• At country level

The UIS estimates certain key items of data that may be missing or incomplete in order to have publishable estimates at the country level. Where this is not possible the UIS imputes missing values for use only for calculating regional and global aggregates.

For the purposes of calculating participation rates by age, the UIS may make one or more of the following:

• An adjustment to account for over- or under-reporting, for example:

o To include enrolments in a type of education – such as private education or special education – not reported by the country; and/or

o To include enrolments in a part of the country not reported by the country.

• An estimate of the number of enrolments in the given age group if the age distribution was not reported by the country.

• A redistribution of enrolments of unknown age (across known ages).

• An estimate of the population in the official age group for small countries (if neither the UN Population Division (UNPD) nor the country itself can provide estimates of their own).

In all cases 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 organizations) or on data from the country for a previous year. These figures may be published: (i) as observed data if the missing items are found in a national source; (ii) as national estimates if the country is persuaded to produce estimates and submit them in place of missing data; or (iii) as UIS estimates, if the estimates are made by the UIS.

The age distribution of enrolments is most commonly estimated from the age distribution reported in a previous year. If the country has never reported the age distribution of enrolments, the age distribution reported in another survey, if available, is used (such as Multiple Indicator Cluster Surveys (MICS) or Demographic Health Surveys (DHS)).

Enrolments of unknown age are redistributed across known ages if they constitute more than 5% of the total enrolments in that level of education. No estimation is made if they are 5% or less.

Population estimates by age for countries with small population – produced only where there are no other suitable estimates available either from UNPD or from the country itself – are made 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.

• At regional and global levels

Regional and global aggregates are derived from both publishable and imputed national data. Publishable data are the data submitted to the UIS by Member States or the result of an explicit estimation made by the Institute based on pre-determined standards. In both cases, these data are sent to Member States for review before they are considered publishable by the UIS.

When data are not available for all countries, the UIS imputes national data for the sole purpose of calculating regional averages. These imputed data are neither published nor otherwise disseminated.

Where data are available for a country for both an earlier and a more recent year than the missing year, a simple linear interpolation is made. Where data are only available for an earlier year, the most recent value is used as an estimate. Similarly, where data are only available for a more recent year, the last value is used as an estimate.

Where the relevant data are not available at all for a country, estimates may be based on another variable which is clearly linked to the item being estimated. For example, enrolments by age may be based on total enrolments.

Where no data are available for the country in any year that can inform the estimate, the unweighted average for the region in which the country lies is used.

Hesin internet-kagin er ikki longur stuðlaður

Rætta uppseting: 4.2.2