Demographic and Health Survey 2013 - Liberia

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

The 2013 Liberia Demographic and Health Survey (LDHS) is designed to provide data for monitoring the population and health situation in Liberia. The 2013 LDHS is the fourth Demographic and Health Survey conducted in Liberia since 1986. The primary objective of the 2013 LDHS is to provide up-to-date estimates of basic demographic and health indicators. Specifically, the 2013 LDHS collected information on fertility levels, marriage, sexual activity, fertility preferences, family planning methods, breastfeeding practices, nutrition, childhood and maternal mortality, maternal and child health, and HIV/AIDS and other sexually transmitted infections (STIs). In addition, the 2013 LDHS provides estimates on HIV prevalence among adult Liberians.

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

National coverage

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

- Household
- Individual/ person
- Children age 0-5 years
- Woman age 15 to 49 years
- Man age 15 to 49 years

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

Sample survey data [ssd]

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

Sample Design
The sampling frame for the 2013 LDHS was developed by the Liberia Institute of Statistics and Geo-Information Services (LISGIS) after the 2008 National Population and Housing Census (NPHC). The sampling frame is similar to that used for the 2009 and 2011 Liberia Malaria Indicator Surveys (LMIS), except that the classification of localities as urban or rural was updated through the application of standardized definitions. The sampling frame excluded nomadic and institutional populations such as residents of hotels, barracks, and prisons. Notably, the sampling frame for the 2013 LDHS differs markedly from that used for the 2007 LDHS, which was based on the 1984 NPHC. Taken together, these differences may complicate data comparisons between surveys.

The 2013 LDHS followed a two-stage sample design that allowed estimates of key indicators for the country as a whole, for urban and rural areas separately, for Greater Monrovia and other urban areas separately, and for each of 15 counties. To facilitate estimates of geographical differentials for certain demographic indicators, the 15 counties were collapsed into five regions as follows:
North Western: Bomi, Grand Cape Mount, and Gbarpolu
South Central: Montserrado, Margibi, and Grand Bassa
South Eastern A: River Cess, Sinoe, and Grand Gedeh
South Eastern B: River Gee, Grand Kru, and Maryland
North Central: Bong, Nimba, and Lofa

Regional data were presented in the 2007 LDHS, the 2009 LMIS, and the 2011 LMIS. However, in contrast with these past surveys, the South Central region now includes Monrovia. Thus, data presented for the South Central region in this report is not directly comparable to that presented in the 2007 LDHS, the 2009 LMIS, or the 2011 LMIS.

The first stage of sample selection involved selecting sample points (clusters) consisting of enumeration areas (EAs) delineated for the 2008 NPHC. Overall, the sample included 322 sample points, 119 in urban areas and 203 in rural areas. To allow for separate estimates of Greater Monrovia and Montserrado as a whole, 44 sample points were selected in Montserrado; 16 to 26 sample points were selected in each of the other 14 counties.

The second stage of selection involved the systemic sampling of households. A household listing operation was undertaken in all the selected EAs from mid-September to mid-October 2012. From these lists, households to be included in the survey were selected. Approximately 30 households were selected from each sample point for a total sample size of 9,677 households. During the listing, geographic coordinates (latitude and longitude) were taken in the center of the populated area of each EA using global positioning system (GPS) units.

Because of the approximately equal sample sizes in each region, the sample is not self-weighting at the national level, and weighting factors have been added to the data file so that the results will be proportional at the national level.

All women age 15-49 who were either permanent residents of the selected households or visitors who stayed in the household the night before the survey were eligible to be interviewed. In half of the households, all men age 15-49 who were either permanent residents of the selected households or visitors who stayed in the household the night before the survey were eligible to be interviewed. In the subsample of households selected for the male survey, blood samples were collected for laboratory testing to detect HIV from eligible women and men who consented; in this same subsample of households, height and weight information was collected from eligible women, men, and children 0-59 months.

Further details on the sample design and implementation are given in Appendix A of the final report.

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

Face-to-face [f2f]

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

Three questionnaires were used for the 2013 LDHS: the Household Questionnaire, the Woman’s Questionnaire, and the Man’s Questionnaire. These questionnaires are based on MEASURE DHS standard survey questionnaires and were adapted to reflect the population and health issues relevant to Liberia. Input was solicited from various stakeholders representing government ministries and agencies, nongovernmental organizations, and international donors.

Given that there are dozens of local languages in Liberia, most of which have no accepted written script and are not taught in the schools, and given that English is widely spoken, it was decided not to attempt to translate the questionnaires into vernaculars. However, many of the questions were broken down into a simpler form of Liberian English that interviewers could use with respondents.

The Household Questionnaire was used to list all the usual members of and visitors to selected households. Some basic demographic information was collected on the characteristics of each person listed, including his or her age, sex, education, and relationship to the head of the household. For children under age 18, survival status of the parents was determined. The data on age and sex of household members obtained in the Household Questionnaire were used to identify women and men who were eligible for individual interview and HIV testing. The Household Questionnaire also collected information on characteristics of the household’s dwelling unit, such as the source of water, type of toilet facility, materials used for the floor of the house, ownership of various durable goods, ownership and use of mosquito nets, and information on household out-of-pocket health-related expenditures. The Household Questionnaire was also used to record height and weight measurements of children 0-59 months and eligible adults. Also recorded was whether or not eligible adults consented to HIV testing.

The Woman’s Questionnaire was used to collect information from all eligible women age 15-49.

The Man’s Questionnaire was administered to all men age 15-49 in the subsample of households selected for the male survey in the 2013 LDHS sample. The Man’s Questionnaire collected much of the same information as the Woman’s Questionnaire, but was shorter because it did not contain a detailed reproductive history or questions on maternal and child health.

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

All questionnaires were returned to the LISGIS central office in Monrovia for data processing, which consisted of office editing, coding of open-ended questions, data entry, and editing computer-identified errors. The data were processed by a team of 12 data entry clerks, two data editors, one data entry supervisor, and two administrators of questionnaires; the latter checked that the clusters were completed according to the sample selection and that all members of the household eligible for individual interview were identified. Secondary editing was led by an LDHS coordinator. Several LISGIS staff took on the responsibility of receiving the blood samples from the field and checking them before sending them to the Montserrado Regional Blood Bank for storage. Data entry and editing using CSPro software was initiated in April 2013 and completed in late August 2013.

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

A total of 9,677 households were selected for the sample, of which 9,386 were occupied. Of the occupied households, 9,333 were successfully interviewed, yielding a response rate of 99 percent.

In the interviewed households, 9,462 eligible women were identified for individual interview; of these, complete interviews were conducted with 9,239 women, yielding a response rate of 98 percent. In the subsample of households selected for the male survey, 4,318 eligible men were identified and 4,118 were successfully interviewed, yielding a response rate of 95 percent. The lower response rate for men was likely due to their more frequent and longer absences from the household.

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

The estimates from a sample survey are affected by two types of errors: nonsampling errors and sampling errors. Nonsampling errors are the results of mistakes made in implementing data collection and data processing, such as failure to locate and interview the correct household, misunderstanding of the questions on the part of either the interviewer or the respondent, and data entry errors. Although numerous efforts were made during the implementation of the 2013 Liberia Demographic and Health Survey to minimize this type of error, nonsampling errors are impossible to avoid and difficult to evaluate statistically.

Sampling errors, on the other hand, can be evaluated statistically. The sample of respondents selected in the 2013 LDHS is only one of many samples that could have been selected from the same population, using the same design and expected size. Each of these samples would yield results that differ somewhat from the results of the actual sample selected. Sampling errors are a measure of the variability between all possible samples. Although the degree of variability is not known exactly, it can be estimated from the survey results.

Sampling error is usually measured in terms of the standard error for a particular statistic (mean, percentage, etc.), which is the square root of the variance. The standard error can be used to calculate confidence intervals within which the true value for the population can reasonably be assumed to fall. For example, for any given statistic calculated from a sample survey, the value of that statistic will fall within a range of plus or minus two times the standard error of that statistic in 95 percent of all possible samples of identical size and design.

If the sample of respondents had been selected as a simple random sample, it would have been possible to use straightforward formulas for calculating sampling errors. However, the 2013 LDHS sample is the result of a multi-stage stratified design, and, consequently, it was necessary to use more complex formulas. Sampling errors are computed in either ISSA or SAS, using programs developed by ICF International. These programs use the Taylor linearization method of variance estimation for survey estimates that are means, proportions, or ratios. The Jackknife repeated replication method is used for variance estimation of more complex statistics such as fertility and mortality rates.

The Taylor linearization method treats any percentage or average as a ratio estimate, r = y/x, where y represents the total sample value for variable y, and x represents the total number of cases in the group or subgroup under consideration.

Further details on sampling errors calculation are given in Appendix B of the final report.

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

Data quality tables were produced to review the quality of the data:
- Household age distribution
- Age distribution of eligible and interviewed women
- Completeness of reporting
- Births by calendar years
- Reporting of age at death in days
- Reporting of age at death in months
- Completeness of information on siblings
- Sibship size and sex ratio of siblings

Note: The tables are presented in APPENDIX C of the final report.

摘要
---------------------------

2013年利比里亚人口与健康调查（LDHS）旨在为利比里亚的人口与健康状况监测提供数据。2013年LDHS是自1986年以来在利比里亚进行的第四次人口与健康调查。2013年LDHS的主要目标是提供最新的人口与健康基本指标估计值。具体而言，2013年LDHS收集了有关生育水平、婚姻、性行为、生育偏好、计划生育方法、母乳喂养实践、营养、儿童和孕产妇死亡率、孕产妇及儿童健康，以及艾滋病/艾滋病病毒（HIV）和其他性传播感染（STIs）的信息。此外，2013年LDHS还提供了利比里亚成人中HIV流行率的估计值。

地理覆盖范围
---------------------------

全国覆盖

分析单元
---------------------------

- 家庭
- 个人/个人
- 0-5岁儿童
- 15至49岁女性
- 15至49岁男性

数据类型
---------------------------

样本调查数据 [ssd]

抽样程序
---------------------------

抽样设计
...
（此处省略部分内容，以下为剩余翻译内容）

数据收集方式
---------------------------

面对面 [f2f]

研究工具
---------------------------

2013年LDHS使用了三种问卷：家庭问卷、女性问卷和男性问卷。这些问卷基于MEASURE DHS标准调查问卷，并针对利比里亚的相关人口与健康问题进行了调整。征求了代表政府部門和機構、非政府組織和国际捐贈者的各種利益相關方的意見。

鉴于利比里亚有数十种地方语言，其中大部分没有公认的书写脚本，且不在学校教授，而英语被广泛使用，因此决定不尝试将问卷翻译成本土语。然而，许多问题被简化为一种利比里亚英语的形式，调查员可以使用这种形式与受访者交流。

家庭问卷用于列出所选家庭的全部常驻成员和访客。收集了关于每个列出人员的某些基本人口信息，包括其年龄、性别、教育程度和与户主的关系。对于18岁以下的儿童，确定了父母的生存状况。家庭问卷中获得的家庭成员年龄和性别数据用于确定有资格进行个别访谈和HIV检测的女性和男性。家庭问卷还收集了有关家庭住宅单元特征的信息，例如水源、厕所设施类型、房屋地板使用的材料、各种耐用品的所有权、蚊帐的所有权和使用情况，以及家庭医疗相关支出信息。家庭问卷还用于记录0-59个月儿童和有资格成人的身高和体重测量结果。还记录了有资格成人是否同意进行HIV检测。

女性问卷用于收集所有有资格的15-49岁女性的信息。

男性问卷用于2013年LDHS样本中选定的男性调查家庭的子样本中的所有15-49岁男性。男性问卷收集了与女性问卷大部分相同的信息，但由于不包含详细的生育史或孕产妇及儿童健康问题，因此较短。

数据清理操作
---------------------------

所有问卷均返回蒙罗维亚的LISGIS中央办公室进行数据处理，包括办公室编辑、开放式问题的编码、数据录入和编辑计算机识别的错误。数据由一个由12名数据录入员、两名数据编辑员、一名数据录入监督员和两名问卷管理员组成的小组处理；后者检查集群是否根据抽样选择完成，以及是否已识别出所有有资格进行个别访谈的家庭成员。二级编辑由LDHS协调员领导。一些LISGIS员工负责从现场接收血液样本，并在将其发送到蒙罗维亚区域血液银行储存之前进行检查。使用CSPro软件进行数据录入和编辑始于2013年4月，并于同年8月底完成。

应答率
---------------------------

总共选择了9,677个家庭作为样本，其中9,386个被占用。在占用的家庭中，有9,333个家庭成功接受了访谈，应答率为99%。

在访谈的家庭中，确定了9,462名有资格进行个别访谈的女性；其中，对9,239名女性进行了完整的访谈，应答率为98%。在选定的男性调查家庭的子样本中，确定了4,318名有资格的男性，其中4,118人成功接受了访谈，应答率为95%。男性应答率较低可能是由于他们更频繁且更长时间的离家。

抽样误差估计
---------------------------

样本调查的估计受到两种类型误差的影响：非抽样误差和抽样误差。非抽样误差是由于在实施数据收集和数据处理过程中出现的错误造成的，例如未能找到和访谈正确的家庭、访谈员或受访者对问题的误解，以及数据录入错误。尽管在实施2013年利比里亚人口与健康调查期间做出了大量努力以最大限度地减少此类错误，但非抽样误差是无法避免且难以进行统计评估的。

另一方面，抽样误差可以统计评估。2013年LDHS中选定的受访者样本只是从同一人群中选择、使用相同设计和预期规模的许多样本之一。每个这样的样本都会产生与实际选定样本结果略有不同的结果。抽样误差是衡量所有可能样本之间变异性的指标。尽管变异程度无法确切知道，但可以从调查结果中估计出来。

抽样误差通常以特定统计量（均值、百分比等）的标准误差来衡量，它是方差的平方根。标准误差可用于计算置信区间，其中可以合理地假设总体真实值将落在该区间内。例如，对于从样本调查中计算出的任何给定统计量，该统计量的值将在95%的所有可能样本（具有相同大小和设计）的标准误差的两倍范围内。

如果受访者样本被选为简单随机样本，则可以使用简单的公式来计算抽样误差。然而，2013年LDHS样本是分层多阶段设计的产物，因此有必要使用更复杂的公式。抽样误差使用ISSA或SAS中的程序计算，这些程序由ICF国际公司开发。这些程序使用泰勒线性化方法对调查估计的均值、比例或比率进行方差估计。对于更复杂的统计量，如生育率和死亡率，使用Jackknife重复复制方法进行方差估计。

泰勒线性化方法将任何百分比或平均数视为比率估计，r = y/x，其中y代表变量y的总样本值，x代表考虑的组或子组的总案例数。

关于抽样误差计算的更多详细信息见最终报告附录B。

数据评估
---------------------------

生成了数据质量表以审查数据质量：
- 家庭年龄分布
- 有资格和接受访谈的妇女年龄分布
- 报告的完整性
- 日历年度出生情况
- 死亡年龄报告（按天数）
- 死亡年龄报告（按月份）
- 兄弟姐妹信息完整性
- 兄弟姐妹家庭规模和性别比

注意：表单在最终报告的附录C中呈现。）