Guide to Research and Writing for the Academic Study of Religion

Search engines use a keyword search similar to that of databases. The more advanced search capabilities, such as Boolean operators, phrase searching and proximity operators, differ from engine to engine. It is worthwhile to become familiar with the advanced search capabilities of a few different search engines and to familiarize yourself with how they select and rank the Web pages they present to you.

In addition to Google Scholar there are a number of other search engines including;  Yahoo (www.yahoo.com), Ask.com (http://www.ask.com), Hot Bot(http://www.hotbot.com) and Bing (www.bing.com),  Carrot2, Giagablast and Microsoft Academic. To search any of these search engines, enter your keyword(s) in the text-box and hit the search button. (Search engines often label their search buttons differently and can be called anything from go, to find, to search).

A search engine creates its own database of Web pages using what are called spiders (among other names). Spiders update the database by following new links on known Web pages and by checking old web pages for changes. New and changed Web pages are then indexed by text, links and other characteristics such as graphics. If you developed a Web page that was not linked to in an already known Web page, then it might not be found by spiders. You would have to submit it manually to the search engine so it could be indexed. When you conduct a search, the engine matches your keywords to the copies of the pages it has indexed and then provides you with a link to the live Web page. Because search engines cannot keep up with the constant change on the Web, sometimes a link will send you to a Web page that no longer exists (a dead link) or to a Web page that no longer has any relevance to your keyword.

A few important characteristics to remember about search engines:

• Automation
  The system for finding Web pages and indexing them is automated with little human intervention and no quality control. This means your access to information is immense but you evaluate for yourself the data you receive!
• Coverage
  Search engines cover different amounts of the Web and different sites. Even the visible Web is impossible to cover completely, though some engines claim to cover more than 90%. What each search engine has indexed will be different so do not limit yourself to one search engine.
• Updating
  Search engines have different schedules for checking the Web. The regularity with which the Web is rechecked and the search engine’s database is updated will affect the Web pages that are retrieved. If most links retrieved are broken (do not go anywhere), then try another search engine.
•  

Presentation of Results

Not only do the search methods of search engines vary but their criteria for ranking the results differ as well.

• Popularity
  Some search engines rank the results based on how popular the link has been to past users. (e.g. Google ranks by the number of times a page has been linked to and the importance of the site from which the link is being made. Hotbot ranks pages by the number of times a page has been selected AND the length of time spent on the page). This method is useful to see which links are most commonly used, but it is important to keep in mind that some sites may be useful to others for reasons that are not useful for you. For instance, if you are researching modern wicca you may end up with more sites containing do-it-yourself spells than sites with any academic relevancy.
• Paid Inclusion
  Some search engines accept paid inclusion sites. This means that the site pays a fee to have their site included whenever certain search terms are used. This necessarily calls into question the results that are retrieved. It is unlikely that a site that is paying for inclusion is presenting objective data or has no vested interest in you seeing their Web page.
• Search Term
  Some search engines rank their results primarily by the location and frequency of your search term in the page. Often your search will return results where your term is not used in the context you were looking for and with no quality control. A search for Tibetan Buddhism could lead you to the personal page of someone obsessed with Tibetan Buddhism but who is little informed, a page of a Tibetan Buddhist organization, a university professor's web site for a class on Tibetan Buddhism or the Wal-Mart site selling a CD of Tibetan Buddhist chanting.
• Categorization of Results
  Some search engines provide what is called a concept cluster. This means that results are grouped by topic.   Yippy (http://www.yippy.com) and Exalead are  examples of  concept clustering search engines.  WolframAlpha ( http://www.wolframalpha.com/) compiles facts and figures in a very informative and easy to read display.

TIP - Remember to always evaluate the sources you have found both for relevancy and credibility!

Search engines also have different search term capabilities though most share some basic common ones. To search effectively it is important to understand what types of search your engine allows and how they do it.

• Boolean
  Though not all search engines allow Boolean searches, they all have a Boolean default. This means that when you enter more than one search term it assumes a Boolean operator between them. Figuring out the default for each search engine can make interpreting your results much easier. The Boolean default is usually AND but not always. Most search engines have some type of Boolean capabilities but not all use the same language or symbols. Some engines use ‘and’ while some use + or AND. Again, figuring out the characteristics particular to your search engine will greatly improve your search ability.
• Truncation
  Many search engines will allow truncation and some such as Google provide some automatic plural searching. The most common indication of truncation is the asterisk (*), though again, this is not always the case.
• Stemming
  Some search engines truncate automatically. This is called stemming or auto truncation. If your search engine auto truncates, then the search term swim would also initiate a search for swims and swimming. If you don't want the engine to truncate then often forcing a phrase search (e.g. "Tibetan Buddhism") will prevent variations on your keyword from being included in the search. In Google you use a plus sign in front of the word to prevent stemming.
• Phrase searching
  Most search engines will recognize phrase searching by placing quotations around the phrase - i.e. "Catholic Church". Some however recognize phrases differently by using parentheses or other methods.
• Field tags
  Specify where your term must occur (e.g. title, URL, etc.)
• Domain type
  Some search engines can limit your search based on the domain type. This means you can specify an American educational Web site with .edu, a governmental Web site with .gov, and a Canadian Web site with .ca. A few domain endings and what they mean are listed in the table below.

Meta-search engines, such as Webcrawler (www.webcrawler.com), and Startpage.com (formerly Ixquick - http://www.startpage.com) function by searching the existing databases of other search engines and sometimes directories as well. They do not have their own databases or indexes. Some return individual lists for each search engine while most return a compiled list often eliminating duplicates.   LightSwitch, allows you to quickly select  to run your search in a variety of search engines and databases.  Meta-search engines can be a great time saver by covering many search engines with one search. However, they are not always the great solution they may appear to be.

One of the drawbacks of meta-search engines is that you have less control over your search. Many cannot handle complex search features as each search engine to which they send your query has different search languages and features. You often cannot tell how your search terms were queried in the other search engines. Furthermore, some of the best search engines prevent meta-search engines from querying their databases. The search engines that they can query are sometimes overloaded already or inaccessible and so the meta-search engines exclude them from their search and gather only the information that is most readily available. Even if you can't specify the search engines to be searched by the meta-search engine, check the help screens to see which search engines are being used.

Further, meta-search engines only retrieve the most relevant hits from each site, usually the top ten percent. This means that your results are not comprehensive. The focus on relevancy can of course work in your favour by cutting down on sites that have minimal relevance to your query. However, the meta-search engines use the relevancy standards of the search engines they query so if the search engine ranks paid inclusion sites first then your results will contain this bias.

Despite all the bad news,  you may find meta-search engines useful when you want to explore more broadly. Note if you create an account on Google, you can use Google to create a customized search engine.