Ask Yourself These Four Questions About Any Digital Collection
Whenever you set out to use any electronic research source, be it a public web search or a specialized database, there are a few questions you should always ask – four to be exact. You may say there are really five or six important questions, or maybe you think there are only three, and that’s ok. But for me, not three but four is the magic number.
So here are my four questions you should always ask, with examples drawn from some of our favourite databases.
1. What is the SCOPE of the database?
Or what is its reach? How would you describe the collection of documents? What topics are covered? How current, how far back in time? What is the geographic extent of the database and what languages are used?
For an Internet search engine such as Google, the scope is all documents within the reach of its automated web-crawling indexer. But for most sites, the scope is limited by subject, geography, &/or time. For some, such as our Saskatchewan Cases database, the user should have a pretty good idea, based on the title, that the scope is limited by geography and subject matter. But what about time? For that we have a scope note conveniently located in the lower-right corner of the search page. For a complex system of databases like CanLII, the scope note is myriad.
For some online services, such as our Find a Lawyer search, the scope is rigorously monitored and crucial to the database’s usefulness. There is one record for every active (practicing) member of the Law Society of Saskatchewan. It is updated instantly. If a record exists, then that person is an active, licensed lawyer in Saskatchewan. If there is no record, then that individual is retired, suspended, inactive, or otherwise not a currently practicing lawyer in the province.
2. What constitutes a RECORD?
What are the individual things the database consists of?
Some case law databases, like Saskatchewan Cases or the Canadian Abridgment, consist of summaries or digests. Others, like CanLII, consist of primary law – judgments and legislation. In others, such as the case law components of Quicklaw and WestlawNext, you can search a combination of summaries and full-text documents. Knowing if you are searching a short summary or full-text decision is very important in formulating your search.
The Saskatchewan Bills database has a unique scope. Each record constitutes a Saskatchewan statute that was affected – created, amended, or repealed – by a Saskatchewan bill. So note that although the database is called “Bills,” the individual records are not bills but statutes, as they were affected by the bills. Thus, there will be a separate record each time a statute is amended, and separate records for each statute affected by single bill. This sounds a bit complicated, but users of the database will agree that there is tremendous advantage to this structure when conducting statutory research.
3. What is the Search SYNTAX?
Know your operators – every search is different. Until you know the basic commands for a site – AND, OR, Phrase, etc. – you should consult the “search tips” or “connectors” every time. Many sites include a link to their search tips, operators, connectors, or “search help.” Some, like Saskatchewan Cases, display the operators right on the search page. In CanLII, scroll over the question marks at the end of the command lines for basic commands, and from there click on help page for more detail.
A closely-related and equally crucial question (and the fifth in the list, if you think the thumb is a finger) – what is the default operator? How will the search interpret two words beside each other in the command line without any connector between them? For our Law Society research databases, it is always a phrase – the search will look for those two words in the same order. That’s a bit unusual. Many search engines default to AND, and some to OR.
Increasingly common, though, is a type of hybrid where the search engine looks for the words (1) together in order (phrase), (2) both included but not necessarily together (AND), and (3) either word (OR) – then an algorithm sorts out the order for you, so the most relevant documents are at the top. Yes, for these types of sites maybe you can ignore everything I’ve said about operators and still get somewhat useful results. Fine. But don’t assume that’s the case – I’m talking to you, Millennials!
4. How are the results ORDERED?
Don’t assume relevance ranking. Our databases, for example, always order the results in reverse chronology (newest to oldest). These days, many databases offer ranking options (usually in a drop-down menu). So if you are looking at a case law database, think about if you want your results by relevance, date, court level or number of cites. If there are a large number of hits (and you can’t think of how to reduce them), then relevancy may be best. For a small number, level of court or number of citations may be better.
Increasingly, searching filters, such as a menu allowing you to limit to documents from only a single jurisdiction, court level, or date span, can be applied after the search – whereas traditionally they were applied beforehand. CanLII and WestlawNext are two good examples from the Canadian legal world. Filtering search results is a very useful feature – it enables you to search broadly, then gives you considerable control over how the results are displayed and what types of records are visible.
[This tip by Ken Fox originally appeared on the Law Society of Saskatchewan Library’s Legal Sourcery Blog]
Start the discussion!