[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: On-line access to archives/archival collections



Craig Wright wrote:
> One of my favorite questions (which rarely elicits a response) is what
> exactly does a hit mean? What useful information can we deduce
> from a page receiving 200 or 2000 hits?

I agree that this is a very important question.  A good overview article on the issues related to hits vs. page views vs. visits is
http://hotwired.lycos.com/webmonkey/98/16/ 

> My general impression is that historical libraries and
> archives, which rarely have significant patronage vis a vis public
> libraries, or in the case of many academic archives vis a vis the
> college/university library -- use these figure to show the powers
> that be that somehow we are serving a whole lot more people
> than our patron visit count indicates.

I think this is a very good point and is true of just about anyone putting content on the Web.  There are two very distinct ways of using server statistics: 
- showing how big the numbers are in order to demonstrate the value of something you've put on the Web,
- honestly determining what the numbers really mean about the strengths and limitations of the site and ways it could be improved.

In order to do the former, "hits" (each request made for a file on the server, whether that's an image, a page that never fully loads, a "404 not found" error page, a probe by a software agent used by a commercial search engine, or even a banner ad put on your site by someone else) are best.  In order to do the latter, much more important are number of page views, type of files requested, last URL visited before making that request, and even type of browser used (all standard data collected by most server log software).  As the article referenced above points out, figuring out which hits belong to the same visitor is often what we most care about but also often tough to determine.  There are tools like cookies that can be used to facilitate this in some cases.

> Perhaps "casual surfer" is too vague a term to be useful. Maybe categorizing
> users/web visitors would be more helpful - if we can deduce useful categories.

It is very helpful to come up with what are often referred to as "user profiles."  A good place to start is Usable Web: Users, Tasks (http://usableweb.com/items/userstasks.html).  These are never absolute categories but just help identify what sorts of attributes you're assuming about your intended user group:
- What sort of browsers are they using?
- Are they accessing the site from home or work?
- Do they tend to print out pages, save them to their hard drive, bookmark them?
- How much experience do they have doing research on the Internet?
- What's their level of experience with the technical terminology of your domain (both archival terms and ones specific to the collection)?
- Do they tend to be self-directed or assigned to do research by someone else?
- Do they visit regularly or just tend to be one-time visitors?

> Statistics suggest that the average webpage is viewed for less than a minute
> (and this is sometimes used as an argument against providing content
> - IMHO a very erroneous position which could easily lead down the slippery
> slope taken by some historical museums)

It's not evidence against providing content but rather against providing pages whose content isn't clearly connected to the rest of the information architecture of the site.  If a user spends a few seconds on a page and it doesn't seem relevant to her, she'll tend to pick one of three choices:
- follow some link on that page to look further,
- go to the search engine located on that site (if there is one),
- hit the "back" button of the browser and go somewhere else.

The search engine may get her there, but it's rarely as effective as a well-labeled link.  Users often pick the wrong search terms and end up lost.

A very valuable source of user data is the log of queries submitted to your site's search engine (if you have one).  As Jared Spool points out, the terms users type in to the query box are the exact terms they would have most likely clicked on if there had been links with those terms as labels.  

If the back button leads back to some other site, then you've just lost a potential user of your materials.  And if that back button leads to one of your other pages, then the user is statistically much less likely to find what she's looking for (again supported by tests done by Jared Spool and company), since they lose that valuable "scent."  We want them to ideally follow a trail of valuable information, rather than having to back out and try again.

Each page should have something of potential use, but it should also include things like global navigation links that help convey the navigational structure of the site.




========================================================
Cal Lee
Electronic Records Project Archivist
Kansas State Historical Society
Phone: 785-272-8681, ext. 280         Fax: 785-272-8682
http://www.kshs.org/archives/recmgt.htm
                        "Obsolete power corrupts obsoletely."
                                                        - Ted Nelson

A posting from the Archives & Archivists LISTSERV List!

To subscribe or unsubscribe, send e-mail to listserv@listserv.muohio.edu
      In body of message:  SUB ARCHIVES firstname lastname
                    *or*:  UNSUB ARCHIVES
To post a message, send e-mail to archives@listserv.muohio.edu

Or to do *anything* (and enjoy doing it!), use the web interface at
     http://listserv.muohio.edu/archives/archives.html

Problems?  Send e-mail to Robert F Schmidt <rschmidt@lib.muohio.edu>