Land and Soil Experimental Research 2013 - Ethiopia

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

The Land and Soil Experimental Research 2013, was conducted as a joint collaboration with The World Bank (LSMS Team), the Central Statistical Agency of Ethiopia (CSA) and the World Agroforestry Center (ICRAF) in an effort to improve the quality of agricultural data, particularly with respect to land area and soil fertility measurements in Ethiopia. The aim of the LASER study was to assess the data quality associated with a number of possible measurement methodologies associated with land area, soil quality, and crop production while piloting the use of each method and assessing the feasibility of implementation in national household surveys. Accurate and timely crop production statistics are critical to adequate government policy responses and the availability of accurate measures are pivotal to establishing credible performance evaluation systems. However, agricultural statistics are often marred by controversy over methods and overall quality, leading to inertia at best, or entirely incorrect policy actions. Major advances in recent years in technologies and practices offer an opportunity to improve on some of the indicators commonly used to measure agricultural performance.

Considerable efforts were made in the 1960s and 1970s, primarily by the Food and Agriculture Organization (FAO), to build a body of knowledge on agricultural statistics based on sound research which, over the years, has proven invaluable to researchers and practitioners in the field of agriculture. However, little new knowledge has been generated over the past few decades and much of the available methodological outputs are now obsolete in view of the changing structure of the sector, driven by global and local trends in both the agronomics of farming and the environment. Measuring land area and soil quality was essential in properly estimating the factors that both promoted and hindered agricultural productivity. It is also critical to assess the accuracy of the key output variable, crop production, in order to validate the methodologies used to collect harvest data as well as analyse the impact of various input measurements on yield estimates. By measuring these components using a variety of methods it was possible to identify the implications of using each and move forward with the superior methods in future household surveys. LASER was implemented across three administrative zones of the Oromia region, namely: East Wellega, West Arsi, and Borena. In total, 1018 households were interviewed, with nearly 1800 agricultural fields selected for objective land area and soil fertility measurement.

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

Regional

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

Households

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

Sample survey data [ssd]

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

The objectives of the sample design for the Land and Soil Experimental Research study were multifaceted and included indicators related to soil properties, crop type, and socio-economic characteristics, among others. Because there were multiple indicators, calculating the sample size based on the variance of a single indicator was not the preferred approach. Instead, practical sampling allocation with implicit stratification was used. Three administrative zones of the Oromia region were selected based primarily on agroecology and geographic diversity. Secondary consideration was made for the availability of local soil research centers that were used for soil processing. The three selected zones were East Wellega, West Arsi and Borena. Using the Central Statistical Agency of Ethiopia (CSA) and the Agricultural Sample Survey (AgSS) as the sampling frame, a total of 85 Enumeration Areas (EAs) were selected.

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

Computer Assisted Personal Interview [capi]

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

Data collection for the study was completed via Computer-Assisted Personal-Interview (CAPI). Each enumerator and supervisor had a personal laptop computer equipped with the Census and Survey Processing System (CSPro), based CAPI application for the Post-Planting, Crop-Cutting, and Post-Harvest questionnaires. Each team was provided with a flash drive, to share data from enumerator to supervisor, and a wireless router, to share consolidated team data with the World Bank project manager. Supervisors were instructed to share data at the close of EA, and only after reviewing all completed questionnaires. Data review and cleaning took place via supervisor review, periodic error reports generated by the World Bank project manager, unplanned CSA supervisor household visits to cross-check responses, and ultimately data review and standard checks (possible value ranges, outliers, etc.).

摘要
---------------------------
2013年土地与土壤实验研究（Land and Soil Experimental Research 2013）是由世界银行（LSMS Team）、埃塞俄比亚中央统计局（Central Statistical Agency of Ethiopia, CSA）和世界农业林业中心（World Agroforestry Center, ICRAF）联合开展的一项研究，旨在提升埃塞俄比亚农业数据的质量，尤其是土地面积和土壤肥力测量方面的数据。LASER研究的目的是评估与土地面积、土壤质量和作物生产相关的一系列可能测量方法的适用性，同时试点使用每种方法并评估其在国家家庭调查中的实施可行性。精确及时的作物生产统计数据对于政府政策的有效响应以及准确措施的可用性至关重要。然而，由于方法上的争议和整体质量的不足，农业统计数据往往存在缺陷，从最好的情况来看是惰性，在最坏的情况下是政策行动的错误。近年来在技术和实践方面取得的重大进展为改进一些常用以衡量农业绩效的指标提供了机会。
在20世纪60年代和70年代，主要由联合国粮食及农业组织（Food and Agriculture Organization, FAO）发起，投入了大量努力，旨在基于可靠的研究建立一套农业统计知识体系，经过多年的实践证明，这套体系对农业领域的学者和实践者极为宝贵。然而，在过去的几十年中，新增的知识有限，且许多现有的方法论成果，鉴于该行业结构的变化，以及农业农艺学和环境方面的全球和地方趋势，已经过时。正确测量土地面积和土壤质量对于准确估计促进和阻碍农业生产力的因素至关重要。评估关键输出变量——作物生产——的准确性，以验证收集收获数据所使用的方法，以及分析各种输入测量对产量估计的影响，也是至关重要的。通过使用各种方法测量这些组成部分，可以识别每种方法的应用影响，并在未来的家庭调查中采用更优越的方法。LASER在奥罗米亚地区的三个行政区域——东维尔加、西阿西和博雷纳——实施。总共对1018个家庭进行了访谈，并选取了近1800个农田进行客观的土地面积和土壤肥力测量。
地理覆盖范围
---------------------------
区域

分析单元
---------------------------
家庭

数据类型
---------------------------
样本调查数据 [ssd]

抽样程序
---------------------------
土地与土壤实验研究样本设计的目标是多方面的，包括与土壤特性、作物类型和社会经济特征相关的指标等。由于存在多个指标，基于单个指标的方差来计算样本大小并不是首选的方法。相反，采用了具有隐含分层的事实抽样分配。主要基于农业生态和地理多样性，在奥罗米亚地区的三个行政区域——东维尔加、西阿西和博雷纳——进行了选择。次要考虑因素是当地土壤研究中心的可用性，这些中心用于土壤处理。所选的三个区域是东维尔加、西阿西和博雷纳。使用埃塞俄比亚中央统计局（CSA）和农业样本调查（AgSS）作为抽样框架，共选择了85个统计区域（EAs）。
数据收集方式
---------------------------
计算机辅助个人访谈 [capi]

数据清洗操作
---------------------------
本研究的数据收集是通过计算机辅助个人访谈（Computer-Assisted Personal-Interview, CAPI）完成的。每位调查员和监督员都配备了一台装有基于CSPro的个人笔记本电脑，该笔记本电脑运行着基于CAPI的应用程序，用于种植后、收割后和收获后的问卷。每个团队都提供了一张闪存驱动器，用于调查员与监督员之间共享数据，以及一个无线路由器，用于将汇总的团队数据与世界银行项目经理共享。监督员被指示在EA结束时共享数据，并且只有在审查了所有完成的问卷之后。数据审查和清洗是通过监督员审查、世界银行项目经理生成的定期错误报告、CSA监督员不计划的 household 访问以交叉核对回答，以及最终的数据审查和标准检查（可能的值范围、异常值等）来进行的。