get_ego_age_distribution

Usage

get_ego_age_distn(ego.dat, only_females = TRUE, warn.single.sex = TRUE)

Arguments

ego.dat: the ego dataset (probably from prep_dhs_sib_histories)
only_females: should only females be used to calculate age distribution? (default: True)
warn.single.sex: when only_females = FALSE, warn if ego.dat holds only one respondent sex, since the result then cannot standardise rates for the other. Set FALSE only when the caller reports the problem itself, as aggregate_maternal_estimates does

Value

dataframe with distribution of respondent ages by 5-year category. Columns age.cat, total and agegrp_prop, plus sex when only_females = FALSE; see Details

Details

ego_dat is assumed to have the columns wwgt, age.cat and sex

The age groups used are those returned by reproductive_age_groups.

When only_females = TRUE (the default), respondents are restricted to females and a single age distribution is returned, with agegrp_prop summing to 1 across age groups.

When only_females = FALSE, a separate age distribution is returned for each respondent sex: the result gains a sex column, and agegrp_prop sums to 1 within each sex. This is what aggregate_maternal_estimates needs in order to weight each sex's age-specific rates by its own respondents' age structure.

There is no male age distribution in DHS data, and you do not need one

only_females = FALSE splits whatever respondents are in ego.dat; it does not conjure a sex that was never interviewed. DHS sibling histories come from the women's file, so a DHS ego.dat is entirely female and only_females = FALSE returns the female distribution with a sex column attached, and warns.

It would be reasonable to conclude from Chap16_AM/AM_rates.do that this blocks reproducing a published male rate, since its get_age_distributions takes the men's age distribution from the men's recode (MR) file. In practice it does not. Published DHS reports standardise both sexes by the age distribution of the survey respondents — the women — which is exactly what only_females = TRUE returns here.

Checked against four published tables spanning DHS phases 4 to 8:

Survey	male rate, this way	published
Malawi 2000	11.064	11.1
Rwanda 2005	7.393	7.39
Rwanda 2014-15	2.961	2.96
Gambia 2019-20	3.133	3.13

Standardising men by a male distribution built from the MR file instead gives 11.162, 7.285 and 2.881 for the first three — further from the published figures in every case.