Technical Notes

The Demographic and Health household Surveys (DHS) started collecting data on education for the de facto and de jure household populations during the second phase of the program (DHS-II). Typically, for every household member and every visitor who spent the night preceding the survey in the house and who was 5 or 6 years old or older[1] the following question was asked: Has (name of household member) ever been to school? If the answer was affirmative, the respondent was then asked, What is the highest level (name of household member) attended? and What is the highest grade (name of household member) completed at that level? Additionally, for every household member older than 5 or 6 but less than 25 years, current school attendance was ascertained by asking: Is (name of household member) still in school? With this information it is possible to examine educational attainment, enrollment levels, gender gaps and other gaps in enrollment, and continuation rates for representative samples of the total populations. In this module, tables based on household-level data are calculated for the de facto population (For more information, go to: ).

The following notes explain certain nuances in data processing and indicator calculations. They serve to answer possible questions, as well as to caution on the use of the data. Many concerns are grounded in the general reason of the (DHS) being intended for the demographic and health sectors, so the education section of the survey is not as robust and controlled for as one would hope. Realizing the need for the education portion to be expanded, a new instrument has been developed in the year 2000 as an add-on for the DHS surveys, called EdData ( For more information, go to: ).

Indicator numbers in the parenthesis show those that are affected by the note.

  1. (Indicators #18-#20) “Recent cohort” refers to different cohorts in different grades, depending on the starting age in a country.

Θ

  1. (Indicators #4-#6, #10-#13, #14-#20) While the indicators by cycle and by grade should be the same, when the cycle in a country matches the grade indicated, these can differ slightly as a result of a missing data factor: when a respondent (usually a mother, but often other members of the household, for example grandmother or uncle) knows what level of schooling the child attends, but is not aware of the grade (s)he is in.

Θ

  1. (Indicators #7-#9) Grade 1 Intake:
  2. Indicators are compiled from data on current grade and level attendance and on number of years completed.
  3. Some countries’ data does not have anyone in the category of 0 years completed of primary, while currently attending school. This could be explained by two factors:
  4. those who are currently in first grade are indicated as having completed first grade, or
  5. those who are currently in first grade are indicated as not being in primary because they have not completed a year.
  6. In other countries’ data there are a few subjects with 0 completed years, but the figure is by far smaller than those with 1 year completed. This could be explained by
  7. answers being not uniformly recorded, or
  8. the survey being done in between the school years or at times when the school is not functioning.

Θ

  1. (Indicators #4-#8, #12-#13) Timing of the survey
  2. The indicators (especially, Net) are affected by the limitations of child’s age at the beginning of the school year and the start of the school year in a particular country being unknown.
  3. The indicators (especially, Net Grade 1 Intake) are highly sensitive to the timing of the survey and therefore can obtain less precision. For example, if primary school entrance age is 6 years and the child is reported to be 7 years old at the time of the survey, (s)he is listed as over the official school age.
  4. The drop-out rates are affected, because there might be more drop-outs later in the school year.
  5. The fieldwork for urban and rural areas not being conducted at the same time in certain countries, there could also be discrepancies in quintiles and localities.

Θ

  1. (Indicators #10-#11, #15-#20, #27) Grade attendance and competition is independent from the level of schooling reported. For example, if an individual has completed primary and is now in secondary 2nd Grade, assuming that the system is 5 years of primary, (s)he is counted as having completed 6 Grades.

Θ

  1. (Indicators #12-#13) Drop-out rates are calculated using the number of children who have reported to have ever been in school as the denominator.

Θ

  1. (Indicators #1-#9, #13-#21, #24-#27) Participation and attainment by older cohorts (for example, 15-19 year olds) of certain levels of education and/or grades being higher than that by younger cohorts, could be explained by the significant number of over-aged children in the system.

Θ

  1. Armenia, Kazakhstan, Kyrgyz Republic, Turkmenistan, and Uzbekistan present an additional challenge in the data being recorded in terms of Basic education, rather than by cycles. A decision was made to divide the Basic education code into Primary cycle, i.e. those attending 4 years of Basic education, and Secondary, i.e. those attending more than 4 years of Basic education, plus an additional number of years after basic education.

Θ

Symbols used:

* Data is not available or considered reliable.

(87.3)Data is considered only marginally reliable; when the number of observations is too small to provide data that are even marginally reliable, the numbers are replaced by asterisks (*).

[1] In general, these questions were asked of household members 5 or 6 years or older. Two exceptions were in Egypt and Morocco, where these questions were asked of the population 3 years or older and 7 years or older, respectively.