Use this option if women are the unit of analysis. Each record will be a woman of childbearing age.

Childbearing age is typically defined as ages 15-49, but some samples include women younger than 15 and/or include only ever-married women. Please consult variable universes and sample descriptions for the exact information for a given sample.

Most variables available under this option relate to women themselves, but some relate to the woman's latest birth (youngest child) or her household.

Use this option if children are the unit of analysis.

Most commonly, child records relate to children under age 5, but some samples include children under age 4 or under age 3. Please consult variable universes and sample descriptions for the exact information for a given sample.

Most variables available under this option relate to the children themselves, but there are also variables that provide information on their mothers and households. Data on children are collected through questions to mothers. Multiple children can have the same mother.

Use this option if births (regardless of current age) are the unit of analysis.

Birth records relate to all births, up to a total of 20 births, to women of childbearing age.

Some variables under this option relate to the characteristics of the births themselves (such as sex and date of birth). The majority of variables provide information on the mothers of those births and their households. Data on births are collected through questions to mothers. Multiple births can have the same mother.

Use this option if household members (of any age and sex) are the unit of analysis.

Household member records relate to all persons, including both usual residents and visitors, recorded on the listing of household members during the completion of the household survey form. In the sampling process, a random selection of households was drawn from a list of all households included in a chosen primary sampling unit (such as an enumeration areafrom the most recent national census). Household members are all persons included in each such randomly selected household.

Most variables under this option relate to the characteristics of the household members themselves (such as their age and relationship to the household head). Other variables provide information on the characteristics of the dwelling unit.

Use this option if men are the unit of analysis. Each record will be an adult non-elderly man, from the randomly selected households included in DHS samples.

Exact age ranges and the presence of marital status restrictions for men vary across samples. Please consult variable universes and sample descriptions for the exact information for a given sample.

Use this option if you are using retrospective calendar data, collected for approximately 5 years before the month of interview for women of childbearing age. Each record will be a month in which a pregnancy, birth, pregnancy termination, or, for some samples, contraceptive use may be recorded. Variables from other parts of the women's questionnaire (such as educational attainment at the time of interview) are also available for calendar data.