Training Course on statistics on ICT in the business ….pptVIP

* Strategies for adapting (improving) Statistical frames for the collection of ICT indicators Where there is available an adequate business register: Usually, the best way of improving the adequacy of the business register is to coordinate General administrative registers (taxes, registrations, licenses, social security etc) Specific registers (such as licenses for operating telecommunications businesses) and Statistical databases (updated from economic censuses). In countries with a well-developed tax or social security system, frames generated directly from associated registers could be an option if the law allows for use of registry information for this purpose. Where there is no adequate statistical business registers, surveys on ICT use by businesses will suffer because of this lack of infrastructure. It will be necessary to consider the possibility of constructing a population frame from other sources such as Lists of live units in economic censuses or External directories such as commercial telephone listings or industry association lists. Since it is unlikely that all businesses will be included in those lists, the frame may suffer from undercoverage, thereby producing biased estimates. It is recommended in that case To compare the coverage of the frame with other (external) sources and To adjust estimates by a process of reweighting. Countries using a population frame that lacks accurate information on industry should include a supplementary question in the survey to classify the main activity of the respondent unit (using ISIC or the national classification). The same can be done to measure the size of the business. Clearly, samples will not be as efficient in this situation (as samples cannot be stratified by industry and/or size). See manual chapter 7 * Sample design Stand-alone surveys on ICT use by businesses are usually designed to collect information on a large number of topics from businesses across a wide range of industries. For cost and