Comparability — Benin [top]

In the 2013 sample, there is substantial heaping in the month of January.

Comparability — Burkina Faso [top]

In 1996 sample, large numbers of respondents are coded to "unknown" because they provided their age (in full years) rather than the date of birth. In 2006, the respondents provided their date of birth, if known.

Comparability — Cameroon [top]

In 1976, large numbers of respondents are coded to "unknown" because they provided their age at last birthday rather than the full date of birth. The 1987 sample only provides the age of the respondents. In 2005, large numbers of respondents are coded to "unknown" because they only provided their year of birth

Comparability — Côte d'Ivoire [top]

In 1998, about 70% of cases had no response.

Comparability — Egypt [top]

There are large amounts of nonresponse in the samples.

Comparability — El Salvador [top]

For the 1992 sample, there is considerable heaping in the month of June (16.73%), compared to the other birth months. This heaping is not present in the 2007 sample.

Comparability — Fiji [top]

In 1976 sample, the month of birth was grouped in two categories: born between January-September and born between September-December. That sample is not included here.

In 1986 sample, large numbers of respondents are coded to "unknown" because they provided their year of birth rather than their full date of birth.

In the 2014 sample, the data has heaping in the month of June.

Comparability — Greece [top]

The Greek samples of 1981 and 1991 only asked the question of children born within the last two calendar years.

Comparability — Guinea [top]

Roughly 95% of the cases had no response.

Comparability — Indonesia [top]

Roughly half the population in 1980 has no response.

Comparability — Iran [top]

In the samples (2006 and 2011), month of birth is reported according to the Hijiri (i.e., Persian) calendar. The Hijiri month was matched with the Gregorian month in which it began. For example, the month Bahman of the Hijiri calendar corresponds to the 30-day period that falls between January and February in the Gregorian calendar. Therefore, the month of Bahman corresponds to January in the harmonized variable. See the unharmonized source variable to recover the information in both the Hijiri and Gregorian calendars.

There are a large number of non-responses. There are an abnormally large number of responses indicating birth in Farvardin (March) or Shahrivar (August).

Comparability — Iraq [top]

There is considerable heaping of the responses in January and especially July, probably as an artifact of the instructions.

Comparability — Jamaica [top]

The variable BIRTHMO is not available for the 1982 sample. The 1991 and 2001 samples are comparable.

There are a large number of non-responses.

Comparability — Lesotho [top]

The 2006 sample has substantial heaping in June.

Comparability — Malawi [top]

For an unknown reason, responses indicating birth in the first six months (January-June) is roughly twice that of the latter six months (July-December).

Comparability — Malaysia [top]

Roughly 40% of the cases had no response.

Comparability — Rwanda [top]

The 1991 sample has many more non-responses than 2002 and 2012. In the 2012 sample, there is substantial heaping in the month of January.

Comparability — Senegal [top]

For the 2013 sample, there is considerable heaping in the month of July (35.76%), compared to the other birth months. The enumeration instructions indicate that if the birth month is unknown, then the month of July should be entered as the response.

Comparability — Thailand [top]

The 1980, 1990 and 2000 Thai samples report both the solar and lunar month of birth. We use programming to create 12 categories (one per month).

Comparability — Vietnam [top]

The variable BIRTHMO is comparable across samples. The only difference is that the 2009 sample has an "unknown" category, while the 1989 and 1999 samples do not.