FWIW: Alexa is really bad. It's largely based on people who have the
Alexa Toolbar Installed? Do you? I don't either.
Compete is generally considered the best service. From experience, I
can see it roughly tracks relative changes. But it doesn't have the
magnitude right—it has us at something like 1/4 of actual traffic. It
also seems to see our short-term Google traffic more than our stable
non-Google traffic. If Google rises or falls 20%, Compete makes it
50%. My theory, though, is that the large errors it makes *might* be
the same across similar sites.
Anyway, that's the story. Alexa is crap. Compete's the best. And nobody is good.
T
On Tue, Mar 16, 2010 at 12:56 PM, Ross Singer <rossfsinger_at_gmail.com> wrote:
> My question would be, are these Alexa statistics even remotely relevant?
>
> To quote that Wikipedia's percentage is orders of magnitude higher --
> how is there any notion that any of this has to with the sort of
> "articles" that we would be able to work with? Facebook is higher
> than Wikipedia, although I wouldn't say libraries' use of Facebook has
> had a tremendous impact.
>
> What percentage of Wikipedia traffic goes from the search result page
> to other Wikipedia pages?
>
> It seems, to a large degree, that comparing Worldcat to Wikipedia is a
> case of apples to oranges: Worldcat to LibraryThing is a much better
> analog.
>
> I seriously think that if library catalogs produced sitemaps that
> included links to Worldcat and even LibraryThing, that this would be a
> simple, cheap way to improve our findability on the web.
>
> -Ross.
>
> On Mon, Mar 15, 2010 at 11:59 AM, Weinheimer Jim <j.weinheimer_at_aur.edu> wrote:
>> If the numbers I gave you are not compelling, here is a report that might interest you:
>> http://projectinfolit.org/pdfs/PIL_Fall2009_Year1Report_12_2009.pdf
>>
>> They discovered that students for school work, students use in this order:
>> 1. Course readings (surprising!)
>> 2. Google
>> 3. Scholarly research databases
>> 4. Opac
>> ...
>>
>> In the library, we see that students use online databases over our library catalogs.
>>
>> But still, it seems to me as if the statistics I quoted are worth something because of the huge discrepancies. If we say that it is OK that our work is only used by 0.016% of the populace, it puts us in a rather precarious position. We can try to say that these represent the "serious" people, but is that true? Is what we make aimed only at 0.016% of the populace? I don't want my job to depend on that.
>>
>> It seems as if this is a very small percentage for us to settle for, especially when we compare it to Wikipedia, where there are many, many people, each of whom is very definitely interested in getting information. Are these people "serious?" I know lots of researchers who go to Wikipedia and they are certainly serious. They use it to get a quick overview and for citations.
>>
>> It seems much more logical to me to infer that people are not looking in Worldcat for all sorts of reasons and that if we want to reach a larger share of the populace, we should place our records in lots of places that receives more visitors.
>>
>> James Weinheimer j.weinheimer_at_aur.edu
>> Director of Library and Information Services
>> The American University of Rome
>> via Pietro Roselli, 4
>> 00153 Rome, Italy
>> voice- 011 39 06 58330919 ext. 258
>> fax-011 39 06 58330992
>>
>>
>> -----Original Message-----
>> From: Next generation catalogs for libraries [mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Laval Hunsucker
>> Sent: Monday, March 15, 2010 4:23 PM
>> To: NGC4LIB_at_LISTSERV.ND.EDU
>> Subject: Re: [NGC4LIB] OCLC and Michigan State at Impasse Over SkyRiver Cataloging, Resource Sharing Costs
>>
>> Thanks very much for the descriptive statistics.
>>
>> I'm not sure how representative these specifically
>> Alexa data are for any definable population, not to
>> mention the population I had in mind : faculty and
>> students affiliated with institutions of higher
>> education in Europe and North America, who
>> aren't necessarily dependent on search-engine
>> approaches -- i.e., how capable these data are of
>> constituting a basis for rejecting a research
>> hypothesis that significantly more than a ( "very,
>> very" ) few members of that population -- and
>> possibly of some other population you might wish
>> to define -- use WorldCat ( to say nothing about
>> knowing that it exists ).
>>
>> Also not sure how comparable in this context
>> LibraryThing really is, to say nothing of Wikipedia.
>>
>> You used the term "can be proven", strong language
>> indeed, and I as researcher had a reason for reacting
>> with concepts like "methodology", "valid and
>> reliable", and "inferential statistics".
>>
>> Needless to say, therefore : the formulations "not
>> sure how representative" and "not sure how
>> comparable" which I employ above are euphemisms
>> for : "You have not yet by any means convinced me."
>> or "I still discern no evidence to justify your earlier
>> categorical claim.", which, as I wrote, doesn't accord
>> with my own experiential and anecdotal data. I think
>> we can conclude that the question has yet to be
>> responsibly investigated ( after, of course, we have
>> operationalized the terms "few", "use", "know
>> about", and of course "people" ;-) ).
>>
>> So -- I'm just gonna have to live with my own
>> perspective for the meantime. Pas grand-chose.
>>
>> None of this is meant to detract from your
>> proposition, below, that "We need other routes
>> as well.", and no, I don't find that the least bit
>> "unfortunate or sad". And also not meant to
>> neglect your important implicit question below,
>> i.e. : What is too much to expect ? But the
>> corollary is of course : What is enough to be
>> satisfied ?
>>
>> However all of that may be, any measures should
>> preferably be founded on good evidence, and by
>> extension therefore on solid research. There's
>> nothing sacrosanct about WorldCat. If it's not worth
>> what it ( in various ways ) costs, out with it. If it is,
>> well, . . ..
>>
>> I like the rest of your post below -- but can't refrain
>> from opining that while, as you say, OCLC and
>> WorldCat are -- like your example the catalogue card
>> -- in principle dispensable and "could disappear", the
>> same is true of librarianship and even of what we now
>> know as libraries. None of these things is an end in
>> itself. At least some of the "values of librarianship"
>> can be considered, for practical purposes, ends in
>> themselves -- but then again, they are neither the
>> creation or prerogative of, nor unique to, librarianship
>> and libraries.
>>
>> Attaining the ideal situation in which librarianship
>> and perhaps even libraries are superfluous is not going
>> to be an easy act, I think -- for all kinds of practical
>> and policy ( *not* philosophical ) reasons. Some may
>> find consolation in this prognosis. With forty-seven
>> years of intensive library involvement, I don't ; I feel
>> no need for such consolation.
>>
>> Anyway, as you've indicated : It's going to be interesting,
>> for quite some time to come.
>>
>>
>> - Laval Hunsucker
>> Breukelen, Nederland
>>
>>
>>
>>
>> ----- Original Message ----
>> From: Weinheimer Jim <j.weinheimer_at_AUR.EDU>
>> To: NGC4LIB_at_LISTSERV.ND.EDU
>> Sent: Mon, March 15, 2010 10:44:26 AM
>> Subject: Re: [NGC4LIB] OCLC and Michigan State at Impasse Over SkyRiver Cataloging, Resource Sharing Costs
>>
>> Laval Hunsucker wrote:
>> <snip>
>> I have no 'proof' immediately available one way or the other ( and would like to know what methodology, what valid and reliable research results, and what apt inferential statistics lie behind the declaration that "it can be proven that very, very few people use Worldcat or even know about it" ; maybe you have a quick reference at hand ), but I've for some years been constantly hearing disciplinary scholars saying in an off-hand fashion that their search in WorldCat yielded such-and-such a result. Isn't at least some of this dependent upon how visible the resource has locally been made, and how well it has been locally marketed ?
>> </snip>
>>
>> If you look at the Alexa site, you can see the statistics:
>> Worldcat: http://www.alexa.com/siteinfo/worldcat.org
>> Yesterday 0.022 +10% (they had a good day yesterday)
>> 7 day 0.0177 -3%
>> 1 month 0.0164 +1%
>> 3 month 0.0164 -7%
>>
>> Therefore, about 0.016% of searches go to Worldcat.
>>
>> Compare to LibraryThing:
>> http://www.alexa.com/siteinfo/librarything.com
>> Yesterday 0.018 -30%
>> 7 day 0.0254 -0.7%
>> 1 month 0.0259 -8%
>> 3 month 0.0277 -1%
>>
>> Therefore, LibraryThing has a much higher percentage of use than Worldcat. (0.027%)
>>
>> Compare to Wikipedia:
>> http://www.alexa.com/siteinfo/wikipedia.org
>> Yesterday 12.51 -5.3%
>> 7 day 12.97 +0.2%
>> 1 month 12.975 -0.09%
>> 3 month 12.748 +13.77%
>>
>> Wikipedia gets a major number of hits.
>>
>> Based on these statistics (which have remained pretty constant) it would make sense to conclude that a record placed into WorldCat will make it available to the least number of people. Placing it in LibraryThing or Wikipedia would increase its use. The conclusion I make from this is that while we can go ahead and put our records into WorldCat, we shouldn't expect too much. We need other routes as well.
>>
>> Perhaps this is unfortunate or sad, but it is a fact nevertheless. What about when we add Google Scholar and Google Books into the equation? A friend of mine at FAO of the UN mentioned that they had recently placed the AGRIS database (an agricultural database) into Google Scholar and the hits went up exponentially. This only makes sense.
>>
>> <snip>
>> May well, but library-centric focus in itself needn't -- witness my experience, but also in principle -- disqualify an instrument as enduser-appropriate and enduser-used, at least in a full-blown academic environment. Other factors can play determinant roles.
>> </snip>
>>
>> Agreed, but products that are objectively better die every day if you can't get people to use them. Libraries have their collections and these collections need to be used, so I don't care how people find out about my materials. I can suggest certain ways, but if people want to use their own methods, that's fine with me. This is the world that is changing in fundamental, and as yet, very unclear ways.
>>
>> What is important is to save libraries (in whatever form they take) and the values of librarianship. We should not confuse this with maintaining "OCLC" and/or "WorldCat." Either one of these entities could disappear and libraries should be able to continue.
>>
>> <snip>
>> Perhaps -- but what will then be left of librarianship after they have been really solved ? Of course, once *all* of the librarianship problems have been solved ( i.e. from the perspective of those outside parties who alone ( can ) lend the field its legitimacy ), librarianship will not exist, or at least need to exist. Shouldn't your consolation ( as an apologist/advocate ) be predicated on the thought that not all of those problems will ever be solved ?
>> </snip>
>>
>> Interesting question. I think there will be many solutions, but no "ultimate solution." As soon as one "solution" is implemented, a dozen more will arise. When you solve these dozen, your original solution will have to be rethought.
>>
>> Some may consider this to be the very essence of futility, but to me, it represents the idea of progress. As new ideas and capabilities arise, you must adapt yourself to them. Just as the card catalog solved a myriad of problems, and created a host of others, we are in a similar situation today. The various ways of producing, and even using, cards evolved, and so will the methods we devise.
>>
>> Jim Weinheimer
>>
>>
>>
>>
>>
>
--
Check out my library at http://www.librarything.com/profile/timspalding
Received on Tue Mar 16 2010 - 13:20:24 EDT