The UIS estimates certain key items of data that may be missing or incomplete. Where data for a country are entirely missing or where an estimate is not based on evidence from or about the country directly, the UIS does not publish the resulting country-level estimates. They are used only for the purposes of calculating regional or global aggregates or averages.
For the purposes of calculating the primary Gross Parity Index, the UIS may need to make one or more of the following types of estimate:
- An adjustment to account for over- or under-reporting, for example:
- To exclude enrolments in other programmes which have been reported together with enrolments at the primary level (very rare and does not usually result in a countrty level publishable estimate);
- To include enrolments in a type of education – such as private education or special education – not reported by the country; and/or
- To include enrolments in a part of the country not reported by the country.
- An estimate of the number of enrolments by sex
- An estimate of the population in the official age group for primary education (if neither UNPD nor the country itself can provide estimates of their own)
In all cases, in the first instance, estimates are based on evidence from the country itself (e.g. information from the data provider on the size of the missing component, via correspondence, publications or data on the ministry’s or National Statistical Office’s (NSO’s) Webpage, or via surveys conducted by other organisations). These figures may be published: as observed data (if the missing items are found in national source or the country submits them); as national estimates (if the country is persuaded to produce estimates and submit them in place of missing data); as UIS estimates (if the estimates are made by the UIS); or, occasionally, as inputs for regional and global aggregates and averages (if the evidence on which the estimates are based is weak or unsubstantiated).
Where no evidence is available for the reporting year from the country, estimates may be based on data already reported by the country in a previous year or on another data item which is available (e.g. total enrolments in primary education) and clearly linked to the missing item. These figures may be published as UIS estimates or, if the data reported are very old or the links with other available data are weak, they may only be used for the construction of regional or global aggregates and averages.
Where there is neither evidence from the country, nor data available in previous years, estimates are based on a similar country. Such figures will only be used for regional or global aggregates and averages.
Over- and under-reporting are typically adjusted by adding or subtracting a given percentage of enrolments thus assuming the same sex distribution as for the reported enrolments.
Over- and under-reporting are typically adjusted by adding or subtracting a given percentage of enrolments thus assuming the same age, grade and sex distribution as for the reported enrolments. (If more detailed evidence is available this will be used but that is typically not the case).
Population estimates – produced only where there are no other suitable estimates available either from UNPD or from the country itself – are produced only for countries which have reported education data to the UIS and for which population estimates from a reliable source are available in some years.