Databases for Survey and Statistical data

Databases for Survey and Statistical Data

Statistical packages provide an excellent solution for a wide class of survey and statistical data processing requirements. These tend to have the following characteristics.

Data structure is simple, so flat files are adequate.
Data collection is short term, so investment in system development is inappropriate.
Complex statistical analysis is central to the project, so a suitable statistical package must form part of the system.
Analysis is limited to a few researchers intimately involved with the data.

In these situations we usually recommend the use of a suitable statistical package, drawing on the available skills in the client organisation.

Many large projects have different characteristics.

On-going data collection, where analysis is based on a particular subset of the data, for example, the current year, or a sub-project.
More complex data structures, with optional or repeating components.
Simple analysis, but large volumes of results.
Wide use of the data, requiring careful documentation.

For such projects it can be worthwhile to invest in a more formally structured database solution. We have experience in a number of such systems, (see Recent Projects). Explicit representation of Meta-Data is important, and we have developed an outline system for adding this to a relational database.

We specialise in systems running on PCs, and look to exploit the full range of software now available, particularly under Microsoft Windows. For larger applications we advocate the use of Client–Server technology, using specialised database systems as servers for data storage and retrieval, while maintaining ease of use with PC packages as clients for data access and analysis.

For organisations with an established Information Technology group we can help to bridge the gap in concepts and terminology between IT specialists and researchers.

Click here for more discussion about the role of relational databases in statistical systems.

Page last updated 30 March 2006.