The Scottish Government

« Previous | Contents | Next »

Listen

CHAPTER THREE: TECHNICAL REVIEW OF THE PILOT WEBSITE

Introduction

3.1 This chapter provides a technical review of the pilot website which looks at the usage, drawing on web usage statistics. The chapter is based on our analysis of the web statistics provided by the International Teledemocracy Centre ( ITC) at Napier University as well as a technical review of the website by ECOTEC's web design expert that makes a comparison with other similar websites. Where appropriate, findings from depth interviews with stakeholders were also included.

Usage of the pilot website

3.2 The web statistics supplied for analysis covered two main phases of the website's life. The first encompassed the period 11 ^th November 2004 - 30 ^th March 2005 and the second period was from Sept 2005 - Aug 2006 (inclusive). A limitation of our analysis was that raw log files were not disclosed by the ITC on the grounds that they contain data confidential to community councils. ECOTEC was therefore not able to verify the accuracy of the generated reports. Since the site does not track unique visitors over time it is not possible to obtain a detailed analysis to correlate visitors' navigation through the site with their comments to it.

3.3 More fundamentally, the pilot website did not employ any form of identity/authentication checking ¹⁶. During the in-depth interviews, stakeholders suggested that it was decided not to use authentication (such as a registration process) as they were concerned this would put people off from using the website (particularly for less experienced internet users) and were keen to keep the website open to people who "lived, worked or travelled through" the area rather than people who necessarily had a postcode there. We accept that this is a difficult area, but given the recent experiences of participatory websites such as Wikipedia ¹⁷ (where there was deliberate injection of untruths and bias into entries) there are important warning signs that it is too easy for disruptive influences to distort the dialogues on such websites. Clearly there are advantages and disadvantages to taking a registration-based approach, although adopting such an approach need not be burdensome or act as a disincentive to genuine users. At the very least, we would expect a project funded at this level to more effectively address the issues of authenticating participants. Given these limitations, it was difficult to know how useful any conclusion based on these statistics could be.

3.4 From the web statistics provided, given the above limitations, we can infer that there was a growing number of 'hits' on the website. We can also say that seven in ten (70%) of all hits were from the UK. The statistics provided do not drill down deeper to provide data on Scotland or the areas in question; therefore we cannot say for sure that all the website users were people with a genuine interest in the area. The pilot website received referrals from a number of places, including Google and a small amount of users have signed up to access the website's RSS (Really Simple Syndication) feed via Bloglines. It may also be worth noting that whilst the amount of users increased over time, the amount of time visitors spent on the site decreased. This suggests users were becoming more familiar with the site.

Website comparison

3.5 We selected Brighton 'Local Issues' ¹⁸ as a comparison website for the pilot project in terms of design and usability. This comparison was made on the basis of being a similar type of community engagement website, and although operating in a different socio-demographic context, ¹⁹ it did have similar aims to the pilot website. The Brighton website was also set up to discuss local issues.

3.6 On the Brighton website, the technology differs slightly in that it was based around a busy online forum of over 200 active members. This was in contrast to the slightly more formalised 'blog' style of the pilot website which uses more formal software. As with all successful community focussed websites, the key to success in our experience is for administrators to create a real sense of community. This was reflected in the fact that members of the Brighton forums have themselves created a 'pledge' scheme whereby each current member can sign up to pledge to enrol four other people to the forum. Admittedly, the Brighton website was based in an area of greater population compared with the pilot websites, however the functionality of the website and its emphasis on the local area all contribute to its popularity.

Encouraging community

3.7 The pilot website was split up into a number of distinct geographical areas, some of which seem dormant and some of which remained reasonably busy after the end of the funding period.

3.8 From our technical review of the website, we conclude that the pilot website was not working hard enough in its effort to encourage the public to participate, through look and feel for example. There did not appear to be enough explicit encouragement for the public to participate. In contrast on the Brighton comparison website there were a series of paragraphs outlining the benefits of participation and attempting to foster a sense of community. Numerous methods were in place for encouraging participation, ranging from online pledges to the ability to apply for and administrate 'satellite' forums around specific subjects.

3.9 Overall, the pilot website was too passive in both tone and approach. Users will not participate unless they can understand why it is in their interest to do so and what will happen as a result. There was no visible effort to reach out to potential users and the website seemed anonymous and remote as a result. Depth interviews with stakeholders backed up this point, they conceded that the public were not engaged in the design and set up of the website, and that not enough was done to advertise the pilot website in local areas. They recognised that engagement with the public would need to be improved in future.

Accessibility

3.10 It is the legal responsibility for all governmental websites to adhere to WCAG 1.0 ' AA' (Priority 2) standards ²⁰. In the case of the pilot project, we have assumed that the websites could be included in this definition since they are government funded. It was outside the scope of this evaluation to fully identify every accessibility issue but it was clear that there were numerous errors in this respect which need to be addressed.

3.11 Some examples of this included:

• Failure to associate labels with controls ²¹ - this was a troubling omission as labels are used extensively by screen readers (a screen reader is software used by people with a visual disability) to allow ease of form-based navigation. Without these controls it can be very difficult for a user with a visual disability to know where they are within the website's structure.
• Failure to identify the targets of links ²² - failing to identify link targets was another major omission. A link target can be identified in numerous ways but the failsafe position is the use of a 'title' attribute in the anchor structure to tell a user when they hover the mouse over the link what the destination is as illustrated in figure 3.1 below.

Figure 3.1 Example of proper use of title attributes

• From our analysis of the code used in the website, we believe that the coders used a very old and low-level method which does not meet current accepted standards. As it stands, the website did not pass basic minimum standard tests for validation. The standard utilised was outmoded (for more details on the technical aspects of this testing see the footnotes). ²³ Note that the Local Channel and Your Community websites also fail W3C validation tests against the AA Priority 2 Standard.

3.12 These are non-trivial issues that may place the pilot website in breach of the provisions of the Disability Discrimination Act ²⁴, meaning that the website may not be accessible to disabled users. It is strongly recommended that the administration team perform an in-depth analysis of the source code of the website.

Usability and design

3.13 The pilot website was fairly usable and functions well in a range of common browsers and platforms. There was a good use of colour to aid usability and a good use of white space to provide room and space. The website contained a 'public' area which included the discussion threads, surveys and general information. There was also a 'private' area which only registered administrators can log into. This area included a private discussion area as well as administrative functions.

3.14 There were a few puzzling anomalies such as non-standard document icons for MS Word and PDF downloads and the lack of clarity regarding RSS feeds. RSS (Really Simple Syndication) feeds are different ways of packaging the same content. Every time someone makes a new post, this post is both displayed in the website itself and is also added to the websites' RSS feed. This feed can be accessed from a huge variety of remote websites and applications. For example, you can read RSS feeds in some web browsers or on dedicated websites such as Rojo.com where a user joins up and adds all their favourite feeds to their own area which they can then log into and read their feeds in one central location. This process was not made good use of on the website even though it can be a good way to get users to access your content.

3.15 There were also a number of usability errors committed by the pilot website administrators, rather than website design/developers. For example, as one post had received over 100 comments, the administrator started a new post to continue the topic. This is bad practice and leads to confusion since it will not be easy or obvious for people new to the debate to participate. Given that many of the administrators were councillors with limited web experience, it might be helpful if a good practice guide on running a community website is created and shared with all administrators.

3.16 The pilot website also contains a 'private' area for registered users (the councillors). The user interface in this section was not well laid out and was not clear in terms of process or operation. Simply presenting a long list of previous entries is not recommended and no attempt was made to order or group things like posts or comments relating to a post logically.

Technology used in the website design

3.17 The technology being used by the pilot website was a standard 'weblog' model under which individuals post entries and then others post comments in response to those entries. The solution was coded in what appeared to be classic ASP (Active Server Pages - this is a programming language). As the website was basically a blog model, our assessment is based on what was apparent from a user's perspective. This will allow us to suggest where the shortcomings in the application are. The Content Management System ( CMS) provided, did not give councillors the facility to add new users, so this has to be done by the ITC.

Tagging

3.18 The ability to 'tag' entries is intrinsic to all major blogging tools. A tag is a user defined word, or series of words, that can be used to describe the article in question.

Figure 3.2 Example of proper use of tagging

Source: ECOTEC

3.19 In figure 3.2, the author has written a post about serving XHTML and has tagged the entry with numerous tags that describe the differing subjects or concepts the entry will discuss. This provides two main advantages. The first advantage is for the end user. As each tag is also a link, they can click on a tag and view all entries made that use that particular tag. The second advantage is for the website itself whereby the website can link to blog search engines. These are search engines specifically setup to cater to blogs. Every time a blog author publishes a new post, details of that post - including the associated tags - are submitted to these search engines. So by submitting these tags, each blog entry raises its visibility and hence its 'find-ability'. This should promote traffic to the website. (The next section discusses dedicated blogging search engines in greater detail).

Blog Search Engines

3.20 There are numerous search engines dedicated solely to searching blogs. The three biggest are: Google ²⁵, Technorati ²⁶ and IceRocket ²⁷. Whenever a blog author publishes a new post, details of that post are automatically (in most common blogging applications) sent to a service called 'ping-o-matic'. This updates all the dedicated blog search engines to let them know your website was updated and what the changes were. This raises both visibility and website find-ability and means people searching for subject matter related to your blog can find you much more easily.

Blog Spam

3.21 It is common that blogs are plagued with blog spam - or unsolicited messages on unrelated topics. This, relatively recent, phenomenon comes in the form of automated scripts attempting to post comments to blog entries to promote their products. Unfortunately, the pilot website has not escaped the attentions of spammers as figure 3.3 below shows.

Figure 3.3 Examples of blog spam from pilot website²⁸

Source: e-community council website

3.22 Unfortunately, by leaving these comments in place, it gave the impression that the website was abandoned and/or non-functional. It was likely that the software employed in the pilot website did not deal automatically with this sort of invasion. This reflects the fact that blog spam has, as we understand it, first affected the pilot sites some time after the end of the project. However most of the 'open source' blog software applications (which are freely available, and free to access) have in-built or plug-in based solutions for dealing with spam entries which would deal with this automatically.

Additional functionality

3.23 In the depth interviews stakeholders also expressed some frustrations with the usability and design of the pilot website, as follows:

• Unable to use underline or bold in posts
• Unable to use colour or different fonts and text sizes in posts
• Formatting in documents uploaded as posts was lost ( e.g. paragraphs and indents)
• Uploading to the website was slow for documents (particularly when using dial-up internet) and can be unstable
• Cannot re-locate or edit files once they have been uploaded (for example an outdated Mission Statement was still on the website)
• Cannot post files to all six areas of the website at once (for example it would be helpful to post the same survey on all six areas of the website)
• Website can be hard to use - particularly for new users of the 'private area'

The website developers, the ITC, regard all of these points as contentious and do make reference in their final evaluation report to recommended improvements to functionality. These may be found in the section entitled "An Upgrade Path for e-Community Council Tools", and appear to identify significant weaknesses in the public area (for example re-organise the homepage and present it more clearly, re-name some main functions, better navigation to archive pages, and an improved search mechanism); and identified five areas for improvement in the community councillor pages. Taken together with the stakeholder views referred to above, this appears to indicate a product in need of further development. It is not clear why these significant issues were not addressed during the research and development phase of the website.

3.24 Stakeholders felt that some additional functionality to improve these aspects would give the website a more professional look and feel and make it easier to use.

Specifically, they felt it would be helpful to add the following functions:

• The Content Management System ( CMS) to allow councillors to add new users and issue passwords ²⁹
• A searchable archive / filing cabinet for all older content
• The facility to interact between all six areas of the website at once - for example it would be helpful to upload content to all Community Council ( CC) areas of the website at once.

Stakeholders said they would like to see these issues addressed in any future re-design of the website. Being able to interact between all six areas of the website in particular will be vital for a 'networked' model for CCs.

Aesthetics and branding

3.23 Branding is often a poorly understood design element within websites. Branding should provide an identity and meaningful context to a user whilst they are on the website. It is commonly defined as :

"The process by which the true character and purpose of the company or organization is communicated."³⁰

3.24 There was no distinctive branding on the pilot website and this was probably hindering the process (both formal and informal) of building a community. Without any local branding (as the RLD final evaluation report acknowledges), the website did not convey enough of a local identity, making it challenging to enthuse potential users or convey to them a positive message about becoming part of the democratic process at local level..

3.25 In depth interviews stakeholders also agreed the websites needed more local individuality such as local photos and information to help create a local identity (or brand). They would also have liked the opportunity to personalise the home pages for each local community to make them more meaningful locally.

Summary

3.26 The pilot website stands at a crossroads in terms of its ongoing viability. There was a very small group of core users who were high level users of the website. Web stats seemed to indicate some growth in users, but it was not possible to be firm about this due to the ambiguous nature of the statistics provided and the lack of access to raw server log files.

3.27 Other websites of a similar nature have better technical solutions and (more importantly) an apparent willingness from website administrators to work hard at building a user group where members want to participate.

3.28 The website was fairly usable but administrators did not make full use of standard web conventions on how to handle online debates, so as not to prevent barriers to participation.

3.29 Some additional functionality would improve users' experiences of the website such as an administrative area so that councillors can manage the website themselves and the ability to use formatting, bold and colours. Shortcomings in the website design, functionality and aesthetics (many acknowledged by ITC themselves) are significant, indicating a considerable amount of work still to be done.

3.30 Improved branding to provide a stronger local identity for each website may encourage uptake.

3.31 Technically, the website was not accessible for disabled users and may not meet legal requirements relating to this important public-service principle.

« Previous | Contents | Next »