Evaluating Web Sources

Although you should generally begin your electronic research by using e-resources available through the Harvard Library, there may be times when you will want to use Google. When you find a source online, you will need to assess whether it is a credible source.

Because web sources can be created by anyone and therefore are riskier in terms of their credibility and authority, they should always be evaluated according to the following criteria:

Who is the author of this site?

As with any source, it's important to identify the author of a website and to become familiar with the author's qualifications. Be skeptical of any web page that does not identify an author or publisher. If you are going to depend on a website as a source of information, you need to determine the author's credentials as well as the purpose and rationale for posting the site in the first place. For example, a site created to serve a particular viewpoint, or to make a monetary profit, might skew information for the author's own ends. Other sites might contain information that is incomplete or wrong. In addition to considering the author, you should also consider the publisher of the site—the place or server on which the document resides (or from which it originates). If this information isn't readily apparent, try backing up several levels (deleting from the right side of the URL). Is the web document linked to a federal agency (.gov), a non-profit site (.org), an educational institution (.edu), or a business (.com)?

Always ask whether the organization sponsoring a particular site is a known, reliable, or suitable site for the document—a research center, for example, a college or university, or a government office.

How accurate and objective is the site?

If you are going to cite information found on a website, it's important to know whether you can trust the accuracy of the facts (i.e., hard data, study results) under discussion in a particular document. First, determine if the factual information on a website can be corroborated elsewhere—through a reference to or citation of a clearly reliable source, for example. A website with data that cannot be confirmed should never be trusted, no matter how perfect it might seem for your purposes.

It's also important to understand the website's point of view or bias. How clear are its purposes? Does the language used on this site suggest a specific ideology or social or political agenda? Advocacy for a particular cause is not in itself a bad thing; when the slant of a site is deceptive or extreme, however, you should reject it as a legitimate research resource. If advertising appears on a web page, try to determine the extent to which it may be influencing informational content: Is it clear where the boundary is between the advertising and information content? Does the data seem manipulated to serve the ads, or are the ads simply used to fund the site?

Sites that have academic or educational content are often non-profit sites and generally follow certain rules. Scholarship relies upon context and usually builds on precedent, so ask these questions:

  • Do you have the sense that the author is taking a position within an ongoing and serious discussion?
  • Does the site demonstrate knowledge of related research—and does the author cite current and reliable sources?
  • If footnotes, bibliographies, and hypertext links are used, do they add authority, credibility, or depth to the argument or only seem to do so?

What is the site's currency and coverage?

Since information on the web is so easily posted, it's especially important to make sure that the sources you consult are timely. Ask these questions:

  • Is the creation date of the document (or of its most recent revision) listed?
  • Is the information up-to-date? Age is relative on the Web: certain documents are timeless—their value is determined completely by their place in the historical record, and a document that is three or four years old can still be "timely" in certain disciplines. In fields where knowledge develops rapidly (the sciences, for example) or data is expected to change (statistics, for example), currency is more critical. As always, if you have questions about whether a source is current enough for your purposes, ask your instructor or a librarian.