------------------------

感兴趣参数（如总数、产量和比率）及其抽样误差的估计过程在报告的附录I中给出，该报告随本元数据附上。选定变量的估计的标准误差和变异系数也作为附件，位于每个报告的末尾。