 |
 |
Judging by your proposed methodology Mies you’re extremely sane and doing everything right. I think Susan G and the guys at Serials Soln COUNTER 360 do a manual check to “clean up” the Title data before normalising into their system. Susan - are you able to give a ball park %error figure for the # of titles that need treatment? In the UK one of our biggest non-COUNTER culprits is a company called IHS - http://uk.ihs.com/ who have always resisted the call to join COUNTER. A new rep has given me fresh hope: 1. “Counter Compliant usage stats – I am looking into this and getting feedback from the IHS Sales team as to what their customers are saying, before I go to the Product Manager’s for a response. I will come back to when I have something to report.” If you deal with this company please lobby your rep! And naming others – Knovel; ICE Virtual Library; British Standards On-Line… I’m sure you have your favourites? C. === From: [log in to unmask] [mailto:[log in to unmask]] On Behalf Of Mies Martin Sent: 29 September 2009 15:09 To: [log in to unmask] Subject: Re: [lib-stats] consistent meta-data from Royal Society of Hi Cliff, Thanks for the note. On the one hand I'm glad to learn that I'm not imagining these sorts of things. When I explain these oddities to my colleagues they think I'm nuts. They be happy to learn that a) I'm not nuts and b) nothing is as consistence as it may seem. Thanks for suggestion concerning using the title as a key field. For our little system I'm considering separate sets of tables (Journal title, year of usage data and publisher) for each of our providers (RSC, IEEE, ACM, ect.). In other words, for each provider we would have separate tables instead of one large database. I know it this might seem rather maddening but this way if one table goes kaput then it would not impact everything else. Moreover it would be rather easy to rebuild a small set of data rather than a whole collections worth. mies Mies Martin Digital Resource Coordinator J.R. Van Pelt / Opie Library Michigan Technological University 1400 Townsend Dr. Houghton, Mi 49931 Ph: (906) 487-2135 email: [log in to unmask]<mailto:[log in to unmask]> ----- Original Message ----- From: "Cliff Spencer" <[log in to unmask]> To: "[log in to unmask]" <[log in to unmask]> Sent: Tuesday, September 29, 2009 5:28:25 AM GMT -05:00 US/Canada Eastern Subject: RE: [lib-stats] consistent meta-data from Royal Society of Chemistry Hi Mies If you are using print issn as your key field (rather than say the title field) in Microsoft Access then you may have to run a validation check to ensure that issn=title. Because there are so many possible data errors in collecting/reporting usage stats there is no ideal key field on which to base subsequent tasks (Title field is far worse but easier to correct). I found missing issns to be a bigger problem than erroneous issn – Access doesn’t like importing missing data say from .csv file (haven’t tried xml import as have gone back to Excel and scripts). I noticed a mail a while ago about a similar data validation problem – i.e. seemingly fixed totals changing month on month – this is also quite common; I think because of the way the data has been collected in the past. Many publishers have improved this (for SUSHI) and although I haven’t done a sample I think the rate of errors has decreased (as the economists would say). May I apologise to the folks who have mailed me off-list with queries and not had a direct reply – and indeed those who have had a longish wait to be added to the list. I’ll be back on the case soon. BW. Cliff. === From: [log in to unmask] [mailto:[log in to unmask]] On Behalf Of Mies Martin Sent: 28 September 2009 13:53 To: [log in to unmask] Subject: [lib-stats] consistent meta-data from Royal Society of Chemistry I'm attempting to build a rather simple but effective Access database that will assist managing our usage statistics. For my test I'm using four years worth of our usage data from the Royal Society of Chemistry. Because I'm in the early stages of this little project I'm creating the initial table by hand. In doing this I've come across an oddity. In the JR1 reports for the Royal Society of Chemistry I'm noticing inconsistencies in the identification of ISSN. Let me explain. In 2006 RSC identified the print ISSN for Annual Reports Section A (Inorganic Chemistry) 1460-4760 and the electronic ISSN 0260-1818. In 2007 and 2008 the print and electronic ISSNs for Annual Reports Section A (Inorganic Chemistry) are identified as follows 0260-1818 and 1460-4760 respectfully. Now, for 2009, that same title the print and electronic ISSN are once again identified as it had been in 2006, i.e. 1460-4760 and 0260-1818. This pattern is not limited to only Annual Reports Section A (Inorganic Chemistry), rather it is reflected in all of the titles list in the JR1 report. Obvious for something that I'm attempting to do this can have important consequences. So I'm wondering whether or not anyone else noticed this oddity with RSC or any other usage statistics provider? Respectfully, mies ~~~~~~~~~ Mies Martin Digital Resource Coordinator J.R. Van Pelt / Opie Library Michigan Technological University 1400 Townsend Dr. Houghton, Mi 49931 Ph: (906) 487-2135 email: [log in to unmask]<mailto:[log in to unmask]>