Total availability of journal articles to Internet users

This paper aims to determine the outcome of the different measures taken towards open access (OA) to peer-reviewed research by measuring aggregate availability of a sample of journal articles. This sample was then used to examine the factors contributing to the availability or non-availability of types of article.

A sample of articles was taken from the Scopus database based on a search by surname. This minimised any skew that would result from a dataset based on subject or source. The results were analysed to determine availability by subject and differences in availability based on source.

Less than 30 per cent of articles are available in their year of publication, rising to nearly 40 per cent in following years. Repositories are responsible for slightly less than 50 per cent of available articles. Substantial differences exist between the practices of scholarly societies and commercial publishers.

Aggregate availability is dependent on a range of activities and current efforts need to be maintained to ensure its continuance. Moves towards OA by commercial publishers are not a major factor.

Non-affiliated researchers are able to access less than half of the peer-reviewed literature.

This study differs from similar work by looking at access from the users’ viewpoint. Rather than looking at the total amount of material to which access is provided, it looks at the proportion of high-value information which is available.