TIP Sheet
EVALUATING WEBSITES
One could argue that critical thinking skills are needed more than ever in the 21st century, if only on the basis of volume of information readily available. We expose ourselves to vast avalanches of unfiltered information on the Internet, where self-publishing is the rule and editorial oversight is optional. Personal websites and weblogs run the gamut from the inane to the cerebral, from the villainous to the highly principled, all protected speech under the First Amendment.
When using online research you need to use sources that are credible, that is, accurate, truthful, and appropriate. Unfortunately, many web authors are willing to sacrifice truth for technical accuracy, or, more commonly, dispense with accuracy altogether in a fit of enthusiasm. Alas, blind zeal is sometimes contagious, like the flu; you have to inoculate yourself against it temporarily in order to appraise the claims it puts forth.
This TIP Sheet discusses some ways to identify the author and his/her/their credentials, the author's purpose, and the possible bias of a website:
The URL
The "address," or URL, is an obvious first place to look for clues about a website. The three-letter extension that follows the "dot" tells you the type of site, and you might begin to infer (guess) the purpose and expertise of the site on this basis alone>
Don't jump to conclusions about a site based on its extension, however. A dot-com site may provide great, objective information (like www.groworganic.com ), while a dot-org may want to "sell" you an idea or ideology (www.cpusa.org, the U.S. Communist Party).
The domain (the part just before the "dot" and three-letter extension) may contain more clues. An organization's name or acronym may constitute the domain, for example, www.loc.gov, the site for the U.S. Library of Congress, or www.latimes.com for the Los Angeles Times newspaper.
Domain names can tip you off that a website is a personal project. Commercial servers like Geocities offer "boilerplate" website construction tools and website hosting, no questions asked. A Geocities website includes the word geocities in the domain. Personal sites may also be marked by a tilde ( ~ ), a percent sign ( % ), a personal name (ssmithers), or the words users, members, or people. The important thing to remember about personal sites is that their authors are an absolute unknown to you, and you will have to get a lot more than a name. Some personal sites are published by hobbyists who have quite a lot of expertise in a highly specialized field of interest, like members.iinet.net.au/~rmine/gctrebs.html (a website by and for trebuchet enthusiasts); others are a factual, visual, or grammatical mess.
After breaking down the URL and identifying as much as possible the authoring group or individual, ask yourself if this author is an appropriate source for this type of information. For example, you probably would not wish to cite organic gardening tips found on an American Nazi Party online discussion group; even if they were perfectly good tips, the source would not be credible. Find similar gardening tips on a credible gardening site.
The page perimeter: author, update, and contact info
Navigation bars and buttons around the page perimeter-top, bottom, sides-offer places to search for clues about the author and his/her/their qualifications. First, identify the author. If an online article does not have an author listed on a byline at the top (by Sue Smithers) or at the bottom (--Sue Smithers), look at the page perimeter for links such as About, Biography, Background, or Who We Are. The author may be a group, organization, or institution (both MLA and APA citation styles allow you to cite an institutional author if no individual author is identified). The author might be the person in whose name a copyright appears.
If you still can't find an author, try truncating the URL, that is, shortening it by gradually deleting elements at the end of the URL until you arrive at what appears to be a home page. This is especially useful if you arrived at an unattributed document through a link from search results. Search results typically link you straight to the content you searched for, bypassing the home page of the site. In this case, you probably won't know unless you look for it who authors or sponsors the information, and you must know this to begin to evaluate it as a source (or to cite it). For example, the online article "Thinking Critically About Conspiracy Theories" appears with no author named. Nevertheless, it clearly is authored by someone, complete with footnotes. To find the author, truncate the URL string www.uea.ac.uk/~j097/CONSP01.htm, omitting the filename of the article itself (in this case, CONSP01.htm). Thus, www.uea.ac.uk/~j097 brings you to a page called "Jerry Goodenough's Home Page"-Jerry Goodenough being the author of the article. His background and qualifications appear on the same page. (This domain, by the way, tells you that the hosting server is the University of East Anglia in the United Kingdom. The tilde (~) tells you this is his personal space on the university site. Goodenough is a philosophy professor at UEA.)
Of course, if the declared author is not who he/she/they claim to be, you may never know this. The best you may be able to do is ascertain whether the content itself is credible. You can do this by searching other, more credible sites, to compare facts and claims. Of course, if you have to go to these lengths to verify the credibility of a source, perhaps you'd be better off simply using the more credible source!
Look at the perimeter of a page to find a date of last update. If the author is citing statistical or other time-sensitive facts, he should either name his own recent sources within the document text, or date the page to demonstrate its currency. Some topics are not time-sensitive; however, a recent update also shows that someone somewhere is actively managing the website. This is, in general, a good thing.
If there is a website manager, he/she/they should provide a contact link to an e-mail address. If there is no functional contact information of any kind, be skeptical and find another source that is willing to supply useable contact information.
The page perimeter: author's purpose and agenda
The author's purpose may be to inform you, persuade you, entertain you, or a combination of all three. Information and entertainment might be offered as an incentive to persuade you to buy a product or a point of view. Depending on the approach, the author may or may not exhibit bias in the pursuit of a purpose. Suspect bias if page content is one-sided, making strong values claims without acknowledging opposing or alternate views. Suspect bias if an author makes startling factual claims without documenting sources. Suspect bias any time the overt purpose of a site is to promote something. This doesn't mean you will always find it, but if you do not at least look for bias, you certainly will not find it even if it is there. Now, the Mayo Clinic's website wishes to promote informed self-management of health care; it offers information on a great many health and medical issues with a high degree of expertise and no particular evidence of bias. On the other hand, Chong's Health Care, which offers articles on Chinese medicine and no doubt wishes to promote understanding of traditional and alternative medicine, also wishes to promote the purchase of an even dozen of its own herbal products. Chong's has a vested interest in selling you something, so it may be a biased source. Still, experts often do sell products and believe the products they sell are beneficial. So even if a source has a vested interest, that source may also be veracious.
Page perimeter elements can link you to statements about the overt intention and purpose of the author-About, Philosophy, What We Do, or Mission Statement, perhaps. Whether the author's stated purpose is, in fact, his actual purpose is another judgment call.
To illustrate, the URL www.ici.org tells me little beyond the fact that it represents a nonprofit group. It could be a fine, upstanding nonprofit group, or an unsavory one. A visit to the site tells me this is the site of the Investment Company Institute. The link About ICI, at the top of the page, tells me a bit more: this is an association that presents itself as promoting high ethical standards in investment companies and public understanding of investment issues. A look around the site would seem to confirm this claim. You might fairly conclude that this site would be a good source for information about sound investment practices or about regulation of investment companies.
Of course, the stated purpose of a website may not represent its actual purpose; you will have to reach a conclusion based on your own background knowledge and what you know of the author, sponsoring organization, and page content. For example, www.dhmo.org appears to be a very elaborate activist site warning against a pervasive and deadly environmental toxin. It has links to the Centers for Disease Control, the National Toxicology Program, and the Environmental Protection Agency, among others. The content is copyrighted by Tom Way and includes a notice of recent update. There is a contact link to an e-mail address. And more than 500 other web pages link in to this site (see below). But read the Dihydrogen Monoxide pages closely, think back to your high school chemistry class, and see if you can discern the author's real purpose.
Looking for links is another way to approach website evaluation. In general, more links to a site suggests more value and credibility. Websites often have lists of links they recommend; look at some of them to see if they seem credible. You can use Google to search for links to a website to get a feel for how many others find this site valuable and credible. To demonstrate, search Google for the Librarians Index to the Internet (lii.org). Then use Google's tools to search for Links to this site. You will find that more than 8,000 other sites link in. If you search for the Investment Company Institute (above), you find about 400 links. The University of East Anglia has more than 1,000 links in. For comparison, an odd conspiracy website called The Watcher Files returns only about 100 links. The value of trying to evaluate a website by searching for links to that website is limited, since really it is the quality of the linking websites themselves that counts. After all, if 8,000 wackos are linking to the Librarian's Index to the Internet, then link count is no help at all in determining the credibility of the LII. Therefore, don't use a links search without using all the other ways you know to evaluate a site.
Computers as persuasive technology
Don't think you are alone in struggling to reach sound conclusions about websites. An emerging field of study called Captology–Computers As Persuasive Technology–examines, among other things, how websites are perceived as credible and how they persuade us to change our behavior. Research conducted in part by Stanford University's Persuasive Technology Lab has shown that, no matter what elements lay people say should weigh most heavily when evaluating a website (author's expertise, sponsoring organization, timeliness, and so on), they tend to reach more or less snap judgments based on how the site looks. Professionals, on the other hand, tend to disregard looks and give more weight to factual content and the reputation and expertise of the author and sponsoring group. Moreover–and no surprise here–we are more likely to perceive a site as credible if we agree with it.
The trick is to train yourself to think like a professional. Be open minded but skeptical. Evaluate factual content. And in the words of the UC Berkeley guide to evaluating websites, "Always look for bias. Especially when you agree with something, check for bias."
For more tips from UC Berkeley on evaluating websites, see www.lib.berkeley.edu/TeachingLib/Guides/Internet/Evaluate.html. For ten tips from the Stanford Persuasive Technology Lab, see http://www.webcredibility.org/guidelines/index.html.