Food Insecurity Experience Scale 2014 - Bangladesh

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

Sustainable Development Goal (SDG) target 2.1 commits countries to end hunger, ensure access by all people to safe, nutritious and sufficient food all year around. Indicator 2.1.2, “Prevalence of moderate or severe food insecurity based on the Food Insecurity Experience Scale (FIES)”, provides internationally-comparable estimates of the proportion of the population facing difficulties in accessing food. More detailed background information is available at http://www.fao.org/in-action/voices-of-the-hungry/fies/en/ .

The FIES-based indicators are compiled using the FIES survey module, containing 8 questions. Two indicators can be computed: 1. The proportion of the population experiencing moderate or severe food insecurity (SDG indicator 2.1.2), 2. The proportion of the population experiencing severe food insecurity. These data were collected by FAO through the Gallup World Poll. General information on the methodology can be found here: https://www.gallup.com/178667/gallup-world-poll-work.aspx. National institutions can also collect FIES data by including the FIES survey module in nationally representative surveys.

Microdata can be used to calculate the indicator 2.1.2 at national level. Instructions for computing this indicator are described in the methodological document available under the "DOCUMENTATION" tab above. Disaggregating results at sub-national level is not encouraged because estimates will suffer from substantial sampling and measurement error.

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

National

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

Individuals

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

Individuals of 15 years or older.

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

Sample survey data [ssd]

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

The population was stratified by geographic regions and each region was further stratified by 3 size categories (division headquarters, municipal centers and rural). In the sampling design Wards (in Metro and MCs) and Thana (rural) were selected using the probability proportional to size (PPS) method.
Exclusions: None
Design effect: 1.28

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

Face-to-face par [f2f]

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

The questionnaire is provided as an external resource in the Documentation Section.

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

Statistical validation assesses the quality of the FIES data collected by testing their consistency with the assumptions of the Rasch model. This analysis involves the interpretation of several statistics that reveal 1) items that do not perform well in a given context, 2) cases with highly erratic response patterns, 3) pairs of items that may be redundant, and 4) the proportion of total variance in the population that is accounted for by the measurement model.

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

The margin of error is estimated as 3.5 .This is calculated around a proportion at the 95% confidence level. The maximum margin of error was calculated assuming a reported percentage of 50% and takes into account the design effect.

摘要
---------------------------
可持续发展目标（SDG）2.1 承诺各国结束饥饿，确保全年人人都能获得安全、营养和充足的食物。指标 2.1.2，‘基于食物不安全经验量表（FIES）的食物不安全普遍程度’，提供了国际上可比较的估计，表明面临食物获取困难的人口比例。更多详细背景信息请参阅 http://www.fao.org/in-action/voices-of-the-hungry/fies/en/ 。

基于 FIES 的指标是通过 FIES 调查模块编制的，包含 8 个问题。可以计算两个指标：1. 经历中度或严重食物不安全的人口比例（SDG 指标 2.1.2），2. 经历严重食物不安全的人口比例。这些数据由联合国粮农组织通过盖洛普世界民意调查收集。关于方法论的详细信息请参阅：https://www.gallup.com/178667/gallup-world-poll-work.aspx。国家机构也可以通过在国家代表性调查中包含 FIES 调查模块来收集 FIES 数据。

微观数据可用于计算国家层面的指标 2.1.2。计算此指标的说明可在上述“文档”标签页下的方法论文件中找到。不建议在次国家层面分解结果，因为估计将受到大量抽样和测量误差的影响。

地理覆盖范围
---------------------------
全国

分析单位
---------------------------
个人

调查范围
---------------------------
15 岁及以上个人。

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

抽样程序
---------------------------
人口按地理区域分层，每个区域再按 3 个规模类别（分局总部、市中心和农村）进一步分层。在抽样设计中，使用按规模比例抽样（PPS）方法选择街区（在都会和市中心）和塔纳（农村）。

排除：无
设计效应：1.28

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

研究工具
---------------------------
问卷作为外部资源提供，可在“文档”部分找到。

清理操作
---------------------------
统计验证通过测试 FIES 数据与 Rasch 模型假设的一致性来评估其质量。此分析涉及解释几个统计量，揭示了以下内容：1) 在特定情境中表现不佳的项目，2) 具有高度波动响应模式的案例，3) 可能冗余的项目对，以及 4) 由测量模型解释的总方差比例在人口中的占比。

抽样误差估计
---------------------------
误差范围估计为 3.5。这是在 95% 置信水平下围绕比例计算的。最大误差范围是在假设报告百分比为 50% 的前提下计算的，并考虑了设计效应。