Since the first LMIS in 2005, the NMCP and its partners have scaled-up malaria interventions in all parts of the country. In order to determine the progress made in malaria control and prevention in Liberia since 2005, the 2009 Liberia Malaria Indicator Survey (LMIS) was designed to provide data on key malaria indicators including mosquito net ownership and use, as well as prompt and effective treatment with ACT.
The key objectives of the 2009 LMIS were to:
• Measure the extent of ownership and use of mosquito bednets
• Assess coverage of the intermittent preventive treatment program to protect pregnant women
• Identify practices used to treat malaria among children under five and the use of specific antimalarial medications
• Measure the prevalence of malaria and anemia among children age 6-59 months
• Assess malaria-related knowledge, attitudes, and practices in the general population.
Another objective of the survey was to transfer knowledge about best practices in survey implementation and to transfer skills to Liberian counterparts related to survey design, training, budgeting, logistics, data collection, monitoring, data processing, analysis, report drafting, and data dissemination.
Kind of data
Sample survey data [ssd]
Unit of analysis
- Children age 0-5
- Women age 15 to 49
Producers and sponsors
National Malaria Control Program
Ministry of Health and Social Welfare
Liberia Institute for Statistics and Geo-Information Services
Ministry of Health and Social Welfare
Technical assistance and medical supplies and equipment for the survey
United States Agency for International Development
The LMIS sample was designed to produce most of the key indicators for the country as a whole, for urban and rural areas separately, and for Monrovia and each of five regions that were formed by grouping the 15 counties. The regional groups are as follows:
1 Greater Monrovia
2 North Western: Bomi, Grand Cape Mount, Gbarpolu
3 South Central: Montserrado (outside Monrovia), Margibi, Grand Bassa
4 South Eastern A: River Cess, Sinoe, Grand Gedeh
5 South Eastern B: River Gee, Grand Kru, Maryland
6 North Central: Bong, Nimba, Lofa
Thus, the sample was not spread geographically in proportion to the population, but rather equally across the regions, with 25 sample points or clusters per region. As a result, the LMIS sample is not selfweighting at the national level and sample weighting factors have been applied to the survey records in order to bring them into proportion.
The survey utilized a two-stage sample design (see Appendix A for details). The first stage involved selecting 150 clusters with probability proportional to size from the list of approximately 7,000 enumeration areas (EAs) covered in the March 2008 National Population and Housing Census. The EA size was the number of residential households residing in the EA recorded in the census. Stratification was achieved by separating each county into urban and rural areas. The urban areas in each county mainly consist of the county capital. Therefore the 15 counties plus Greater Monrovia (which has only urban areas) were stratified into 31 sampling strata, 15 rural strata and 16 urban strata. Samples were selected independently in every stratum, with a predetermined number of EAs to be selected. Implicit stratification was achieved in each of the explicit sampling stratum by sorting the sampling frame according to districts and clan within each of the sampling stratum and by using the probability proportional to size selection procedure. Among the 150 EAs (clusters) selected, 69 were in urban areas and 81 were in rural areas.
In the second stage, for all of the selected EAs, a fixed number of households (30) was selected using an equal probability systematic sampling from a list of households in the EA. Because the census was still fresh (March 2008), it was decided to use the census household results as the sampling frame for household selection in the second stage, thus avoiding having to undertake a costly separate household listing operation. This involved borrowing the census questionnaire books for each of the selected EAs or clusters and copying information for all the occupied residential households recorded in the census book. These lists served as the sampling frame for household selection.
All women age 15-49 years who were either permanent residents of the households in the sample or visitors present in the household on the night before the survey were eligible to be interviewed in the survey. In addition, all children age 6-59 months who were listed in the household were eligible for the anemia and malaria testing component.
Note: See detailed description of the sample design in Appendix A of the final report.
Of the 4,485 households selected in the sample, 4,285 were found occupied at the time of the fieldwork. The shortfall is due to households that were away for an extended period of time, dwellings that could not be found in the field, and dwellings that were found to be vacant or destroyed. Of the existing households, 4,162 were successfully interviewed, yielding a household response rate of 97 percent.
In the households interviewed in the survey, a total of 4,512 eligible women were identified, of whom 4,397 were successfully interviewed yielding a response rate of 98 percent. The household response rates are slightly lower in the urban than rural sample, though they are almost equal for women. The principal reason for nonresponse among eligible women was the failure to find them at home despite repeated visits to the household.
Note: See summarized response rates in Table 1.2 of the final report.
Dates of collection
Mode of data collection
Two questionnaires were used in the LMIS: a Household Questionnaire and a Woman’s Questionnaire for all women age 15-49 in the selected households. Both instruments were based on the model Malaria Indicator Survey questionnaires developed by the Roll Back Malaria and DHS programs, as well as on previous surveys conducted in Liberia, including the 2005 LMIS and the 2007 LDHS. In consultation with the Technical Committee, NMCP and Macro staff modified the model questionnaires to reflect relevant issues of malaria in Liberia. 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 and visitors in the selected households. Some basic information was collected on the characteristics of each person listed, including age, sex, and relationship to the head of the household. The main purpose of the Household Questionnaire was to identify women who were eligible for the individual interview and children age 6-59 months for anemia and malaria testing. The household questionnaire also collected information on characteristics of the household's dwelling unit, such as the source of water, type of toilet facilities, materials used for the floor, roof, and walls of the house, ownership of various durable goods, and ownership and use of mosquito nets. In addition, this questionnaire was also used to record consent and results with regard to the anemia and malaria testing of young children.
The Woman’s Questionnaire was used to collect information from all women age 15-49 years and covered the following topics:
• Background characteristics (age, residential history, education, literacy, religion, dialect)
• Full reproductive history and child mortality
• Prenatal care and preventive malaria treatment for most recent birth
• Prevalence and treatment of fever among children under five
• Knowledge about malaria (symptoms, causes, ways to avoid, types of medicines, etc.).
Because almost all of the questions had been included in previous surveys and NMCP had experience with anemia and malaria testing, no formal pretest was held.
The processing of the LMIS questionnaire data began a few weeks after the fieldwork commenced. Completed questionnaires were returned periodically from the field to the NMCP office in Monrovia, where they were coded by data processing personnel recruited and trained for this task. The data processing staff consisted of a supervisor and an assistant from NMCP, a questionnaire administrator, five data entry operators, and two data editors, all of whom were trained by a Macro data processing specialist. Data were entered using the CSPro computer package. All data were entered twice (100 percent verification). The concurrent processing of the data was a distinct advantage for data quality, since NMCP was able to advise field teams of errors detected during data entry. The data entry and editing phase of the survey was completed in early May 2009.
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 2009 Liberia Malaria Indicator Survey (LMIS) 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 2009 LMIS 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.
A 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 and 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 2009 LMIS sample is the result of a multi-stage stratified design, and, consequently, it was necessary to use more complex formulae. The computer software used to calculate sampling errors for the 2009 LMIS is a Macro SAS procedure. This procedure used the Taylor linearization method of variance estimation for survey estimates that are means or proportions. The Jackknife repeated replication method is used for variance estimation of more complex statistics such as fertility and mortality rates.
Note: See detailed estimate of sampling error calculation in APPENDIX B of the final report.
Use of the dataset must be acknowledged using a citation which would include:
- the Identification of the Primary Investigator
- the title of the survey (including country, acronym and year of implementation)
- the survey reference number
- the source and date of download
Disclaimer and copyrights
The user of the data acknowledges that the original collector of the data, the authorized distributor of the data, and the relevant funding agency bear no responsibility for use of the data or for interpretations or inferences based upon such uses.