Guidelines for Computer Research

Best Guide to Canadian Legal Research
Copyright © 1995-2008 Catherine P. Best
Conditions and terms of use

	Introduction
	Differences between computerized research tools
	When computer research should be used
	Limitations of computer research
	Keeping down the cost of computer research
	References

Introduction

Computerized research has become an essential part of legal research. Computerized research tools are very effective if used discriminately, with knowledge of their contents, syntax, and limitations.

On most research projects, it is wise to use both print and electronic sources. You will find that each method retrieves many of the same sources, but there will always be sources that one method retrieves which the other did not. You may want to use a variety of computerized tools, ranging from unedited collections of recent cases on free Internet sites to commercial databases containing comprehensive primary source collections and value-added features.

Differences between Computerized Research Tools

There is a bewildering array of on-line commercial databases, Internet sites, and CD-ROM products to choose from, each with its own search syntax and rules. If you don't understand the search syntax for the system you are using, you are probably not using it very effectively. Try to find out the following information about the product you are using:

What information is included in the product, and how current is it?

Does the product use boolean logic? If so, what is the default when you use two words separated by a space:

	Are the terms searched as a phrase, as in Lexis and the new version of Quicklaw?
	Are they searched as though there were an OR between them, as in WestlaweCARSWELL?
	Are they searched as though there were an AND between them, as in Folioviews?

Does the product permit proximity searching, where you look for terms within a certain proximity, such as within 5 words of each other, or within the same paragraph?

Can you use parentheses in your search query?

What is the unit of text being searched? Are you searching within an entire case for your terms, or within one paragraph?

How are plurals and truncation treated?

	If you put in the singular form of a word, will it be automatically searched in both plural and singular forms?
	Are other word forms automatically searched? If not, what command can you use to force the product to search variant word endings?

Can you search on fields, such as a date field or jurisdiction field? If so, what fields are available in the different databases and what is the search syntax for conducting a field search?

Does the product search for synonyms? Even those search engines using "natural language" search syntax will not automatically include synonyms in a search query.

How are your search results ranked, if at all? Common ranking methods include statistical relevancy and reverse chronological order. However, some products use other ranking methods, such as table of contents order. The quality of relevancy ranking varies dramatically from one product to another.

How are you being charged for your use? Is it based on time spent on-line, number of documents retrieved or printed, number of questions asked, or flat fee unlimited usage?

The Search Syntax for Canadian Legal Databases table provides some of this information for the legal databases most commonly used for Canadian research, including Quicklaw, WestlaweCARSWELL, and CanLII. The Internet Legal Sites table summarises search syntax for commonly used free Internet legal sites. A summary for Internet general search engines is available on the Search Engine Features Chart by Greg Notess.

A review of these tables reveals that a user who assumes that search syntax from one database or search engine is transferable to others will achieve poor results, and in many cases the exact opposite of what he or she intended. Some of these errors can be overcome by using the search templates included with these products, or relying on natural language searching options available in some products.

Although the templates included in many of the products enable users to create basic queries without needing to memorise Boolean commands, there are pitfalls for the unwary.

	The templates do not permit the researcher to create sophisticated queries, so they are not the best way to search full text documents such as cases and articles.
	The failure to truncate a word, or to include synonyms as alternate search terms, will still result in important documents being missed.
	Those used to the luxury of automatic plurals searching must remember, when using Folioviews 3.1 and most Internet search engines, to truncate in order to search both singular and plural forms.

Canadian lawyers have electronic access to a vast amount of primary source material through commercial databases and free Internet sites. This has led to a tendency to go straight to keyword searches to find relevant cases, ignoring other types of finding tools and secondary sources which were the traditional entry points for research. The latter approach has value, because it forces the researcher to place the problem within a conceptual framework, and exposes her to the collected wisdom of others regarding the issues she is researching. Good research should embrace both the narrow fact-specific approach enabled by computer research, and the broader conceptual approach enabled by secondary sources. As well, standardized indexing and references in secondary sources can lead the researcher to relevant cases that would not have been located in a keyword search. Electronic research must still be combined with a review of the print sources for truly effective research.

When Computer Research Should be Used

The volume of case law means that effective legal research must be conducted in computerized collections of full text cases. However, there are challenges to researching abstract legal concepts in this way. It should be done in conjunction with a review of relevant secondary sources.

Some tasks are particularly suited to computer research:

	finding history and judicial consideration of cases
	searching statutes and regulations
	finding judicial consideration of statutes
	looking for cases dealing with an issue identified by a very particular term, such as "renvoi", "poison pill", or "piercing the corporate veil"
	doing searches by classification scheme in publications such as the Abridgment Digests on WestlaweCARSWELL or West's American Digest System
	doing periodical searches
	doing fact specific research, such as finding cases involving accidents suffered while horseback riding
	finding cases involving a particular expert, judge or counsel
	restricting your search to material within a certain date range, or within a particular jurisdiction or court level.

Some material can only be researched using the computer:

	you may not have access to a comprehensive library
	the material may be too current to find using print sources
	the material may be from another jurisdiction
	the material may be unreported

Limitations of Computer Research

Lack of Historical Coverage

Some full text Canadian databases lack historical coverage. Major exceptions are WestlaweCARSWELL and Quicklaw. However, even these databases are missing older reported cases. Free Internet sites such as CanLII vary in coverage from one jurisdiction to another, and have less historical coverage. Some of the commercial digest databases will give more historical depth, particularly the Canadian Abridgment Digests on WestlaweCARSWELL and the Canadian Case Summaries on Quicklaw. Consider whether you need to review additional sources for the period before the database you are using commenced coverage.

Literal Approach

When you conduct computer research, you get only what you asked for. You do not see the surrounding material that would jump out at you from a page in a book. You must therefore have a healthy suspicion index.

Think of all the various ways your problem could have been described in the abstract or full text, and the various spellings or word forms the terms could take.

When searching for cases dealing with negligence, use the truncation features of the service to search for "negligence", "negligent" and "negligently".

Think of analogous terms to describe the same concept or article, such as "vehicle", "automobile", "car".

	Think of variant spelling of words. For example, is the American spelling, the English spelling, or the Canadian spelling of a word used? Is it "labor" or "labour"? Is it "reorganisation" or "reorganization"?
	If available, use a natural language search engine that will automatically search for word variants. WestlaweCARSWELL has an excellent natural language search engine, and CanLII's search engine looks for word variants.

Typographical Errors

Some databases, particularly those without editorial oversight such as the free Internet databases, contain typographical errors. As a result, you may miss a case simply because a word was spelled wrong in a way you could not anticipate.

In some cases, you can anticipate a wrong spelling. For example, when looking for cases citing the decision of the Supreme Court of Canada in Kamloops v. Neilsen you can anticipate that "Neilsen" may have been spelled in any of four different ways: Neilsen, Nielsen, Neilson or Nielson. You can overcome this problem by searching with a wild character substituted for the potentially misspelled letters.

Statute Databases may not be Current

We tend to assume that computer databases will be more current than their paper counterparts. In the case of some statute databases this is not true. Be especially sure to check the scope note for statutes databases.

Keeping Down the Cost of Computer Research

Used by an experienced researcher with an appreciation of its strengths and weaknesses, computer research can be a very effective and efficient research tool. However, in order to prevent the cost of on-line research in commercial databases using time-based or transactional charges from becoming prohibitive, several rules must be adhered to:

	Determine the databases you want to search in and construct your search query before you sign on. This is important for databases that assign a fee to each search conducted, and for time-based charges.
	Use the computer in conjunction with your print-based research, rather than as a replacement for it. For example, review secondary sources and leading cases before going on-line.
	Find out what is available free of charge on Internet sites such as CanLII, and use this resource in conjunction with the commercial databases to make your research more cost-efficient.
	Find out whether there is a way to "stop the clock" in the course of a search to permit you to rephrase your query or deal with an interruption.
	Approach subject searches with caution where your search terms are very general.
	Understand how you are being charged so that you can take a cost-effective approach.

Because most law firms are on flat-rate accounts they do not suffer the immediate consequences of inefficient computer research. However, usage history is a major factor in negotiating renewal rates for these accounts. As well, clients are often billed based on actual usage rather than the flat-rate fee. It is therefore worthwhile to ensure that these expensive resources are used both effectively and efficiently.

References

Best, Everything old is new again: the proliferation of case law and whether there is a remedy.

Botluk, Strategies for Online Legal Research.

Law Library Resource Exchange.

MacEllven, Legal Research Handbook, 4th ed. (Toronto: Butterworths, 1998).

Peoples, "The death of the digest and the pitfalls of modern electronic research: what is the modern legal researcher to do?" (2005) 97 Law Library Journal 661.

Wren & Wren, Using Computers in Legal Research (Madison: Adams & Ambrose, 1999).

Zivanovic, Guide to Electronic Legal Research (Toronto: Butterworths, 2002).