Internet access: an uneven picture

• consistency of service access and quality in libraries across the UK
• rigour and consistency of application of Acceptable Use Policies (AUPs)
• clarity, consistency and visibility of policies about internet filtering
• consistency of front-ends for users.
  

The project used unobtrusive testing (‘mystery shopping’), with 14 different UK library authorities (eight English, four Scottish, two Welsh) being visited. Where possible neighbouring libraries were chosen, the hypothesis being that two libraries that were close but under a different local authority control could conceivably be visited by the same people and thus any differences in service would be noticeable to the users. Travel from central Scotland to the location had to be relatively easy and inexpensive.

The same researcher visited all 14 libraries. His story was that he was not a library member but wanted to access his email using the library computers. The researcher had no means of proving his address, only being able to show identification such as credit cards or workplace ID in order to gain access to the service.

The researcher attempted to access 25 different websites while using the library computers. The checklist of websites he tried to visit incorporated:

• chat sites
• web-based email
• web 2.0 sites such as MySpace and Faceparty and Bebo
• sexual health information sites
• advice sites on sexuality issues
• dating services
• downloading sites
• gambling sites.

The 14 libraries in the study will not be identified, since the aim is only to raise general issues of concern.

Consistency of service access
Twelve of the 14 libraries allowed the researcher access to the internet despite the fact he was not a member and had limited ID. The two libraries that refused access did so on the basis of him not carrying acceptable ID that showed his address. Interestingly one of the libraries that refused access stated they would have been happy with one non-photographic item of ID that stated address. This means that a bill or similar would have been acceptable, raising the potential issue of identity theft and whether a bill that did not belong to the individual would have been enough.

Only one pair of libraries used the same network and shared interface/AUP/filtering standards. These library services also allowed members to borrow items such as books from the neighbouring authority, offering an attractive set of reciprocal entitlements all round.

Application of AUPs
In only one of the 12 libraries where access was granted did the staff make any attempt to explain what acceptable use of the internet constituted and what the user’s responsibilities were. Indeed, in two of the 12 libraries, staff helpfully logged the researcher on to the computers, bypassing the on-screen AUP on his behalf.

Policies on internet filtering
Among the 12 sites providing access there was no consistency in internet filtering. Two libraries blocked nothing on the checklist, and others blocked varying numbers of sites in different categories. The most commonly blocked were chat sites (50 per cent), an advice site for gay teenagers (33 per cent) and a gambling site (33 per cent). None of these sites were illegal, and their availability in other library services tested raises questions about consistency of access to legitimate content across the country.

Some blocking was overt, with a clear screen being displayed saying that the site had been blocked and advising the user to contact staff if they wanted it to be unblocked. This raises some issues, since it is questionable whether a teenager would ask staff for a site on sexuality to be unblocked, given the personal nature of the content.

However, some blocking did seem to be covert: what appeared to be 404 error messages were displayed, presented in such a way as to suggest they were not standard 404 errors and thus hint that it was the filtering software blocking the site rather than the site being unavailable. It is difficult to be certain about this, however, without knowing the intricacies of the filtering software used.

Consistent front-end
One of the most disappointing aspects of the study was the huge variability of quality of the front-end screen faced by users when they first log on to the system. Only one site visited had produced a coherent information and library-orientated front-end which had the sole aim of operating as an information gateway for users. All others pointed to council website pages, or library homepages within the council website. These varied greatly in quality, ease of use, and usefulness. Indeed, the researcher found several of them confusing and with links that did not seem at all relevant to the majority of library users, such as council tax payment links, and links to councillor information.

It did seem that, for the majority of the libraries visited, the e-government goals of the local authority rather than the information needs of the library user were given priority.

Rather worryingly, only one library of the 12 offered a novice internet guide on the computer desktop, which meant any novice users would have to rely on the library staff to guide them.

Questions to ponder?
While all the libraries had staff who were extremely helpful to the researcher, even when access to the internet was denied, there were occasions when the desire to be helpful was potentially putting the library at risk. Since an AUP forms a contract between the library user and the library, the fact it was not explained to the researcher in 11 out of the 12 libraries granting internet access raises issues of potential liability. In addition, in two of the library services, staff logged the researcher on to the terminal, clicking the AUP screen to proceed. In this scenario, who exactly would be responsible for any misuse of the internet that resulted, either malicious or accidental?

Recommendations
The findings of the project suggest the following issues need to be debated:

• The need for thorough training in public libraries on public access computing. Staff members need to understand what the AUP is there for and what its limitations are, and to ensure that all users who are accessing the internet using library facilities understand what their responsibilities and liabilities are. The rights and wrongs of passing on the liability to the user notwithstanding, it seems unwise to have a policy in place and not ensure that all staff members both understand it and adhere to it. While it could be argued that the two libraries out of the 14 who refused access were being restrictive, they were adhering to the policy of the organisation and thus their actions were entirely justifiable.
• Libraries’ need to consider whether their priority is to provide an unbiased information service or merely act as an agent for the local authority. It is regrettable that some public libraries seem to believe that simply placing the local authority web page on library terminals is a good way of providing an information source for users. The variability in quality of these sites could mean that novice users are being presented with a gateway that is of no relevance to them or, worse, that could hinder their information seeking. More research needs to be undertaken on how such gateways influence the information seeking of public library users.

It is reasonable to expect that public library users throughout the UK should receive a comparable service. Access to the internet is one of the trickiest services to manage, but it does seem that resources are being wasted on a large scale, with every authority developing their own AUP, filtering policies, front-ends and the like. Surely a movement towards an acceptable national model would ensure public library users throughout the country received a comparable service, and that their quality of access to web-based information was not based on their postcode.

David McMenemy is a Lecturer in the Department of Computer and Information Sciences, University of Strathclyde (david.mcmenemy@cis.strath.ac.uk).