I know a lot of folks avoid Twitter because it be kind of a mess / useless timesink / brain drain / dumpster fire. And sometimes it can. But it also can lead you to a lot of interesting people, like Spencer Greenhalgh.  Spencer is a PhD candidate at Michigan State University who came to my attention in December 2015 because of his work with R and Twitter. And we’ve had some conversations and he’s pointed me toward some great resources.

And he also gave me a terrific question recently – a question about Google Scholar.

Spencer was having trouble searching Google Scholar ( https://scholar.google.com/ ) for his topic of choice – Twitter and graduate students / programs.

I love a good search question. So I jumped into Google Scholar and started messing around, and wasn’t having much luck. Then I thought, “Wonder if Google’s full-word wildcard works in Google Scholar?” Surprise! It does.

A Quick Look at Google’s Full-Word Wildcard

Google uses * as a full-word wildcard. That is, if you put a * in a phrase, you will get phrases that have any word in place of the * in the phrase. My stock example for using this is searching Google for “three * mice”:

You’ll see when running this search that “three blind mice” matched my search, but also “three little mice,” “three bad mice,” etc.

I don’t have the room in this article to get into an exploration of the potential of the full-word wildcard, but I do want to add that if you use more than one * you will get different results. You see the results for “three * mice” above, but what happens when you search for “three * * mice”?

You’re still getting “three blind mice” but you’re also getting phrases like “three strains of mice”. Using two wildcards opened up  your phrase searching a bit.

Bottom line: you don’t have to use just one. It will make a tremendous amount of difference if you use two wildcards or more, though in my experience you don’t get much different after about six wildcards.

(I’m working on a screencast to show how to use the wildcard, and I’ll get into more detail there.)

What does this have to do with Google Scholar? The full-word wildcard is useful there for doing proximity searching.

Spencer’s Problem: Graduate Students and Twitter

Spencer was searching Google Scholar for information on Twitter use in graduate programs, and he wasn’t having much luck.  Twitter and graduate programs, as you might imagine, are common words on Google Scholar, and those two concepts don’t easily translate into a phrase.

So let’s make a fake phrase with Google’s full-word wildcard. Instead of trying to figure out how people might describe information about Twitter as it relates to graduate programs and students, let’s just fake it with this search:

“Twitter * graduate programs”

Hmm. That’s not doing me a lot of good. How about two wildcards?

That’s doing me even less good. How about we focus on graduate students instead of graduate programs.

Aha! Now we’re getting somewhere. You’ll see in the screen shot that the query matched to “Twitter-using graduate students,” “Twitter with graduate students,” and “Twitter to graduate students.” If we’d resorted to testing phrases, we probably would have gotten around to these eventually, but isn’t it easier to try it this way?

Let’s add another wildcard and make it “Twitter * * graduate students”.

Another set of useful results, very different from the first set. But we’re getting enough relevant results to know we’re on the right track.

Now what? Do we add another wildcard? No; we go backwards by leaving a single wildcard in place and reversing the query: “graduate students * twitter” .

That worked as well! Now let’s try for the grand finale, “graduate students * * Twitter”

A little overlap here, but still plenty of useful results.

The point is that full-word wildcards are not just for Google searching a known phrase, though it can be useful for that. Instead it’s invaluable as a proximity search tool when you’re trying to explore two concepts that can’t easily be linked as or with a common phrase. And as you can see, that’s especially useful in Google Scholar.