Wildcards II

A student may ask us:

Are there any other examples of wildcard searches?

Our reply may be:

If we are researching a thesis or academic research project in environmental science, we may save time instead of searching separately for the words pollute, pollutes, pollution, polluter, polluted, polluting, pollutant, and pollutants, by simply searching for pollut$.

The search results should find all of these variants for us in a single effort. Another example is if a TU student researching the subject of secondary school education needs to find out information about teen mentors. Instead of searching for teens and teenagers and mentor and mentors and mentoring and mentorship, it is possible to search for teen* AND mentor*.

Similarly, if a TU student is doing research on feminism, instead of doing separate searches for the words feminist, feminism, feminists, and feminine, they may choose to search the single term femini* and get the same results.

A search for the term politic* should bring up the results that a search engine would usually produce for separate searches for the words politics, political, and politicians.

The student may wish to know:

Is there any time that we should not use wildcards?

Our answer would be:

Some research databases provide their own wildcard searches, looking for alternate spellings of search terms. In those cases, it is best not to add any wildcards.

The student may ask:

How can I tell whether the database I am using has its own wildcard searches?

We may say:

To be sure, first try any search without wildcards, and if the result is not extensive enough, with enough alternate terms included, then it should be clear that the database has not looked for variant terms. In this case, then it may be useful to try wildcards as a supplement to the original search. 

While it usually takes more time than more researchers want to spend to read the instructions on databases, some may also be helpful on this subject. For example, the website of PubMed, a free search engine accessing primarily the MEDLINE database of references and abstracts on life sciences and biomedical topics. The United States National Library of Medicine (NLM) at the National Institutes of Health maintains the database as part of the Entrez system of information retrieval. As a learning resources feature, part of PubMed Online Training, a webpage informs researchers:

• Truncation, or finding all terms that begin with a given string of text, is generally not a recommended search technique for PubMed. Truncation bypasses Automatic Term Mapping and automatic explosion. Also, in such a large database, variations of a text string can become overwhelming and unmanageable. Especially when multiple terms are truncated, PubMed may even time out and fail in an attempt to run the search. However, there are times when limited use of truncation can be useful.
• The PubMed truncation symbol is the asterisk (*), sometimes referred to as a “wildcard.”
• For example, let’s say you are trying to search all terms that have the root, mimic. If you search mimic* PubMed will retrieve words such as mimic, mimics, mimicking, etc.
• PubMed supports only end-truncation. There is no single character symbol.
• PubMed restricts retrieval to the first 600 variations of the truncated term. When this occurs, PubMed will display a warning message.

This means that using the asterisk to shorten a search term is not recommended for PubMed because too many search results appear and the researcher may be lost in all the material that is available on different subjects related to original search term. PubMed uses the database search term truncation, which just means shortening, as when a word ending is replaced by a symbol such as the asterisk (*) or question mark (?).

(All images courtesy of Wikimedia Commons)