Census of Agriculture, 2007 - Vanuatu

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

The Agriculture Census is envisioned with the following objectives:

- To provide data on the structure of agriculture as well as forestry and fisheries in Vanuatu;
- To provide data that will be used as benchmark for current agricultural statistics; and
- To provide sampling frame for surveys on agriculture (crops and livestock), fisheries and forestry.

Specifically, the Agriculture Census Phase II aims:

- To determine the structure and characteristics of the agricultural activities of the households in Vanuatu such as crop gardening, coconut/cocoa/ coffee/kava/vanilla/pepper farming, tending of cattle and other livestock activities, forestry-related activities and fishing operations;
- To determine the number and distribution of household engaged in crop gardening, coconut/cocoa/coffee/kava/vanilla/pepper farming, tending of cattle and other livestock activities, forestry-related activities
and fishing operations at the island level; and
- To provide data on the farm/holding/sub-holding area, quantity of the crops grown/sold, number of cattle and other livestock kept as of the day of enumeration, quantity of fisheries species gathered/caught, etc

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

Regional coverage

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

Households

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

The statistical unit was the agricultural holding (farm), defined as any economic unit of agricultural production (such as a garden or temporary and/or permanent crops or cattle raising, cocoa/coffee/kava/vanilla/pepper area/plantation) under single management, comprising all livestock kept and all land wholly or partly used for agricultural production purposes, without regard to title, legal form or size. Information on agricultural holdings was collected according to their legal status.

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

Census/enumeration data [cen]

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

Sampling method:

The 18 major islands were classified as:
• small – number of households engaged in agricultural activities less than 500 (Torres, Paama, Erromango, Aniwa, Aneityum and Futuna);
• medium – number of households engaged in agricultural activities 500-1,999 (Banks, Malo, Maewo, Ambrym, Epi and Shepherds); and
• large – number of households operating agricultural activities 2,000 or more (Efate, Malekula, Ambae, Pentecost and Tanna).

In determining the number of households to be interviewed in each island and in each enumeration area (EA):

- for small islands, all households were listed and the identified households engaged in agricultural activities were enumerated;
- for medium-sized islands, one-third of the sample EAs in these islands were selected and all households were listed and those found to be engaged in agricultural activities were interviewed; and
- for large islands, one-third of the total EAs were selected in each island and all households listed. Of households found to have a crop garden, coconut sub-holding or kava sub-holding, one-third were selected to be further interviewed. In addition, all households listed and involved in the subholding of cattle and cash crops like cocoa, coffee (for Tanna only), vanilla and pepper (10 or more plants) were also enumerated.

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

Face-to-face [f2f]

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

Two types of questionnaires were used: one for listing the households (Phase 1) and one for data collection (Phase 2). Separate questionnaires were provided for different agricultural activities (nine questionnaires in total). Only one of the 16 core items recommended by FAO for the WCA 2010 was not covered by the CA 2007.

Phase I: Census Listing
Phase II: Surveys
Form 1.1 - Household
Form 1.2 - Crop Garden
Form 1.2A - Gardener's Form
Form 1.3 - Kava
Form 1.4 - Coconut
Form 2 - Cocoa
Form 3 - Coffee
Form 4 - Vanilla
Form 5 - Pepper
Form 6 - Cattle
Form 7 - Commercial Farm
Form A - List of Activities
Form B1 - Control Sheet for all small and medium sized islands
Form B2 - Control Sheet for Santo, Pentecost and Ambae
Form B3 - Control Sheet for Ambrym and Malekula
Form B4 - Control Sheet for Efate and Tanna

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

Data editing:

Eight data entry operators were hired by the project to do the data encoding of the Phase I of the project. This was the first-hands on as far as the software is concerned for all the data entry operators. Before the actual data entry, the data processing expert had all eight operators plus the supervisors on a training session for a few days. At the end of the training session, they were familiar with the software and then started the actual data encoding. The processing of data for Phase I of the project took the entire month of June 2006 to be completed. During the Phase II of the project, the expert set up the system and trained the local staff on system operation for two weeks and then left for his home country. Since the project staff and the data entry operators who were hired were already familiar with CsPro, the whole data processing was done without the presence of the consultant. The expert later came for his final mission to prepare the data for tabulation and generate the required tables using the table specifications for that purpose. The machine data processing of the forms was done using CsPro. Data encoding, data cleaning and tabulation were done using data entry, batch edit and cross tab applications respectively. Control and management of the data entry of the forms and data cleaning of the batch files were done using SCIPS (Survey / Census Integrated Processing System), a Visual Basic 6 (VB6) program developed by the expert designed to integrate the different phases of data capture and data cleaning of any survey/census. The program facilitates the assignment of folios to keyers that resulted to automatic recording of the data capture status of each batch/folio and eliminated errors in the encoding of the geographic identification codes. It also made the data cleaning easier since SCIPS enabled the users to correct errors found by the data consistency and completeness check programs without printing the generated error list.

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

The number of households to be interviewed is based on the sampling methodology that is used in the census. The 15 major islands were classified as:

1. small - if the number of households engaged in agricultural activities is less than 500; in this case, Torres, Paama and Erromango are under this category.
2. medium - if the number of households engaged in agricultural activities is between 500 - 1,999; Banks, Malo, Maewo, Ambrym, Epi and Shepherds belong to this group.
3. large - if the number of households operating agricultural activities is 2,000 or more; Santo, Efate, Malekula, Ambae, Pentecost and Tanna were considered to be large islands.

In selecting the number of households to be interviewed in each island, the following was carried out:

a. For Erromango, Torres and Paama, all households were listed and those households engaged in agricultural activities were enumerated;
b. For Banks, Malo, Maewo, Ambrym, Epi and Shepherds, 1/3 of the sample EAs in these islands were selected and all households were listed and those engaged in agricultural activities were interviewed for their involvement in these activities; and
c. For Santo, Efate, Malekula, Ambae, Pentecost and Tanna, 1/3 of the total EAs were also selected in each island and all households were listed in these islands, after which only 1/3 of the households engaged in agricultural activities were further interviewed if they were involved in crop garden, coconut sub-holding and kava sub-holding. In addition to this, all households in the selected EAs of these islands that were involved in the sub-holding of cattle and cash crops (with 10 trees or more) like cocoa, coffee (for Tanna only), vanilla and pepper were enumerated.

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

The CA 2007 data were disseminated through a printed publication, which included the final census results. The results were also published on institution's website.

摘要
---------------------------
农业普查旨在实现以下目标：
- 提供关于瓦努阿图农业结构、林业和渔业的资料；
- 提供可作为当前农业统计数据基准的资料；
- 为农业（作物和畜牧业）、渔业和林业调查提供抽样框架。
具体而言，农业普查第二阶段旨在：
- 确定瓦努阿图家庭从事的农业活动结构及特征，如作物园艺、椰子/可可/咖啡/胡椒/胡椒树/辣椒种植、牲畜饲养及其他畜牧业活动、与林业相关的活动及捕鱼作业；
- 确定在岛屿层面上从事作物园艺、椰子/可可/咖啡/胡椒/胡椒树/辣椒种植、牲畜饲养及其他畜牧业活动、与林业相关的活动及捕鱼作业的家庭数量及分布；
- 提供关于农场/经营单位/次级经营单位面积、种植/销售作物数量、截至统计日饲养的牲畜数量、收集/捕获的渔业物种数量等数据。
地理覆盖范围
---------------------------
区域覆盖

分析单元
---------------------------
家庭

总体
---------------------------
统计单元为农业经营单位（农场），定义为任何农业生产经济单位（如花园或临时和/或永久作物或牲畜饲养、可可/咖啡/胡椒树/胡椒树/辣椒种植园或种植园），在单一管理下，包括所有饲养的牲畜和所有完全或部分用于农业生产目的的土地，不考虑所有权、法律形式或规模。根据其法律状态收集农业经营单位信息。

数据类型
---------------------------
普查/枚举数据 [cen]

抽样程序
---------------------------
抽样方法：
将18个主要岛屿分类为：
• 小型岛屿 – 从事农业活动的家庭数量少于500（托雷斯、帕阿玛、埃罗芒戈、阿尼瓦、阿内蒂乌姆和富图纳）；
• 中型岛屿 – 从事农业活动的家庭数量为500-1,999（班克斯、马洛、马埃沃、埃皮、埃皮和牧羊人）；
• 大型岛屿 – 从事农业活动的家庭数量为2,000或更多（埃法特、马勒库拉、安巴埃、圣灵岛和塔纳岛）。
在确定每个岛屿和每个统计区域（EA）中要访谈的家庭数量时：
- 对于小型岛屿，列出所有家庭，并对从事农业活动的识别家庭进行统计；
- 对于中型岛屿，选择这些岛屿中三分之一的样本EA，列出所有家庭，并对从事农业活动的家庭进行访谈；
- 对于大型岛屿，在每个岛屿中选择三分之一的总EA，列出所有家庭。在发现有作物花园、椰子次级经营单位或胡椒树次级经营单位的家庭中，选择三分之一进行进一步访谈。此外，所有列出的家庭，如果涉及牲畜饲养和现金作物（如可可、咖啡（仅塔纳岛）、香草和胡椒（10棵或更多植物）的次级经营单位，也进行了统计。

数据收集方式
---------------------------
面对面 [f2f]

研究工具
---------------------------
使用了两种类型的问卷：一种用于列出家庭（第一阶段），另一种用于数据收集（第二阶段）。为不同的农业活动提供了单独的问卷（总共九份问卷）。只有16项核心项目中有1项未被FAO推荐的WCA 2010所涵盖。
第一阶段：普查登记
第二阶段：调查
表格1.1 - 家庭
表格1.2 - 作物花园
表格1.2A - 园艺师表格
表格1.3 - 胡椒
表格1.4 - 椰子
表格2 - 可可
表格3 - 咖啡
表格4 - 香草
表格5 - 辣椒
表格6 - 牲畜
表格7 - 商业农场
表格A - 活动清单
表格B1 - 所有小型和中型岛屿的控制表
表格B2 - 圣托、圣灵岛和安巴埃的控制表
表格B3 - 安巴埃和马勒库拉的控制表
表格B4 - 埃法特和塔纳的控制表

数据清理操作
---------------------------
数据编辑：
项目雇佣了8名数据录入员进行项目第一阶段的数据编码。就软件而言，这是所有数据录入员的第一手操作。在实际数据录入之前，数据处理专家对包括8名录入员和监督员在内的所有人进行了为期几天的培训。培训结束后，他们熟悉了软件，然后开始了实际的数据编码。项目第一阶段的数据处理耗时整整一个月的2006年6月。在项目的第二阶段，专家建立了系统，并在两周内对当地员工进行了系统操作培训，然后返回了祖国。由于项目员工和数据录入员已经熟悉CsPro，整个数据处理过程无需咨询人员在场。专家后来再次前来，完成了数据的整理，并使用相应的表格规范生成了所需的表格。表格的机器数据处理使用CsPro完成。数据编码、数据清理和制表分别使用数据录入、批量编辑和交叉制表应用程序完成。表格和批量文件的数据录入控制及数据清理使用SCIPS（调查/普查集成处理系统）进行管理，SCIPS是由专家开发的Visual Basic 6（VB6）程序，旨在集成任何调查/普查的不同阶段的数据捕获和数据清理。该程序简化了用户对数据一致性检查和完整性检查程序发现错误的纠正，无需打印生成的错误列表。它还使得数据清理变得更加容易，因为SCIPS允许用户在不打印错误列表的情况下纠正数据一致性检查和完整性检查程序发现的错误。

抽样误差估计
---------------------------
要访谈的家庭数量基于普查中使用的抽样方法。15个主要岛屿被分类为：
1. 小型岛屿 – 从事农业活动的家庭数量少于500；在这种情况下，托雷斯、帕阿玛和埃罗芒戈属于这一类别。
2. 中型岛屿 – 从事农业活动的家庭数量在500-1,999之间；班克斯、马洛、马埃沃、埃皮、埃皮和牧羊人属于这一组。
3. 大型岛屿 – 从事农业活动的家庭数量为2,000或更多；圣托、埃法特、马勒库拉、安巴埃、圣灵岛和塔纳岛被认为是大型岛屿。
在确定每个岛屿中要访谈的家庭数量时，执行以下操作：
a. 对于埃罗芒戈、托雷斯和帕阿玛，列出所有家庭，并对从事农业活动的家庭进行统计；
b. 对于班克斯、马洛、马埃沃、埃皮、埃皮和牧羊人，选择这些岛屿中三分之一的样本EA，列出所有家庭，并对从事农业活动的家庭进行访谈，以了解他们在这些活动中的参与情况；
c. 对于圣托、埃法特、马勒库拉、安巴埃、圣灵岛和塔纳岛，在每个岛屿中也选择三分之一的总EA，列出所有岛屿上的家庭，然后如果他们在作物花园、椰子次级经营单位和胡椒树次级经营单位中参与农业活动，则进一步访谈三分之一的家庭。此外，还统计了这些岛屿上选定的EA中涉及牲畜饲养和现金作物（如可可、咖啡（仅塔纳岛）、香草和胡椒（10棵或更多植物）次级经营单位的所有家庭。

数据评估
---------------------------
2007年CA数据通过印刷出版物进行传播，其中包含最终普查结果。结果还发布在机构的网站上。