gms | German Medical Science

GMS Medizin — Bibliothek — Information.

Arbeitsgemeinschaft für Medizinisches Bibliothekswesen (AGMB)

ISSN 1865-066X

ScholarlyStats@MedUniVienna: Are usage statistics now a piece of cake?

ScholarlyStats@MedUniVienna: Nutzungsstatistik leicht gemacht?

Original Contribution

Search Medline for

GMS Med Bibl Inf 2008;8(1):Doc09

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/journals/mbi/2008-8/mbi000106.shtml

Published: June 26, 2008

© 2008 Dollfuss.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Abstract

COUNTER (Counting Online Usage of Networked Electronic Resources) standardized usage data from publishers of electronic library resources. Notwithstanding the collection of differently formated usage data from a variety of platforms to piece together an all-encompassing usage report is still cumbersome. MPS technologies is offering exactly this kind of service. The company collects usage data from publisher platforms on behalf of the library and pulls them together into uniform spreadsheet reports. Additionally several statisistical analysis are generated based on the consolidated reports.

Since 2007 the university library at the Medical University Vienna holds a contract with MPS Technologies. The following report will give insight into service coverage, functionality results and experience with ScholarlyStats. Furthermore a few points are listed which have to be considered and discussed in the library before signing this contract.

Keywords: ScholarlyStats, outsourcing, usage statistics, e-journals, e-books, databases, electronic library resources

Zusammenfassung

Die von den Verlagen zur Verfügung gestellten Nutzungsdaten elektronischer Medien wurden durch COUNTER (Counting Online Usage of Networked Electronic Resources) wesentlich vereinheitlicht. Trotzdem bleibt es eine zeitaufwändige Arbeit, die unterschiedlich gestalteten Berichte von den Plattformen herunter zu laden, um daraus einen einheitlichen Nutzungsbericht und statistische Auswertungen zu gewinnen. MPS Technologies bietet nun den Bibliotheken genau dieses Service an. Die Firma sammelt Nutzungsdaten von unterschiedlichsten Verlagen, erstellt einheitliche Excel-Tabellen und generiert daraus statistische Analysen.

Die Universitätsbibliothek der Medizinischen Universität Wien schloss 2007 einen Servicevertrag mit MPS Technologies ab. Der folgende Bericht soll Einblicke geben in Implementation, Funktion, Serviceumfang und Erfahrungen mit ScholarlyStats und listet auch einige Punkte auf, die es vor dem Vertragsabschluss zu bedenken und zu klären gilt.

Schlüsselwörter: ScholarlyStats, Nutzungsstatistik, Auslagerung, e-Journals, e-Books, Datenbanken, elektronische Ressourcen


ScholarlyStats@MedUniVienna

Introduction

The emergence of e-journals, e-books, databases and other web based electronic media unthoughtly showered publishers and subscribers with a plethora of detailed usage data hitherto unknown from print media. A publisher's server meticulously logs access and usage of its stored contend to a file. From this file a server-based software generates a comprehensive usage report which can be accessed by the publisher and subscribers. Based on this data both sides will draw their conclusions, not always in unison.

Initially each publisher bred and formated his own peculiar usage report. So it was indeed a daring feat for the librarians to pull together all the different reports from all the publisher's websites. In 2002 COUNTER (Counting Online Usage of Networked Electronic Resources) was introduced which led to better standardized usage reports. But still it is a cumbersome task for the e-journal account manager to compile one single spreadsheet showing all licensed titles and download figures. This is mainly because each COUNTER report is formated in a slightly different manner according to the programmer's taste. Hence libraries frequently base their decisions on an ad-hoc analysis of individual publishers in the run-up of license renewal negotiations.

Generally and ironically speaking we are warned not to trust statistics drawn by someone else. Nevertheless the Indian based Company MPS Technologies sells exactly this service to libraries [1]. They offer to collect usage data from a variety of platforms on behalf of the library and pull them together into one single uniform spreadsheet report. Moreover with some simple statistics they will provide the librarians with a quick and easy to grasp analysis of their institute's download figures.

Implementation

Our library opted for the smallest service package including 9 out from around 50 platforms to choose. This choice became part of the contract and there is no way to change platforms during a calendar year. Implementation of ScholarlyStats was carried out quickly and without any troubles. Everything is web based so there was no need to install any software. The first task of the library's ScholarlyStats manager was to activate his account with the Library Activation Code in order to register and to set up for user name, password and security check question. Thereafter access details for the selected platforms had to be entered, i.e. user name, password and URL (Figure 1 [Fig. 1]). These are used by the company's data collectors to access and download monthly usage data from the publisher platforms on behalf of the library.

Usage data

ScholarlyStats collects monthly usage data from different vendors (Figure 2 [Fig. 2]). On average publishers need one month to process their logfiles and to generate usage reports. ScholarlyStats has to wait for this one month time span and starts collecting data at the beginning of the following month. So usage data from e.g. March will turn up on ScholarlyStats by mid of May. Usually this is not a problem since usage analysis is not performed daily. As available they pull data together into one uniform spreadsheet. According to ScholarlyStats they could even handle usage data which are not COUNTER compliant. However, Firmature T. [2] reports on odd figures from non compliant platforms which turned up unaltered in the consolidated report. The downloaded raw data are stored separately on ScholarlyStats. These are easily accessible in oder to retrace inconsistent results if necessary.

Recently ScholarlyStats launched release 1.6 featuring new e-book reports [3]. Furthermore a Collection Platform Column was added to the spreadsheet report. So it is now possible to break down platform usage data into several platform collection reports. Hence in a consolidated report usage data from e.g. the platform HighWire Press will be easily separable into collection reports for AAAS, BMJ, Sage and so on. Of course the new relase made up the platform design and scholarlystats' logo using small letters. The ScholarlyStats platform now features six different consolidated spreadsheet reports for usage data:

  • Journal Report 1: Fulltext article requests by journal title, platform and month (Figure 3 [Fig. 3])
  • Database Report 1: Searches and sessions by month and database
  • Database Report 2: Turnaways by month and database
  • Database Report 3: Searches and sessions by month and service
  • Book Report 1: Title requests (since relase 1.6)
  • Book Report 2: Section requests (since relase 1.6)

Usage statistics

Based on its consolidated usage data ScholarlyStats generates several statistics. These so-called "Dashboard Reports" give a quick survey and analysis of an institution's download figures. There are 15 different spreadsheet reports available:

  • Dashboard Report 1: Number of titles by platform
  • Dashboard Report 2: Total requests by platform
  • Dashboard Report 3: Average usage by platform
  • Dashboard Report 4: Top 10 journals by platform
  • Dashboard Report 5: Low usage journals (including zeros)
  • Dashboard Report 6: Zero usage journals
  • Dashboard Report 7: Low usage journals (excluding zeros)
  • Dashboard Report 8: Top 50 journals across platforms
  • Dashboard Report 9: Proportional usage of journal titles

Relase 1.6 in June 2008 featured the following new e-book reports

  • Book Dashboard Report 1: Number of books by collection platform
  • Book Dashboard Report 2: Number of book requests by collection platform
  • Book Dashboard Report 3: Number of section requests by collection platform
  • Book Dashboard Report 4: Average book use by collection platform
  • Book Dashboard Report 5: Top use books by collection platform
  • Book Dashboard Report 6: Zero usage books report

These Dashboard Reports (DR) can quickly reveal some surprisingly interessing details about an institution's download volume with just a few mouse clicks. This attained preprocessed information gets more valuable and deliveres a better survey if ScholarlyStats covers more platforms to collect data from. Otherwise the librarian still has to glue together results from ScholarlyStats with self pulled data from other vendors to complete the analysis.

As with all statistical figures, one has to be cautious. This is particularly true with DR1 and DR3. ScholarlyStats assumes that the number of e-journals found in a collected raw data report from a publisher or aggregator platform equals the number of licensed e-journals at an institution. Such as Blackwell-Synergy only generates usage reports listing all 1043 titles in a body no matter if the library holds subscriptions to all. Of course this report then just denotes the download figures for the institution leaving non subscribed journals with zero usage. But DR3 relies on the accurate numbers of title holdings from DR1 hence it could produce utterly wrong figures if not critically watched and corrected by the librarian. Anyway, viewed with common sense Dashbord Reports provide a valuable service.

A library's ScholarlyStats account administrator can also register other collegues in order to give them individual access to the report area of the platform. Older reports are stored away in the archive area and can be reviewed there.

Integration Partners

The library may choose to send its usage data to a third party, a so-called Integration Partner, e.g. a subscription agent. This Integration Partner in turn could use the download figures to automatically calculate an institution's price per download per journal as an additional service to the library. However, hitherto only a few Integration Partners are offered and there might be other objections against supplying vendors with usage data. The university library at the Medical University Vienna does not use this feature yet.

ScholarlyStats features a SUSHI 1.0 interface, a protocol for the automatic delivery of usage data from one system to another through XML. So usage data can be send to an external electronic resources evaluation tool [4]. Or in the future the library's own Electronic Resource Management (ERM) system could be an Integration Partner receiving usage data directly from ScholarlyStats.

Considerations befor signing a contract with ScholarlyStats

Before the outsourcing of usage data collection to ScholarlyStats a few considerations might be helpful:

  • Does outsourcing of usage data collection pay-off for your institution?
  • How many and which platforms should be covered by ScholarlyStats?
  • Are the registered lists of IP ranges up-to-date?
  • If there is a choice - which usage data should be collected from a platform?
  • Who is going to administer the library's account at ScholarlyStats?
  • Who is communicating via e-mail with the ScholarlyStats team?
  • Who analyses the ScholarlyStats reports with critical common sense?

Upshot

The implementation of ScholarlyStats goes quick and smooth. Communication with the ScholarlyStats team via e-mail is easy and reliable. Collection of usage data and formating them into one single spreadsheet report is a highly valuable service for a busy library. The benefit of this service gets bigger if ScholarlyStats covers most or all platforms onto which a library holds e-journal, e-book and database subscriptions. A critical approach to the provided reports and a further in-depth analysis and report of download figures by a librarian who is firm in statistics will be still inevitable. Otherwise you will pay for data that will sit idle. However, this essential task is eased a lot by ScholarlyStats.


References

1.
MPS Technologies Ltd. For a clearer view of your usage statistics. ScholarlyStats brochure. Online available at: https://www.scholarlystats.com/sstats/ScholarlyStats-NewBrochure.pdf. External link
2.
Firmature T. A review of ScholarlyStats. The Charleston Advisor. 2006;8(2):38. Online available at: https://www.scholarlystats.com/sstats/TwilaReviewInCharlestonAdvisor.pdf. External link
3.
MPS Technologies Ltd. ScholarlyStats 1.6. ScholarlyStats Release 1.6 Notes 2008.
4.
McMeekin J, Sullivan S. ScholarlyStats at University of Melbourne. Presentation at CEIRC Forum on Usage Statistics for E-resources. 2007. Online available at: http://eprints.rclis.org/archive/00011307/01/mcmeekin.pdf. External link