Namibia Census of Agriculture 2013/2014 - Communal Sector - Namibia

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

The Namibia Census of Agriculture (NCA) 2013/14 covered the communal and commercial farming sectors throughout the country. The objective of the NCA 2013/14 is to obtain baseline agricultural production and structural variables for the communal and commercial farming sectors at the national and regional levels. The census will provide statistics to improve planning and decision-making in the agricultural sector and satisfy the information needs of the socio-economic database being set up by the Namibia Statistics Agency (NSA).

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

The geographical coverage for the NCA-CN 2013/2014 covers only the agricultural households and the population living in those households who are engaged in subsistence farming activities, thus the coverage for the communal sector is only the rural areas and the semi-urban areas.

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

The study was done on the population for the communal sector and a significant number of agricultural households in the rural communal areas of Namibia including the semi-urban areas around the urban centers. The census collected detailed data on crop production, crop storage, livestock production and fish farming. The unit of analysis is individuals, agricultural household, agricultural holdings, parcels and plots.

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

The NCA-CN 2013/2014 is designed to cover all the agricultural households members regardless of age who reside in the holding on the day of enumeration.

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

Sample survey data [ssd]

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

Sample size
A total sample size of 10,550 agricultural households was determined to give reasonably reliable estimates at the regional level for the most important variables. The proportional allocation of this sample did not yield the minimum sample size for some of the sparsely populated regions hence a power allocation with some adjustments had to be carried out as a compromise procedure while keeping the overall national sample fixed.

In general,10 agricultural households were sampled from each of the selected PSUs thus having a larger spread of the sample across the population of agricultural households. However, in some of the southern regions having less communal farming activities, the sample size per PSU was raised to 16 agricultural households.

Sample design
A stratified two stage cluster sample design for the communal sector was used. At the first stage, primary sampling units (PSUs) were selected with Probability Proportional to Size (PPS) from the sampling frame based on the Enumeration Areas of 2011 Population and Housing Census. The size measure of a PSU in the sampling frame was the number of agricultural households which was derived from the questions included in the 2011 Population and Housing Census as per the World Programme of Agriculture 2010 (WCA 2010) Handbook recommendations.

The main strata was the regions which are also the primary domains of estimation. The frame units (PSUs) were further stratified implicitly by the constituencies within the regions. The list of agricultural households prepared within a selected PSU formed the secondary sampling frame from which a sample of agricultural households was selected systematically.

A third stage of sampling was introduced to measure objectively the average yields of the three major crops namely; Maize, Sorghum and Millet for the purpose of estimating the production instead of the farmer's estimates. Hence a crop cutting experiment was conducted to measure the average yield of these crops. A list of plots under each of these crops in a sampled PSU was made using the plot information of the selected households within the PSU. These lists then formed the sampling frames for each of the crops in the PSU.

Three plots were then randomly selected from each of the crop lists. If the list contained less than 3 plots then all were included in the experiment. An area was marked within the selected plot according to the FAO guidelines. Matured crops inside this marked area were cut and the weights from wet as well as dry grains were recorded. Grains were weighted while wet as well as dry. These figures were then used to estimate the average yield of each of the crops.

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

Face-to-face [f2f]

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

Computer Assisted Personal Interview (CAPI) replaced the conventional paper questionnaire.

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

Data editing took place at a number of stages throughout the processing, including:
Office editing and coding, Data editing and data cleaning using Stata 13 and CSPro 5.0,
These phases were carried out over a period of 18 months. Out of this period, the designing of tabulation programs, and the generation, verification and correction of tables lasted for 10 months.

Response rate
---------------------------

Response rates were computed for each region and the overall response was 95.9 percent.

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

Sampling errors have been calculated for 15 selected indicators. The sampling errors are based on the variances of the estimated figures which in turn depend on the sample sizes and the variation observed in the population units.

Data appraisal
---------------------------

N/A

摘要
---------------------------
2013/14年纳米比亚农业普查（NCA）涵盖了全国范围内的公有和商业农业部门。NCA 2013/14的目标是获取全国及地区层面公有和商业农业部门的农业生产基础数据和结构变量。此次普查将为农业部门规划与决策提供统计数据，并满足纳米比亚统计局（NSA）建立的社会经济数据库的信息需求。
地理覆盖范围
---------------------------
NCA-CN 2013/2014的地理覆盖范围仅限于从事自给自足农业活动的农业家庭及其居住者，因此公有部门的覆盖范围仅包括农村和半城市地区。
分析单元
---------------------------
研究针对的是纳米比亚公有部门的人口以及包括半城市地区在内的农村公有区域的众多农业家庭。普查收集了关于作物生产、作物储存、畜牧业生产和渔业养殖的详细数据。分析单元包括个人、农业家庭、农业经营单位、地块和土地。
总体
---------------------------
NCA-CN 2013/2014旨在涵盖所有居住在经营单位且在调查日居住的、不论年龄的农业家庭成员。
数据类型
---------------------------
样本调查数据 [ssd]
抽样程序
---------------------------
样本量
为在地区层面合理可靠地估计最重要的变量，确定了10,550个农业家庭的总体样本量。由于按比例分配的样本量未能满足一些人口稀少的地区的最低样本量，因此不得不在保持总体国家样本量不变的情况下，采取了一些调整的权力分配作为折衷方案。
一般而言，从每个选定的PSU中抽取了10个农业家庭作为样本，从而在农业家庭人口中实现了样本的更大分布。然而，在一些南部地区，由于公有农业活动较少，每个PSU的样本量增加到16个农业家庭。
样本设计
对于公有部门，使用了分层两阶段集群样本设计。在第一阶段，根据2011年人口和住房普查的计数区域，从抽样框架中基于概率成比例大小（PPS）选择了初级抽样单位（PSU）。抽样框架中PSU的大小度量是农业家庭数量，该数量是根据2010年世界农业计划（WCA 2010）手册建议包含在2011年人口和住房普查中的问题推导出来的。
主要层是地区，也是估计的主要领域。框架单元（PSU）在地区内的选区内部隐式分层。在选定的PSU内准备的农业家庭清单形成了二级抽样框架，从中系统地选择了农业家庭的样本。
引入了第三阶段的抽样，以客观地测量三种主要作物（玉米、高粱和小米）的平均产量，用于估计产量而不是农民的估计。因此，进行了一次作物切割实验来测量这些作物的平均产量。在样本PSU中，使用选定的家庭在PSU内的地块信息制作了每种作物下的地块清单。然后，这些清单形成了PSU中每种作物的抽样框架。
然后从每种作物的清单中随机选择了三个地块。如果清单中包含少于3个地块，则将所有地块都包括在实验中。根据FAO指南，在选定的地块内标记了一个区域。在这个标记区域内的成熟作物被割下，湿重和干重谷物都被记录下来。在湿和干状态下称重谷物，然后使用这些数据来估计每种作物的平均产量。
数据收集方式
---------------------------
面对面 [f2f]
研究工具
---------------------------
计算机辅助个人访谈（CAPI）取代了传统的纸质问卷。
清理操作
---------------------------
数据编辑在整个处理过程中发生，包括：
办公室编辑和编码，使用Stata 13和CSPro 5.0进行数据编辑和数据清理。
这些阶段在18个月的时间内完成。在这段时间里，表格程序的设计、表格的生成、验证和修正持续了10个月。
响应率
---------------------------
为每个地区和整体计算了响应率，整体响应率为95.9%。
抽样误差估计
---------------------------
计算了15个选定指标的调查误差。抽样误差基于估计数字的方差，而方差又取决于样本量和人口单位观察到的变异。
数据评估
---------------------------
N/A