Analysis and development of statistical methods is often supported through the use of statistical computing.
At a basic level, statistical computing packages can be used to interrogate data and produce summary and graphical information.
Statistical methods can be developed and evaluated by writing programs. Occasionally, these programs may also be used as part of statistical production systems.
It is important that statistical computing programs are fit for purpose and perform as expected. There should be an effective quality assurance process to ensure results are correct.
Programs that are used for production should also include code to identify and notify the user of any unexpected errors. It is also important that these programs consider how they fit in with the wider production system.
Traditionally, statistical offices have made use of licensed statistical computing software such as SAS, SPSS and Stata. More recently, there is a move to investigate open source packages like R, which can offer a cost-effective and flexible alternative. When choosing statistical programming software it is worth considering whether open source can provide a suitable solution.
Training & Support Services
Support is available via ONS’s Methodology Advisory Service (MAS). Work is undertaken on an individual project basis with services charged on a cost recovery basis. For more information, please contact MAS.
The internet contains a wide range of useful advice and programming tips for statistical computing languages. Typing specific problems in search engines generally leads to the most useful results. See official SAS support and detailed information on R.