Integrated Living Conditions Survey 2007 - Armenia

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

The Integrated Living Conditions Survey (ILCS), conducted annually by the NSS National Statistical Service of the Republic of Armenia, formed the basis for monitoring living conditions in Armenia. The ILCS is a universally recognized best-practice survey for collecting data to inform about the living standards of households. The ILCS comprises comprehensive and valuable data on the welfare of households and separate individuals which gives the NSS an opportunity to provide the public with up to date information on the population’s income, expenditures, the level of poverty and the other changes in living standards on an annual basis.

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

Urban and rural communities

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

- Households;
- Individuals.

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

Sample survey data [ssd]

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

During the 2001-2003 surveys two-stage random sample was used; the first stage covered the selection of settlements - cities and villages, while the second stage was focused on the selection of households in these settlements. The surveys were conducted on the principle of monthly rotation of households by clusters (sample units). In 2002 and 2003 the number of households was 387 with the sample covering 14 cities and 30 villages in 2002 and 17 cities and 20 villages in 2003.

During the 2004-2006 surveys the sampling frame for the ILCS was built using the database of addresses for the 2001 Population Census; the database was developed with the World Bank technical assistance. The database of addresses of all households in Armenia was divided into 48 strata including 12 communities of Yerevan city. The households from other regions (marzes) were grouped according to the following three categories: big towns with 15,000 and more population; villages, and other towns. Big towns formed 16 strata (the only exception was the Vayots Dzor marz where there are no big towns). The villages and other towns formed 10 strata each. According to this division, a random, two-step sample stratified at marz level was developed. All marzes, as well as all urban and rural settlements were included in the sample population according to the share of population residing in those settlements as percent to the total population in the country.
In the first step, the settlements, i.e. primary sample units, were selected: 43 towns out of 48 or 90 percent of all towns in Armenia were surveyed during the year; also 216 villages out of 951 or 23 percent of all villages in the country were covered by the survey. In the second step, the respondent households were selected: 6,816 households (5,088 from urban and 1,728 from rural settlements). As a result, for the first time since 1996 survey data were representative at the marz level.

During the 2007-2012 surveys the sampling frame for ILCS was designed according to the database of addresses for the 2001 Population Census, which was developed with the World Bank technical assistance. The sample consisted of two parts: core sample and oversample.

1) For the creation of core sample, the sample frame (database of addresses of all households in Armenia) was divided into 48 strata including 12 communities of Yerevan city. The households from other regions (marzes) were grouped according to three categories: large towns (with population of 15000 and higher), villages and other towns. Large towns formed by 16 groups (strata), while the villages and towns formed by 10 strata each. According to that division, a random, two-step sample stratified at the marz level was developed. All marzes, as well as all urban and rural settlements were included in the sample population according to the share of households residing in those settlements as percent to the total households in the country. In the first step, using the PPS method the enumeration units (i.e., primary sample units to be surveyed during the year) were selected. 2007 sample includes 48 urban and 18 rural enumeration areas per month.
2) The oversample was drawn from the list of villages included in MCA-Armenia Rural Roads Rehabilitation Project. The enumeration areas of villages that were already in the core sample were excluded from that list. From the remaining enumeration areas 18 enumeration areas were selected per month. Thus, the rural sample size was doubled.
3) After merging the core sample and oversample, the survey households were selected in the second step. 656 households were surveyed per month, from which 368 from urban and 288 from rural settlements. Each month 82 interviewers had conducted field work, and their workload included 8 households per month. In 2007 number of surveyed households was 7,872 (4,416 from urban and 3,456 from rural areas).

For the survey 2013 the sample frame for ILCS was designed in accordance with the database of addresses of all private households in the country developed on basis of the 2001 Population Census results, with the technical assistance of the World Bank. The method of systematic representative probability sampling was used to frame the sample. For the purpose of drawing the sample, the sample frame was divided into 32 strata including 12 communities of Yerevan City (currently, the administrative districts).
According to this division, a two-tier sample was drawn stratified by regions and by Yerevan. All regions and Yerevan, as well as all urban and rural communities were included in the sample in accordance to the shares of their resident households within the total number of households in the country. In the first round, enumeration areas - that is primary sample units to be surveyed during the year - were selected. The ILCS 2013 sample included 32 enumeration areas in urban and 16 enumeration areas in rural communities per month.
The households to be surveyed were selected in the second round. A total of 432 households were surveyed per month, of which 279 and 153 households from urban and rural communities, respectively. Every month 48 interviewers went on field work with a workload of 9 households per month.

The sample frame for 2014-2016 was designed in accordance with the database of addresses of all private households in the country developed on basis of the 2011 Population Census results, with the technical assistance of the World Bank. The method of systematic representative probability sampling was used to frame the sample.
For drawing the sample, the sample frame was divided into 32 strata including 12 communities of Yerevan City (currently, the administrative districts). According to this division, a two-tier sample was drawn stratified by regions and by Yerevan. All regions and Yerevan, as well as all urban and rural communities were included in the sample in accordance to the shares of their resident households within the total number of households in the country. In the first round, enumeration areas - that is primary sample units to be surveyed during the year - were selected. The ILCS 2014 sample included 30 enumeration areas in urban and 18 enumeration areas in rural communities per month.
The method of representative probability sampling was used to frame the sample. At regional level, all communities were grouped into two categories - towns and villages. According to this division, a two-tier sample was drawn stratified by regions and by Yerevan. All regions and Yerevan, as well as all rural and urban communities were included in the sample in accordance to the shares of their resident households within the total number of households in the country. In the first round, enumeration districts - that is primary sample units to be surveyed during the year - were selected. The ILCS 2015 sample included 30 enumeration districts in urban and 18 enumeration districts in rural communities per month.

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

Face-to-face [f2f]

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

The Questionnaire is filled in by the interviewer during the least five visits to households per month. During face-to-face interviews with the household head or another knowledgeable adult member, the interviewer collects information on the composition and housing conditions of the household, the employment status, educational level and health condition of the members, availability and use of land, livestock, and agricultural machinery, monetary and commodity flows between households, and other information.

The 2007 survey questionnaire had the following sections: (1) "List of Household Members", (2) "Migration", (3) "Housing and Dwelling Conditions", (4) "Employment", (5) "Education", (6) "Agriculture", (7) "Food Production", (8) "Monetary and Commodity Flows between Households", (9) "Health (General) and Healthcare", (10) "Debts", (11) "Subjective Assessment of Living Conditions", (12) "Provision of Services", (13) "Social Assistance", (14) "Households as Employers for Service Personnel", and (15) "Household Monthly Consumption of Energy Resources".

The Diary is completed directly by the household for one month. Every day the household would record all its expenditures on food, non-food products and services, also giving a detailed description of such purchases; e.g. for food products the name, quantity, cost, and place of purchase of the product is recorded. Besides, the household records its consumption of food products received and used from its own land and livestock, as well as from other sources (e.g. gifts, humanitarian aid). Non-food products and services purchased or received for free are also recorded in the diary. Then, the household records its income received during the month. At the end of the month, information on rarely used food products, durable goods and ceremonies is recorded, as well. The records in the diary are verified by the interviewer in the course of 5 mandatory visits to the household during the survey month.

The Survey Diary has the following sections: (1) food purchased during the day, (2) food consumed at home during the day, (3) expenditures on food consumed away from home, (4) non-food products purchased and services obtained, (5) non-food products and services received free of charge, (6) household income and monetary inflows, (7) food products, which are usually consumed in small quantities during the day, (8) list of real estate, durable goods, and ceremonies. The interviewer’s manual provides detailed instructions for completing the questionnaire and the diary.

The Questionnaire, the Diary and the Interviewer's Manual are revised and adjusted, as appropriate, prior to the launch of the survey. Starting from 2012, data are codified under the "Classification of Individual Consumption by Purpose" (COICOP) classifier.

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

In 2007 the survey covered 48 urban and 289 rural areas with the sample size of 7,872 households.

摘要
---------------------------
由亚美尼亚共和国国家统计服务局（NSS）每年进行的综合生活条件调查（ILCS），构成了监测亚美尼亚生活条件的基础。ILCS是一项被普遍认可的最佳实践调查，用于收集有关家庭生活水平的资料。ILCS包含了关于家庭和个体福利的全面且宝贵的数据，为NSS提供了向公众提供关于人口收入、支出、贫困水平以及其他生活水平年度变化最新信息的机遇。
地理覆盖范围
---------------------------
城市和农村社区
分析单元
---------------------------
- 家庭；
- 个人。
数据类型
---------------------------
样本调查数据 [ssd]
抽样程序
---------------------------
在2001-2003年的调查中，采用了两阶段随机抽样方法；第一阶段涉及定居点——城市和村庄的选择，而第二阶段则专注于这些定居点中的家庭选择。调查遵循以家庭群组为单位的月度轮换原则。在2002年和2003年，家庭数量为387户，样本覆盖了2002年的14个城市和30个村庄，以及2003年的17个城市和20个村庄。
在2004-2006年的调查中，ILCS的抽样框架是利用2001年人口普查的地址数据库构建的；该数据库是在世界银行的技
术援助下开发的。亚美尼亚所有家庭的地址数据库被划分为48个层次，包括12个叶里温市社区。来自其他地区（marzes）的家庭根据以下三个类别进行分组：人口超过15,000的大城镇、村庄和其他城镇。大城镇形成了16个层次（唯一的例外是瓦约茨佐尔marz，那里没有大城镇）。村庄和其他城镇各自形成了10个层次。根据这种划分，按照marz级别的分层随机两阶段样本被开发。所有marzes，以及所有城市和农村定居点，根据居住在这些定居点的人口占国家总人口的百分比被纳入样本人口。在第一步中，选择了定居点，即初级样本单位：在当年进行调查的48个城镇中的43个，以及在全国951个村庄中的216个村庄被调查。在第二步中，选择了受访家庭：6,816户家庭（5,088户来自城市定居点，1,728户来自农村定居点）。因此，自1996年以来，第一次在marz级别上具有代表性的调查数据。
在2007-2012年的调查中，ILCS的抽样框架是根据2001年人口普查的地址数据库设计的，该数据库是在世界银行的技术援助下开发的。样本由两部分组成：核心样本和超样本。
1) 核心样本的创建涉及将抽样框架（亚美尼亚所有家庭的地址数据库）划分为48个层次，包括12个叶里温市社区。来自其他地区（marzes）的家庭根据以下三个类别进行分组：人口超过15,000的大城镇、村庄和其他城镇。大城镇形成了16个层次（唯一的例外是瓦约茨佐尔marz，那里没有大城镇）。村庄和其他城镇各自形成了10个层次。根据这种划分，按照marz级别的分层随机两阶段样本被开发。所有marzes，以及所有城市和农村定居点，根据居住在这些定居点的家庭占国家总家庭的百分比被纳入样本人口。在第一步中，使用PPS方法选择了调查的辅助单元（即当年将被调查的初级样本单位）。2007年的样本包括每月48个城市和18个农村辅助单元。
2) 超样本是从包括在MCA-Armenia农村道路修复项目中的村庄名单中抽取的。已经包括在核心样本中的村庄的辅助单元被排除在该名单之外。从剩余的辅助单元中，每月选择18个辅助单元。因此，农村样本量翻倍。
3) 在合并核心样本和超样本后，在第二步中选择了调查家庭。每月调查656户家庭，其中368户来自城市定居点，288户来自农村定居点。每月有82名调查员进行实地工作，他们的工作量是每月8户。2007年调查的家庭数量为7,872户（4,416户来自城市，3,456户来自农村地区）。
对于2013年的调查，ILCS的抽样框架是根据基于2001年人口普查结果的全国所有私人家庭的地址数据库设计的，该数据库是在世界银行的技术援助下开发的。使用了系统代表性概率抽样方法来构建样本。为了抽取样本，抽样框架被划分为32个层次，包括12个叶里温市社区（目前，为行政区域）。
根据这种划分，按照地区和叶里温进行了两阶段分层抽样。所有地区和叶里温，以及所有城市和农村社区，根据其居民家庭在国家总家庭数中的比例被纳入样本。在第一轮中，选择了辅助单元——即当年将被调查的初级样本单位。ILCS 2013的样本包括每月城市社区32个辅助单元和农村社区16个辅助单元。
在第二轮中，选择了将被调查的家庭。每月调查432户家庭，其中279户来自城市社区，153户来自农村社区。每月有48名调查员进行实地工作，他们的工作量是每月9户。
2014-2016年的样本框架是根据基于2011年人口普查结果的全国所有私人家庭的地址数据库设计的，该数据库是在世界银行的技术援助下开发的。使用了系统代表性概率抽样方法来构建样本。
在抽取样本时，抽样框架被划分为32个层次，包括12个叶里温市社区（目前，为行政区域）。根据这种划分，按照地区和叶里温进行了两阶段分层抽样。所有地区和叶里温，以及所有城市和农村社区，根据其居民家庭在国家总家庭数中的比例被纳入样本。在第一轮中，选择了辅助单元——即当年将被调查的初级样本单位。ILCS 2014的样本包括每月城市社区30个辅助单元和农村社区18个辅助单元。
抽样方法使用了代表性概率抽样。在区域层面，所有社区被分为两类——城镇和村庄。根据这种划分，按照地区和叶里温进行了两阶段分层抽样。所有地区和叶里温，以及所有农村和城市社区，根据其居民家庭在国家总家庭数中的比例被纳入样本。在第一轮中，选择了辅助单元——即当年将被调查的初级样本单位。ILCS 2015的样本包括每月城市社区30个辅助单元和农村社区18个辅助单元。
数据收集方式
---------------------------
面对面 [f2f]
研究工具
---------------------------
调查问卷由调查员在每月至少五次访问家庭时填写。在与户主或其他知识渊博的成年成员进行面对面访谈时，调查员收集有关家庭构成和住房条件、成员的就业状况、教育水平和健康状况、土地、牲畜和农业机械的可用性和使用情况、家庭间的货币和商品流动以及其他信息。
2007年的调查问卷包括以下部分：（1）“家庭成员清单”，（2）“迁移”，（3）“住房和居住条件”，（4）“就业”，（5）“教育”，（6）“农业”，（7）“食品生产”，（8）“家庭间的货币和商品流动”，（9）“健康（一般）和医疗保健”，（10）“债务”，（11）“生活条件的主观评估”，（12）“服务提供”，（13）“社会援助”，（14）“家庭作为服务人员雇主”，以及（15）“家庭每月能源资源消耗”。
调查日记由家庭直接填写一个月。每天家庭会记录其所有在食品、非食品产品和服务的支出，并对这些购买提供详细的描述；例如，对于食品产品，记录产品的名称、数量、成本和购买地点。此外，家庭还会记录其从自己的土地和牲畜以及从其他来源（例如，礼物、人道主义援助）收到的食品产品的消费和使用情况。非食品产品和服务的购买或免费获得也会在日记中记录。然后，家庭会记录其在本月收到的收入。月底，还会记录很少使用的食品产品、耐用消费品和仪式的信息。调查员在调查月份的5次强制访问过程中验证日记中的记录。
调查日记包括以下部分：（1）当天的食品购买，（2）当天的家庭食品消费，（3）在家庭外消费的食品支出，（4）购买的非食品产品和获得的服务，（5）免费获得的非食品产品和服
务，（6）家庭收入和货币流入，（7）通常在一天中少量消费的食品产品，（8）不动产、耐用消费品和仪式清单。调查员手册提供了填写问卷和日记的详细说明。
问卷、日记和调查员手册在调查开始前根据需要进行了修订和调整。从2012年开始，数据根据“按目的分类的个人消费”（COICOP）分类器进行编码。
响应率
---------------------------
在2007年，调查覆盖了48个城市和289个农村地区，样本量为7,872户。