This is of course rather self-evident to begin with - you find things mostly online (and only very rarely through the physical library on campus), and you do so by using search terms that you type into a search engine such as Google Scholar.

That said, using search terms efficiently is a bit of an acquired skill that I will try to demonstrate to you through an example:

These days I am considering writing a paper on facebook and art education. So, what would the search terms for something like this be? "Facebook" is one - obviously. However, I can think of other related ones also - such as "social networks" and "web 2.0."
  • Writing these terms and words into the search field as facebook social network web 2.0 will get me to pages where any one or two or all of these terms may have been used. And since one of these is a common word, "social," I would be quite likely to get millions of results. 
  • Writing these words as facebook+social+network+web 2.0 will take me to pages that have all these words, not necessarily together, but spread throughout the text.
  • Writing "web 2.0 social network facebook" inside quotation marks however will take me to a text in which these words have been used together, exactly as I wrote them here - which in this case would be impossible since what I wrote makes no sense whatsoever.
But then - here is the thing: All of these results will give me everything about facebook and social networks and web 2.0 sharing, whereas I only need stuff about education, and specifically art education. If I write "art education"+facebook, I do not get much of anything it seems. If I write "art education"+"social networks" I do get more, however the results are somewhat uninspiring. So, I write education+facebook to see what there is available that relates to the general field - and there I do find quite a bit of stuff that I can probably use as good background material. I then refine the search or use alternative searches with "social networks" and "web 2.0," coupled with "education," also by inserting different words such as creativity, art, practice and so on.

From this it already emerges that my specific query has gone unanswered and I now know that not too many art educators out there appear to have written about facebook as a learning domain. On the other hand (and much more importantly!) I can also see that the components of my interest are covered - people have written about facebook and education. And when it comes to art education I already know that plenty of others have written in this field from previous literature reviews that I held on that subject. And all of this when put together is brilliant news for me! It means that I have an original idea that I arrived at 'bisociatively' and which I can now develop further by combining ideas from several different theoretical domains, namely "art education," and "social networks/web 2.0" under whose jurisdiction "facebook" falls.