Reading and Evaluating Web Sites

Prof. Chris Toulouse
Department of Sociology
Brooklyn College

Print and the web again Book design is easy. In part because book design is limited by physical form, and in part because it's had centuries to evolve, printed volumes follow a familiar linear convention: cover, title, publishing information, table of contents, introduction, chapters 1-X, conclusion, appendix, references etc. There are similar models for the essay, the newspaper article, the magazine piece and so on. Web design is difficult. In part because there are so many nonlinear possibilities, and in part because people are only just getting used to the differences between the fixed world of text in print (when projects were products and they were always finished) and the fluid world of text on the web (where projects are processes and they are never ever finished). Nevertheless, it's possible to discern some basic patterns for organizing web sites and navigating readers around them.

Basic design types The circle: the simplest web site design, picture-based sites often work this way, with a first page leading to a second to a third and so on. The chances of the reader getting lost are small. You can only go backwards or forwards. The wheel: the form resource pages should take, with the home page at the hub with spokes linking to subsidiary pages. If there is always a clearly marked link back to the home page the chances of the reader getting lost are small. The pyramid: a more complex version of the wheel, where subsidiary pages are themselves home pages to sections of web sites. Without clear navigation tools the chances of the reader getting lost are fairly high. "The camel": my term, from the colloquial phrase "a camel is a horse designed by a committee." Company sites are often camels because different departments are responsible for different sections and they do not coordinate properly with one another (and because there often seems to be hump of information somewhere and it's not always clear why it's there...)

Basic navigation tools The navigation bar: the simplest navigation tool is a horizontal table with cells for the home page and each of the pages on a web site. If the page author establishes the convention of putting a navigation bar at the top and/or bottom of each page, readers should not get lost. The navigation frame: the same information you find in a navigation bar, but this time it's placed in a frame that always appears at the top, right or bottom of the browser window. This is more convenient to the reader in so far as the contents of the site is always within view; however, frames slow downloading and can impose an idiosyncratic design that can confuse readers and cause them to crash their browsers. The site map or site index: a client-side image map or a multi-celled table or a long list displaying all the pages on a web site. Invaluable in helping readers find their way around pyramids. The search organized site: minimal navigation information on the home page is accompanied by a search facility which calls web pages out of a cgi-bin. Huge institutional sites like newspapers and large corporations use this kind of navigation aide. It can be helpful but it comes with two drawbacks: (1) readers must express their inquiries in terms that are on the keyword list inside the search engine or they will get the ubiquitous "no items found" message; and (2) cgi-bins generate long unwieldy web addresses that very often do not work from outside their host web site.

There is often a lot of useful information about the source
in the address of a web site.

What type of account is it? Private accounts (directories on an organizational server) have a tilde ~ in the address. Private accounts generally cost $20 a month in the US. Commerce is prohibited and most private accounts do not have server script facilities. Domain accounts have common suffixes: com for commercial org for nonprofit edu for education (ac in the UK) gov for the federal government mil for military Domain accounts generally cost $50 a month in the US. Commerce is permitted and most domain accounts run server scripts to provide page counters, facilities for processing forms, message boards etc. Addresses on servers in other nations often carry a national marker -eg. uk for Britain, au for Australia, fr for France etc.

Hey, where am I? Lopping of sections of the address: The most important navigational aide in the hands of the web site reader is to delete sections of the address /between forward slashes/ and hit return. That way you travel further and further up in the hierarchy toward the host directory. The hidden directory trick: By deleting sections of the address and hitting return you may come upon a hidden directory. Here's why... The home page of a directory is usually called index.html. You can get away with typing an address without including the index.html -for eg. typing in www.urbsoc.org gets you right to the index.html page. However, what happens when you leave the index.html out of the address but there is no index.html in the directory? Then Netscape displays a list of the files in that directory instead. Try www.urbsoc.org/courses and take a look. The general attitude to this feature among page authors seems to be that it is messy and unprofessional. However, it is also a godsend to a reader trying to fathom the organization of a web site.

Remember (in spite of what college professors and publishers would like you to believe) you don't need a Ph.D and books to your name to produce a very useful web site.