Cites & Insights: Surveying volume 5

For those of you who don’t read Cites & Insights, and I’ll have a post Real Soon Now about the growing interdependence of the journal and this blog, and why you should read it–you can skip this long post if you like. (So can everyone else, of course…)

This piece would have appeared in “Bibs & Blather,” the “editorial” section of C&I (and also the alternate title of the journal, partly as a reminder to me not to take it too seriously), in previous volumes. There may be more like it: Informal analyses of how the journal fits together and apparent readership.

In past volumes, I’ve tended to look at full and partial pages. I’m not sure why; they’ve always been stored as Word documents, and nothing’s easier than highlighting part of a Word document and getting a partial word count. That’s what I did this year–partly because the typography changed in February 2005, partly because another change late in the year can result in quite a few more words per page. (Best example: the Mid-Fall Issue was theoretically the shortest of the year, at 20 pages–but in fact has more words than seven other issues.)

The volume contains a total of 244,065 words, not including overhead (table of contents, masthead, headers, footers). I checked wording on each story and prepared a spreadsheet categorizing stories by the traditional C&I sections and, in some cases, by what seemed like sensible groupings. Here’s the spreadsheet (just the summary columns) if you want to play with it yourself (although I can’t imagine why).

Here’s what I conclude about volume 5, and it’s not quite what I would have thought without doing the measurements:

  • Going by the topical divisions I used, coverage was reasonably varied, with–and this really is a surprise–blogging being the single largest topic (13.5% of the total), with copyright close behind (12.9%).
  • After that, the “top ten” are filled out by Google and OCA (9.3%), C&I itself (8.7%), library futures (8.1%), Library Access to Scholarship (7.6%), Trends & Quick Takes (6.5%), net media other than blogging/Google/OCA (4.7%), Bibs & Blather (4.5%), and “offtopic perspectives” about old movies on DVD (4.4%).
  • Censorware got short shrift for obvious reasons (0.4%), and there was only one “disContent” reprint and expansion (1%); two other traditional sections also barely appeared–ebooks & etext (1.1%, unless you include Google/OCA) and The Good Stuff (1.8%).
  • The “reporting” experiment accounted for 2.5% of the volume.

In the middle were sections such as Interesting & Peculiar Products (3.4%), The Library Stuff (3.3%), PC Progress (2.7%).

You could group these in different ways: Arguably, 27.5% of the volume was about net media (including Google, OCA, and blogging), 22.2% was copyright-related (if you include Google and OCA), and 16.9% was access-related (again including Google and OCA). How much was directly library-related? I’d argue at least 35%, but you could slice that several different ways.

[If you find numbers adding up to more than 100%, that’s exactly right, because I logged some essays under more than one category. The biggest percentage in the spreadsheet is for Perspectives, with 45.7% of all words, but that’s meaningless.)

I’d love to do a five-year trend analysis (obviously, discussion of net media has increased enormously in the last two years, for example), but it turns out to be more work than the results may deserve. Last year was more varied than I thought, and that’s a good thing.

As for popularity, I might take a closer look at that later, after a full scan of all C&I logs (on the current server) takes place. For 2005 readership, one thing stands out immediately: “Investigating the Biblioblogosphere” was read far more widely than anything else (with the highest issue readership and an independent HTML readership much higher than any other essay), with the possible exception of issues or essays that have been copied to other websites (which has happened in the past and is perfectly OK, but makes readership measurements impossible).

2 Responses to “Cites & Insights: Surveying volume 5”

  1. What did you ever do before the days of automated word counts? 🙂

  2. walt says:

    Well, when it came to book-writing, I got pretty good at guesstimating. Fortunately, by the time I was doing columns where the word length is specified, automated word counts were available…and it turns out that the usual word-length specification is wrong for my columns, ’cause I use short words.