Response to BEPress report

[Liblicense-L Listowner <liblicen@pantheon.yale.edu>]

Ann: I wonder if you would be so kind as to post my response, 
below, to the recent exchanges on Liblicense for me.

Thank you so much.  Peter Shepherd

* ******

On Metrics 

I  have read with interest the comments of Joe Esposito and Phil 
Davis on the Berkeley Electronic Press report on the potential 
inflationary effects of internet robots, crawlers, etc. on online 
usage statistics. As far as usage reports are concerned, 
there seem to me to be two relevant scenarios, in one of which 
there should be no inflation of the usage statistics due to 
internet robots, etc., while in the other scenarios there is a 
potentail effect. Both scenarios are described below:

Scenario 1: Usage of subscription-based online resources, whose 
access is limited to a closed community ( this is the scenario in 
a typical library or library consortium to which vendors supply 
COUNTER reports). This type of usage, as Phil Davis indicates, is 
very unlikely to be susceptible to internet robots, etc.

Scenario 2: Usage of online resources in an open access 
environment, or on other platforms without access control: here 
the effect of internet robots, etc. should be taken into account, 
as there are likely to be platforms that have no filters in place 
to protect against such robots. Such platforms are, 
however, also going to be visited by Google, Yahoo and other 
legitimate web search engines which may sweep the site multiple 
times

To summarise, in Scenario 1, which covers the existing COUNTER 
usage reports for libraries and library consortia, I think it is 
reasonable to assume that the inflationary effect of internet 
robots and their ilk on the COUNTER usage statistics is 
negligible. In Scenario 2 such robots do,in all likelihood, have 
an inflationary effect. An important question, however, is 
whether such inflation is proportionately the same across all 
vendors. Perhaps this is a reasonable assumption, perhaps 
not. Most large vendors maintain lists of internet robots and 
discard usage statistics from such sources; but do they all use 
the same lists?  

As far as COUNTER is concerned, we are investigating possible 
solutions to mitigate the effect of internet robots and their 
ilk. It becomes increasingly important to do so as the prospect 
of COUNTER usage statistics being reported in open access and 
global situations grows.

Peter Shepherd
Director
COUNTER