> Originally, since these numbers are so far off what others
> had, I thought
> these statistics were incorrect so I checked and rechecked
> them. But they
> are correct. They also demonstrate what I mentioned above, since I can
> assume that nobody will be able to put in a subject heading/subject
> subdivision search, so I conclude that my users are searching
> originally by
> keyword, then clicking on the subjects or authors that they
> see. (Titles are
> only series, so this is minimal)
As I think others have said - stats are not the full story. Although the
stats you have are interesting, you then draw a conclusion based on an
assumed explanation for the stats. There are two major issues (as I see
it):
You assume that 'subject' searching is somehow more desirable - but
there is no evidence to support this conclusion - do users more often
get what they want and/or need when they use a subject search?
You assume that the 'subject' searches are done using appropriate
subject headings - but there is no evidence (that you quote) to say that
they aren't just putting keywords into the subject search box
Also, I note from your construction of your Advanced search page that
the 'subject browse' isn't actually a direct search of the catalogue -
presumably the stats count a 'subject browse' each time a user clicks
this 'browse' option, rather than actually carrying out the search?
That said, I'm not trying to write off your results, which are
interesting - it's just that I don't really trust statistics to give us
the detailed information we need to make the kind of analysis you are
attempting here.
Owen
Received on Thu Feb 14 2008 - 04:26:30 EST