What is the purpose of the site? i.e., is it academic and historical? Is it educational? What level is its target? (High achool? College) Is it a travel-site? Is it encyclopedic?
How extensive is the site? Is your page part of a large multi-page site? How easily can you find the home page and basic info about who is responsible for the site?
Is there a clear list of sources, as well as background of the main author?
Is there adequate Web-site upkeep?
What is the quality of the bibliography and notation – does it seem to cite authoritative sources?
Does it have links to other sites?
Does it link to or use primary sources?
NOTE WELL that one article or page from a multi-page web-site may not have some of the info you are looking for. Often you need to go to the site home page to find info on the overall purpose of the site, authors for the site, and sometimes most importantly, who is paying for the site and what do they expect in return for your use?
Encyclopedic sources – You often cannot judge all the articles the same. When judging encyclopedic sources especially, whether those links and references are there or not should be one way to determine reliability. If the article does NOT have those links and at least a good bibliography as well as clear notes – then it is probably not even good for background information.
You should keep the following criteria in mind when making your assessment for those sources that are print books or articles
Examine aspects like index and table of contents. Are chapters subdivided? How precise in the index? How easily can you find some specific piece of info in a large book? What does the preface say – does it inform about the purpose, does it justify the book? Sometimes long scholarly articles will also have these features – address any sense of organization of the article. How does an article fit the journal in which it appears?
Does it have footnotes? Endnotes? Bibliography?
Do not settle for the fact that something is backed by a reputable institution. How does the book or article prove its validity?