Unlike paper documents, which can be anything from a scruffily copied handout to an impressive bound tome, everything you see on the Net takes pretty much the same form. Certainly, it comes at you through the same interface - a window on a computer screen. Although it is now possible to enhance World Wide Web pages to some extent, Internet documents still have much more in common with each other than paper documents do.

 Surely it's what a document says, not what it looks like that matters? True enough. But all of us have almost subconsciously devised rough and ready guides that help us decide the trustworthiness of what we read or hear. If you saw a closely typed flier glued to a lamppost you would probably treat its message with caution. If you read substantially the same thing in a national newspaper you might be inclined to take it more seriously. This isn't fair, of course - it's rank prejudice. But given the volume of information that floods over us, prejudices of this kind make a useful filter.

 On the Internet, though, everyone can publish to the world. And if they have a gift for visual presentation they can make their work look every bit as good as the most deep-pocketed publisher can. If you have sought out information from a source you already trust in the "real world", you should be on firm ground. But one of the charms of hypertext in general, and the Web in particular, is the ease with which it takes sources you never dreamed of. And unless you keep your wits about you it is all too easy to forget where you are getting your information from.

 Take Newspage for example. It takes around 20,000 articles daily from more than 700 third-party sources and divides them into 2,500 "topics" to make them easy to browse. It doesn't generate any information itself, and does not assess the accuracy of any of the material if feeds through. Given the volume of material passing through the site it couldn't judge it all, even if it wanted to. As a result, alongside stories from news organisations like Reuters and the Chicago Tribune you can read unedited press releases fed by PR Newswire, Business Wire and several others. It's all too easy to mistake a well-written release for an unbiased (or at least less biased) news report, especially if it comes to you from the same Web site in the same visual style.

 You get a similar problem when you use Web indexes like Yahoo or Internet search engines like AltaVista. They aren't in the business of getting you the most authoritative pages dealing with your favourite topic - just those that best fit the keywords you gave it. Look on Yahoo for information about Comet Hale-Bopp and you'll find a couple of dozen sites about the Hale Bopp Controversy, nearly all of them supporting a variety of unscientific conspiracy theories about an imagined "alien spacecraft" trailing the comet. Look for information on life on Mars and alongside Planet's Science's impeccably researched Web section you will find 23 sites about the Cydonia Region, nearly all claiming there is a face on Mars's surface, and spinning wildly improbable theories to explain it. Astonishingly, many of these pseudo-science sites display awards from one of the many organisations that make it their business to dish such things out. But don't forget that with hundreds of sites popping up every day, they can't spend more than a few minutes reading each one - if it is attractive and looks vaguely plausible that may be enough to get them mentioned.

 You have probably developed your own "data filters" to help you find sources you trust on the internet, and I'd welcome hearing what they are. Meanwhile, here are a few tips of my own:

 First, always keep in mind the source of what you are reading. If you reached the site from your bookmark list, or learnt about it for a trusted source, this shouldn't be a problem. But if you found a site through a search engine or by following links, it may not be clear who created the page you are reading. 

• Quite often somewhere near the top or bottom of a page you will find a link labelled "home" or a small icon which will lead you back to the home page the main index of a Web site's material. 
• Take a look at the address of the page you are looking at. If it's something like http://www.something.com/directory/subdirectory/ you may be able to find the origin of the page by copying the address, pasting it back in and going "up a level" a few times until you find something that looks like a contents page. To go "up a level" from the address above, go to http://www.something.com/directory/. 
• Many pages also include a "feedback" link or an e-mail address which connects you to the author of a page. If the data you found is particularly useful you may wish to e-mail the author direct to get some idea of their qualifications and their sources. 
• If you still have no luck you can try using the "view source" option on your browser and searching through the resulting text for a recognisable e-mail address.

 When in doubt, stick to sources you already know and trust. There are thousands of journals, magazines, newspapers and radio and TV programmes that have been around for years and have a reputation to maintain. Accessing "old media" publications may not be the most adventurous way to use the Internet, but is still one of the best ways of guaranteeing a minimum level of quality. And even here, the Net holds some advantages over paper: you can sample from a broader range of publications (particularly American ones) than you might find in a library, and you can often search them by keyword.

 Only if these tried and tested sources don't have the information you need, should you enter less well charted territory, and when you do, keep a sharp lookout. Even when you start with links from a site you respect, with just a couple more clicks, you can find yourself in bandit country.

With both Altavista and Infoseek you can use the same syntax. To check what people have said about our site for example, you would type link:http://www.newscientist.com/ -url:http://www.newscientist.com (the -url part excludes links to our site from within our site). D.B.

Of course you then have to decide whether you trust what those people think!

 Neil Blenkiron ( neilb@vicnet.net.au ) writes, "There are, as always, other alternatives - compare the information found on a number of websites and see where the middle ground lies. Occams Razor would then suggest the closest possible ideal or 'truth'. I have dyslexia, and have recently tried to find good accurate information on my disability. Out of 50 websites I found, 15 were commercial, 10 were from people with dyslexia and the rest from academia or community organisations catering to the needs of people with dyslexia. Using the above method I found the commercial sites (the ones without the tilde '~') to be the least accurate, followed by the academic sites. The web pages published by people with dyslexia agreed not only with each other in the essential symptoms/challenges that are faced with people who have dyslexia but also with a large proportion of those sites offered by community organisations. Respected sources like academic sites either glossed over the facts or presented a stereotypical picture (eg dyslexics only have difficulties accessing textual and mathatical material)."

 Certainly any information from commercial organisations should be carefully scrutinised as they can have a hidden (or not so hidden) agenda - to sell you something, and certainly academics will find it difficult to get as close to a subject as the people actually affected by it. But anecdotal experience (like that from community groups) can also be dangerous to rely on, and merely averaging what you find on web sites might lead you to believe that (for example) UFOs exist - there are far more sites publicising them than debunking them... D.B.

 Your column discusses the practice of according a website with a tilde in its address less authority than a more established site. For many subjects the only source of authoritative or recent information is the 'lone voice in the wilderness' site, ie, the one that in pre-web days, would be the single mimeograph machine and the hand addressed envelopes. Consider, for example, the ongoing child sexual abuse Witchhunt in Wenatchee, WA (USA) wherein dozens of defendants, many of them of borderline intelligence, stand accused by one police officer who seems particularly able to elicit incriminating statements from children but refuses to use a tape recorder which would provide documentation of threats. Although there are now several 'respectable' sites that discuss the issue, for quite some time the most authoritative site was a "tilde site".

 Much of my research involves iconoclastic discoveries that contradict the generally accepted wisdom. Much of my sources are no where near what would be classified as a 'First Tier, Peer Reviewed, Scholarly Journal', indeed, most of my leads appear in what would be more aptly described as 'scientific Vanity Press' wherein the author of the article pays to have it published. Many of the items discussed are the subject of various anectdotal claims previously propounded by those with little
training and considerable vested interests.

 Although I am most assuredly willing to admit that there is a great deal of nonsense out there about UFOs or Multiple Personalities, etc. I find that often the underdog...the lone voice in the wilderness...the man who stands alone...or as you might phrase it, the tilde site is indeed the voice of reason and the voice of progress. 

Don Bradley