The Internet offers a host of high quality research materials and, unfortuately, a lot of junk. You must evaluate critically the quality of all documents that you use.
Professor Cunningham's criteria for evaluating a research source follows:
Who is the author of the item? A major player in the field? Or an unknown? Written by a researcher or a journalist? Is the author's name, institution and contact address given? Be wary if any of the three are missing.
What is the reputation of any sponsoring organization? Is there any potential bias? For example, was the article written by a member of a commercial organization which is plugging one of its products?
What is the date of authorship? Is the material obsolete or outdated? Be suspicious if no date given when the material was written (as opposed to when the WWW page was created).
Is the author aware of previous relevant literature? Does the item have an appropriate bibliography? Does the author support his or her statements with convincing evidence? Or just opinions?
Is there evidence that the item has been refereed or reviewed? Are any critiques of the work available?
The scope of a document includes the breadth and depth of both its intended and its actual coverage. Are entire subareas of the subject matter missing? Is the presentation balanced across different subareas and different prospectives?
The intented audience of the item should dictate the level of writing and detail. Is the material written for a general or scholarly audience? For a computer scientist or a biologist? As a research paper or a tutorial?
Are the details and calculations accurate? Are there errors in the mathematics or programming? Are the research procedures thoroughly documented? Are there any flaws in the author's logic?