Bibliographic analysis for fun and collection development

You know how you get a brand new hammer and suddenly you notice all these nails sticking out?

So I've been working more with Ref2RIS. And in the meantime some of my colleagues and I were talking about analysing researchers' bibliographies for nefarious purposes, and I suddenly realised that doing such a thing might also help me get the handle I desperately need on one of the subject areas I'm attempting to be a liaison librarian for without having had any handover or background in.

And then I realised that, instead of staring glumly at some PhD thesis bibliography and having my eyes glaze over, I could just run it through Ref2RIS, pull all the references into Endnote, and sort by journal title.

It did take me two hours to create the conversion file, but on the other hand I'm getting quicker at that. And then I sorted, and did a quick count, and came up with the following data:

The bibliography for this thesis contained 133 references, of which 1 was a website, 9 were books/reports/manuals, and the bulk of 123 were journal articles from 27 different journals.

16 journals were used for only 1 reference each;
2 journals for 2 references;
2 journals for 3 references;
1 journal for 4;
2 journals for 5;
1 journal for 12;
1 journal for 18;
1 journal for 19;
1 journal for 34 references (over a quarter of the entire bibliography)

I also discovered that this last journal is one that our library doesn't hold.... (We do hold everything that was used 4 or more times; I got bored before checking the less-used journal titles.)

Obviously more research is required

• to find out if this is a significant gap in our collection or a fluke of this particular thesis; and
• to figure out if there are any other interesting patterns in usage;

but if the researchers have had the courtesy to all use the same citation style then it should be pretty quick research.