Affiliations: Faculty of Economics and Management, University of Latvia, Riga, Latvia
Note:  Corresponding author: Daina Šķiltere, Faculty of
Economics and Management, University of Latvia, Central Statistical Bureau of
Latvia, Riga, Latvia. E-mail: [email protected]
Abstract: National Statistical Offices (NSO's) need data for the production of
statistics. Apart from data obtained through surveys, NSO's are increasingly
using other data sources (for example, data collected by other organizations).
Disadvantages here come from the fact that very often an NSO uses an
administrative data source for a purpose different than the one for which the
data was originally collected. As a result of this difference, the 'statistical'
usability of a data source needs to be thoroughly studied by an NSO before its use. Some more possible problems of statistics are the
following: data sources, recent data, regularity, timeliness of data, updating
data, time series, data frequency (annual, quarterly, monthly or daily data
required), data comparability (is it possible to compare one country's
performance against another or group of countries etc.), data costs. Sometimes,
the required data do not exist; data from different sources are not always
comparable. Also, the quality of some information needs an improvement. It is therefore of vital importance that a systematic approach is available to
determine the quality of statistical data. For this purpose a quality framework
have been developed by authors to determine whether statistical data is of
quality required for their intended use.
Keywords: Statistical data quality, quality evaluation