Several errors occur regularly in validation checks.

• simple code mistakes, like forgetting to close a tag > or not closing it properly
• not including ALT text for images which describe them for users with poor eyesight
• broken links either within the site or to external sites
• use of out of date code - called deprecated tags
• incomplete reference to javascript included on a page

Common errors fall into two categories: those that an author simply gets wrong and those that he may not know are wrong. In the former group would be the laziness of just not bothering to add some text to describe an image or not checking links before publishing pages. It is for the latter, though, that the validation check can be particularly useful. Many designers will utilise commercial or freely available code to add enhancements to a page - often with javascript or more complex html that he could not possibly have written himself and so has just taken as correct. The validator will come to his rescue there (although not in examining the javascript itself). Broken links are probably the most annoying thing we encounter on the web and may occur, in the case of external sites, when that site developer makes changes to it and the author will not be aware.

Keeping up-to-date with CSS coding, which is developing very rapidly, is not easy. Many designers may be using software to construct code pages that was written a few years ago - software like Adobe Dreamweaver is expensive and not everyone can afford to buy every latest version, for instance - and this software can place code in a page which has been superseded. It may still work on 95% of users' browsers but it is still going to be marked as invalid by w3c!

Read More