Most search engines work this way.

• Capitalization doesn't matter -- Rhode Island and rhode island are the same search.
• Most search engines search for all the words in your search in any order -- if you type in rhode island real estate, the search engine will look for pages with the word rhode and the word island and the word real and the word estate anywhere in the page. Obviously Rhode Island will tend to be together, but the others could be anywhere and in any order.
• Frequently appearing words like "a," "of," and "the" will be dropped by most search engines -- so if you search for System of a Down, you will get pages with the words "system" and "down" anywhere and in any order.
• Quotation marks are your secret weapon, they will make most search engines search for you phrase as a single unit -- So "United States of America" or "System of a Down" will get searched as single units rather than separate words, and the "ofs" and "a" will get counted.

So, you've done your search. You have some results. Probably a lot of results. Probably a lot of pages of results. How do you find what you need? How do you know that the best page for you is not 30 pages into the results?

The first step is to think about how the search engine you are choosing ranks the results. You may have noticed this process in some of the Library's online research tools and the way they rank records of periodical articles -- the default is usually newest articles appear first, but there are other ways to arrange the article records, including alphabetically by author, alphabetically title of periodical, and (most importantly for this discussion) by relevance.

Relevance sorting is done by formulas (called "algorithms") run by the search engine. Sadly, the exact workings of these formulae are secrets of the companies that create the search engines or databases, but they all look at at least some of the following:

• Proximity -- how close the terms are. If you search system of a down, pages with the band name "System of a Down" will be higher in the rankings than pages with phrases like "on Thursday, the library's computer system crashed. While IT staff worked diligently, most library functions were down for at least three hours."
• Frequency -- a page that has the term(s) mentioned a lot will rank higher than pages that mention it only once or twice.
• Completeness -- for search engines that do not require every term to appear, pages with all of the terms will rank higher than pages with only some of them.
• Locations -- most search engines can tell where the terms appear and increase the rank of a page with the terms in the title or section headings above pages where the terms just appear in the body.
• Links -- some search engines, notably Google, examine the links to the page to increase the rank of pages that have more links from similar pages.

While you can't know exactly which of these criteria your search engine is using, the important thing is to realize that something like this is going on, and, by understanding that, you can guess what is going on. You might try similar searches in more than one search engine to find out if one ranks material you find useful higher.

Once you have selected a couple of pages to look at, the next step is evaluating the page to see how useful it is. This is pretty much the same process you would use for evaluating a book or an article, but with some extra steps. This is a critical step! It is easier to find material online than in print, but it is way easier for that material to be wrong, fraudulent, or biased, too!

Standard Evaluation Tools (print and online)

• Relevance -- is it really on your topic? If you are studying the physiological effects of steroids, for example, am ESPN page on a baseball player accused of steroid use is probably considerably less useful than a page from a medical research center listing effects on the body and linking to scholarly studies that support the information.
• Currency -- when was the page written? Does age affect the information? A page listing the population of New Orleans that was last updated in 2000 is not so much wrong as very out of date. Unlike print materials, with web pages you need to look at publication date and the last update.
• Authority -- who wrote it? As for any source, you want to look for the author's credentials, including relevant degrees and/or affiliation with an appropriate organization or institution. You also want to consider the possibility of author bias. Remember, a biased author is not bad, necessarily, but that bias needs to be balanced by other sources (not doing this is a type of Academic Dishonesty).
• Purpose -- what is the source intended to do? Does it explain a topic? Does it try to persuade you to think or act a certain way? Does it intend to entertain you?

Web-Specific Evaluation Tools

• Merchandise --if a web site is mostly trying to sell you tee shirts, you may want to look for another site.
• Hosting site -- many sites are hosted on domains owned by organizations who have a side in a debate. For example, a page hosted on the Sierra Club web site will almost certainly have a bias towards environmentalism, for example. It is very easy to edit the URL down to the domain -- the part of the URL before the first single slash (so the domain of http://web.uri.edu/library/ is http://web.uri.edu/). You can look for a link labeled "about" or "about us" or something similar and see if the hosting organization has a bias relating to the topic.