e

A key research need for the voluntary sector is consistent, reliable and useful data on the finances of voluntary organisations. This paper discusses challenges which arise in collecting data from charity accounts, and describes a new process, developed by the Third Sector Research Centre and the National Council for Voluntary Organisations, for collecting data from that source. The process aims to create a shared, consistent dataset that will provide a rigorous base for future study of the voluntary sector. We devised a sample of 10,000 charities, designed to permit totals to be estimated with accuracy (for example, total statutory income) by size, region and sub-sector, while also giving some insight into proportions (for example, the proportion of organisations in receipt of statutory income). Data was then captured from annual audited accounts for charities, and entered via a web-based form. Procedures were then developed for the classification of text strings describing income sources into a limited number of categories that could then be used to analyse the data in more detail. The paper then discusses a number of data quality issues found in the process of collecting data; these relate to the source material, the development of checks to ensure the accuracy of data, and challenges entailed by the accurate classification of account data