Often, the DHS EdData survey is linked to a Demographic and Health Survey (DHS) conducted in the same country. The DHS survey is designed to provide current and reliable information on key indicators of social development, including fertility levels and trends, family planning knowledge and use, and maternal and child health. The most recent DHS in Malawi was conducted from July through November 2000. When the DHS survey and the DHS EdData survey are linked in a country, households sampled for the DHS survey are revisited and in-depth information on education is collected. For each household, data from the two surveys are statistically linked to create a joint data set that provides information on a wide range of topics.
Typically, a DHS EdData survey begins fieldwork within a month or two of the completion of the DHS survey to maximise the chances of locating and interviewing the same households interviewed for the DHS survey and to increase the likelihood that the household characteristics (such as composition and wealth) are unchanged during the period of both surveys. However, in Malawi, there was a 17-month gap between the end of the fieldwork for the 2000 Malawi DHS and the start of fieldwork for the 2002 MDES. Because of this gap, the 2000 Malawi DHS households were not revisited for the 2002 MDES. However, as discussed below, the sampling frame used for the 2000 Malawi DHS, which was based on enumeration areas defined in the 1998 Malawi Census of Population and Housing, was used to structure the sampling frame for the 2002 MDES.
The principal aim of the 2002 Malawi DHS EdData Survey (MDES) is to provide upto-date information on education among children of primary school age (age 6-13). The survey focuses on factors influencing household decisions about children's school attendance. These data supplement the data collected by the Ministry of Education, Science, and Technology by focusing on attendance rather than enrolment and exploring the costs of schooling (monetary and non-monetary) and parent/guardian attitudes about schooling.The survey provides data on topics such as the age of children's first school attendance and dropout; the reasons for overage first-time enrolment in school, never enrolling in school, and dropout; the frequency of and reasons for pupil absenteeism; household expenditures on schooling and other contributions to schooling; distances and travel times to schools; and parent/guardian perceptions of school quality and the benefits and disadvantages of schooling.
The 2002 MDES was designed to supplement education data sources and to provide data to assist policy-makers in evaluating education programmes in the country. In broad terms, the 2002 MDES aims to:
• Provide baseline data on key education indicators
• Assist in the evaluation of Malawi's education programmes
• Advance survey methodology in Malawi and contribute to national and international databases.
In more specific terms, the 2002 MDES was designed to:
• Provide data on the schooling status of Malawian children of primary school age and on factors influencing whether children ever enrol in school and why pupils drop out of school
• Quantify household expenditures on children's schooling and examine differential patterns of expenditure by various background characteristics
• Measure parent/guardian attitudes about schoolin - including their perceptions of the quality of schooling and of the effects of Free Primary Education, to provide an understanding of attitudes that shape parents' and guardians. willingness to send their children to school
• Measure the frequency of pupil absenteeism and the reasons for missing school in order to suggest approaches to maximise pupil attendance.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
The survey focuses on factors influencing household decisions about children's school attendance.
The following themes are covered in the survey:
- Age of children's first school attendance and dropout
- Reasons for overage first-time enrolment in school
- Never enrolling in school, and dropout
- Frequency of and reasons for pupil absenteeism
- Household expenditures on schooling and other contributions to schooling;
- Distances and travel times to schools;
- Parent/guardian perceptions of school quality and the benefits and disadvantages of schooling
The principal aim of the 2002 Malawi DHS EdData Survey (MDES) is to provide up-to-date information on education among children of primary school age (age 6-13)
Producers and sponsors
National Statistical Office (NSO)
Government of Malawi
Ministry of Education, Science and Technology
Center for Educational Research and Training
United States Agency for International Development
Department for International Development
Canadian International Development Agency
German Agency for Technical Cooperation
Japan International Cooperation Agency
USAID Office of Education in the Bureau for Economic Growth, Agriculture and Trade
Department for International Development
Canadian International Development Agency
The sample for the 2002 MDES was based on the sampling frame for the 2000 MDHS, which was designed to provide estimates of health and demographic indicators. The discussion in this section first addresses the sample design for the 2000 Malawi DHS, then the subsequent design for the 2002 MDES.
The 2000 Malawi DHS was designed to provide estimates at the national and regional levels and for urban and rural areas. It was also designed to provide estimates of some health and demographic indicators at the sub-regional level in 11 districts.
The 2000 Malawi DHS sample points (clusters) were systematically sampled from a list of enumeration areas (EAs) defined in the 1998 Malawi Census of Population and Housing. A total of 560 clusters were drawn from the census sample frame: 449 in rural areas and 111 in urban areas. After selecting the 560 clusters, the NSO trained teams to conduct the comprehensive listing of households and to update maps in the selected clusters. Nine listing teams conducted a comprehensive listing of households and updated maps in the selected clusters, from April through May 2000. This exercise provided a basis for second-stage sampling for the 2000 Malawi DHS-and later, for the 2002 MDES.
After the listing operation, households to be included in the 2000 Malawi DHS were selected; the number of households selected per cluster was inversely proportional to the size of the cluster. In the Malawi DHS sampling frame, as in the 2002 MDES sampling frame, the number of EAs selected in each district was not proportional to the total population; rather, urban areas were oversampled in order to generate unbiased urban estimates.
As part of the 2002 MDES pre-test, a verification exercise was conducted in one urban and two rural enumeration areas around Zomba to estimate what percentage of households identified at the time of the 2000 household listing would be found during the 2002 MDES fieldwork. During this verification exercise-using structure numbers that were written on buildings during the household listing, and the name of the household head at the time of the listing exercise-92 percent of the urban and 95 percent of the rural households were located. These results suggested that the household listing conducted in 2000 as part of the Malawi DHS remained usable for purposes of the 2002 MDES.
While structures and households were still identifiable, in many instances, the household head (and sometimes the entire household) had changed between 2000 and 2002. In 52 percent of the households in the urban area and 15 percent of the households in the rural areas, the name of the household head was different in 2002 than in 2000. In other words, household composition had changed for over half of the households in the urban area and for one-seventh of the households in the rural areas, supporting the decision not to try to link information from the 2000 Malawi DHS and 2002 MDES at the household level.
For the 2002 MDES, 129 EAs-111 in rural areas and 18 in urban areas-were selected from the 560 EAs in the 2000 Malawi DHS sample.The 2002 MDES was designed to provide estimates at the national and regional levels and for urban and rural areas.
A total of 3,866 households were selected, of which 3,325 were occupied. Of the 3,325 occupied households, 3,290 were interviewed successfully, yielding a household response rate of 99 percent. In the interviewed households, 2,048 parents/guardians were identified to be interviewed. Completed interviews were conducted with all of these parents/guardians, yielding a response rate of 100 percent. Since the parent/guardians responded to the questions for their children and the children for whom they were responsible, the Eligible Child Questionnaire response rate reflects the percentage of eligible children for whom data were collected. A total of 3,755 eligible children were identified and data were collected on 3,752 of these children, yielding a response rate of nearly 100 percent.
In conjunction with the 2000 MDHS, before the final household selection, a complete household listing operation was completed for each selected EA. This household listing was also used for the 2002 MDES. Based on these household lists, the household selection was then implemented to maintain a selfweighted sample in each domain but the sampling rates differ between districts. Therefore, the total 2000 MDHS and the 2002 MDES samples are weighted, and a final weighting adjustment is required to provide national estimates.
Dates of Data Collection
Data Collection Mode
Six interviewing teams carried out data collection for the 2002 MDES. Each team was composed of one supervisor, six interviewers, and one driver. Staff from NSO and MoEST coordinated and supervised fieldwork activities. ORC Macro staff also participated in field supervision. In the field, local guides assisted interviewing teams in locating selected households for interviews.
Data Collection Notes
Training of field staff for the main survey was conducted over a 2-week period in May 2002. A total of 46 persons participated in the main survey training for interviewers, including the 6 supervisors. The training was conducted using the DHS EdData survey training procedures, including instruction in general interviewing techniques and field procedures, class presentations on the questionnaires, mock interviews between participants, and classroom tests. The training included practiceinterviews using the questionnaire in English and the two local languages into which the questionnaires had been ranslated—Chichewa and Chitumbuka. Discussions of the translations were also an important part of the training programme. Supervisors were trained during a 1-day session.
All children age 6-14 were targeted to have data collected about them from a parent or guardian.
National Statistical Office
Government of Malawi
Ministry of Education, Science and Technoloy
Government of Malawi
Three questionnaires were used for the 2002 MDES: the Household Questionnaire, the Parent/Guardian Questionnaire, and the Eligible Child Questionnaire.
The three purposes of the MDES Household Questionnaire were to
1) list all household members and visitors to the household,
2) identify which children were eligible (qualified) to be covered by the Eligible Child Questionnaire, and
3) identify a parent or guardian as the respondent for each eligible child. Children age 6-14 were eligible to be covered by the Eligible Child Questionnaire.
The Parent/Guardian Questionnaire collected background information on each parent/guardian respondent and on general education issues. Information was collected on the parent/guardian’s age, education, literacy, and religion. Questions were asked about the walking time and distance to the nearest primary and secondary schools, and about household participation in school activities. Information was also collected on each primary school attended by the children for whom the parent/guardian responded, including the school type and location, the reason for selection of that school, and perceived school quality.
The Eligible Child Questionnaire collected different kinds of information about each eligible child, depending on the child’s schooling status. While the subject of the Eligible Child Questionnaire was the eligible child and his/her schooling, the respondent for the questionnaire was the child’s parent/guardian, as the purpose of the questionnaire was to collect information on issues from the parent/guardian’s perspective. Data were collected on the following topics, according to a child’s schooling status:
• Schooling background and participation during the current school year (attended school during the 2002 school year, dropped out of school, or never attended school)
• Frequency of and reasons for pupil absenteeism, household expenditures on schooling, other costs of schooling (for children who attended school during the 2001 school year)
• Reasons for dropping out of school (for children who have dropped out of school)
• Reasons for not attending school during the 2002 school year (for children who have never attended school)
• Children’s eating patterns
In April, the questionnaires were pre-tested in Chichewa in and around Zomba. A total of 108 households were interviewed and 120 Parent/Guardian Questionnaires and 367 Eligible Child Questionnaires were completed. Based on the results of the pre-test, minor changes in the pre-test survey questionnaires were made before the main survey fieldwork was conducted.
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
National Statistical Office and ORC Macro. 2003. Malawi DHS EdData Survey 2002:
Education Data for Decision-making. Calverton, Maryland U.S.A.: National Statistical Office and ORC Macro.
Accessed from www.nsomalawi.mw on February 2, 2013.
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.