APSR WorkshopAPSR Workshop

4 May 20064 May 2006

Arthur Sale

The OA Advantage, accession The OA Advantage, accession

policies, and the RQFpolicies, and the RQF

APSR Workshop: 4 May 2006 2

Open Access RepositoriesOpen Access Repositories

• UTas’ present repository is an UTas’ present repository is an Open Open Access RepositoryAccess Repository (OAR). (OAR).

• This means that it is oriented towards the This means that it is oriented towards the goal of the Open Archives Initiative: goal of the Open Archives Initiative: making research results available to all, making research results available to all, free, online, and at the time of asking.free, online, and at the time of asking.

• The primary readers are other The primary readers are other researchers looking for information.researchers looking for information.

APSR Workshop: 4 May 2006 3

StatusStatus

Prototype has been running since Prototype has been running since 20042004 operated operated by the by the School of ComputingSchool of Computing

About to become About to become University LibraryUniversity Library responsibility…responsibility…

Deposit is required for School of Computing staff Deposit is required for School of Computing staff and studentsand students; voluntary for others; voluntary for others

PhD theses are mirrored in OAR PhD theses are mirrored in OAR as well as separate ADT repositoryas well as separate ADT repository

APSR Workshop: 4 May 2006 4

What it is What it is notnot……

• Not a vehicle for electronic publishing of Not a vehicle for electronic publishing of journals nor e-booksjournals nor e-books

• Not a repository for images or text of historical Not a repository for images or text of historical or local collections of objectsor local collections of objects

• Not a service based primarily around Not a service based primarily around preservation (archiving)preservation (archiving)

• Not an e-research facility (raw research data)Not an e-research facility (raw research data)

APSR Workshop: 4 May 2006 5

Why?Why?

Services with distinct characteristics need Services with distinct characteristics need separate and more expensive repositories, for separate and more expensive repositories, for example image collections, e-publishing (though example image collections, e-publishing (though maybe using the same software base)maybe using the same software base)

The highest priority is therefore afforded to :The highest priority is therefore afforded to :

– gaining an advantage in citation metrics (for research gaining an advantage in citation metrics (for research standing reasons) and standing reasons) and

– the technology is mature, cheap and the goal the technology is mature, cheap and the goal achievable in a short time frame, decision-makers achievable in a short time frame, decision-makers willing.willing.

APSR Workshop: 4 May 2006 6

The Open Access AdvantageThe Open Access Advantage

• Research documents which are online Research documents which are online attract attract 0.5x to 2.5x more citations0.5x to 2.5x more citations than than if not, depending on discipline.if not, depending on discipline.

• An OAR is the best way to put a research An OAR is the best way to put a research document online, because it is indexed document online, because it is indexed and has a metadata harvesting scheme. and has a metadata harvesting scheme. It also facilitates other services based on It also facilitates other services based on the OAR’s existence.the OAR’s existence.

APSR Workshop: 4 May 2006 7

Well-established researchWell-established research• Harnad S & Brody T (2004). ‘Harnad S & Brody T (2004). ‘

Comparing the Impact of Open Access (OA) vs. Non-OA ArComparing the Impact of Open Access (OA) vs. Non-OA Articles in the Same Journalsticles in the Same Journals’. ’. D-Lib MagazineD-Lib Magazine, 10(6), ISSN 1082–9873. , 10(6), ISSN 1082–9873.

• Lawrence, S. (2001) Lawrence, S. (2001) Online or Invisible?Online or Invisible?, , NatureNature 411 411 (2001) (6837), p. 521. (2001) (6837), p. 521.

• Kurtz MJ, Eichhorn G, Accomazzi A, Grant C, Demleitner Kurtz MJ, Eichhorn G, Accomazzi A, Grant C, Demleitner M, Henneken E, Murray SS (2005). M, Henneken E, Murray SS (2005). The effect of use and access on citationsThe effect of use and access on citations. . Information Information Processing & ManagementProcessing & Management 41(6): 1395–1402. 41(6): 1395–1402.

• Antelman K (2004). Antelman K (2004). Do Open-Access Articles have a Greater Research ImpactDo Open-Access Articles have a Greater Research Impact??. . College and Research LibrariesCollege and Research Libraries 65(5):372–382. 65(5):372–382.

APSR Workshop: 4 May 2006 8

Antelman (2004)Antelman (2004)

APSR Workshop: 4 May 2006 9

Harnad Harnad et al et al (2004)(2004)

APSR Workshop: 4 May 2006 10

Lawrence (2001)Lawrence (2001)

119,924 computer science articles119,924 computer science articles

APSR Workshop: 4 May 2006 11

Lawrence 2Lawrence 2

% online articles cited more than offline% online articles cited more than offline

APSR Workshop: 4 May 2006 12

Accession policiesAccession policies

There are two main classes of accession There are two main classes of accession policies:policies:

1.1. VoluntaryVoluntary – deposit is at the author’s discretion – deposit is at the author’s discretion

2.2. MandatoryMandatory – deposit is required by the author’s – deposit is required by the author’s institution (or by a grant funder eg Wellcome institution (or by a grant funder eg Wellcome Trust, NIH’s policy, RCUK’s current discussions Trust, NIH’s policy, RCUK’s current discussions or by national policy eg Netherlands and France, or by national policy eg Netherlands and France, EU proposal)EU proposal)

NOTE: ‘deposit’ is not the same as ‘open access’; NOTE: ‘deposit’ is not the same as ‘open access’; deposit = acquiring the e-copy. deposit = acquiring the e-copy.

APSR Workshop: 4 May 2006 13

The evidence - 1The evidence - 1

OARs that operate a OARs that operate a voluntaryvoluntary deposit deposit policy collect at most policy collect at most 15%15% of the available of the available documents.documents.

This seems to be the case no matter what This seems to be the case no matter what incentives are offered or what promotional incentives are offered or what promotional strategies are tried. It is worldwide strategies are tried. It is worldwide experience. Low promotion activity levels experience. Low promotion activity levels leads to deposits at even lower levels. leads to deposits at even lower levels. Avoidable work is avoided, even if trivial.Avoidable work is avoided, even if trivial.

APSR Workshop: 4 May 2006 14

The evidence - 2The evidence - 2

Where the authors are Where the authors are requiredrequired to to deposit in an OAR, deposit in an OAR, 80%+80%+ of the of the available documents can be collected.available documents can be collected.

Hard enforcement of the policy is not Hard enforcement of the policy is not required; compliance exceeding 80% required; compliance exceeding 80% is fuss-free; compliance grows over 2-is fuss-free; compliance grows over 2-3 years.3 years.

APSR Workshop: 4 May 2006 15

Author attitudesAuthor attitudes

Seminal JISC study by Swan & Brown, Seminal JISC study by Swan & Brown, Open Open access self-archiving: An author studyaccess self-archiving: An author study. . 2005. 2005. http://eprints.ecs.soton.ac.uk/10999/http://eprints.ecs.soton.ac.uk/10999/

• 81%81% of authors will deposit willingly if required of authors will deposit willingly if required by institutionby institution

• 13%13% will deposit reluctantly if required will deposit reluctantly if required

• Only Only 5%5% say they will refuse to deposit even if say they will refuse to deposit even if requiredrequired

APSR Workshop: 4 May 2006 16

The proofThe proof

See:See:

Sale, Prof Arthur (2006) Sale, Prof Arthur (2006) Comparison of IR content Comparison of IR content policies in Australiapolicies in Australia. . First MondayFirst Monday 1111(4). (4). April 2006. April 2006. http://www.firstmonday.org/issues/issue11_4/sale/http://www.firstmonday.org/issues/issue11_4/sale/

Sale, Prof Arthur (2006) Sale, Prof Arthur (2006) The impact of mandatory The impact of mandatory policies on ETD acquisitionpolicies on ETD acquisition.. D-Lib MagazineD-Lib Magazine 1212(4). April 2006. (4). April 2006. http://www.dlib.org/dlib/april06/sale/04sale.htmlhttp://www.dlib.org/dlib/april06/sale/04sale.html

APSR Workshop: 4 May 2006 17

Research articlesResearch articlesThe The First MondayFirst Monday paper deals with research paper deals with research

articles in OARs. The following slide articles in OARs. The following slide shows shows

• the number of deposited research the number of deposited research documents published in a particular year documents published in a particular year

• as a percentage of the officially reported as a percentage of the officially reported DEST research output DEST research output

• for all seven Australian universities that for all seven Australian universities that had an OAR established prior to 2004.had an OAR established prior to 2004.

APSR Workshop: 4 May 2006 18

OAR chartOAR chart% of DEST output

0%5%

10%15%20%25%30%35%40%45%50%

ANU

Curtin

Melb

ourne

Mon

ash

Que

ensla

ndQ

UT

Tasman

ia

2004 2005

APSR Workshop: 4 May 2006 19

ThesesThesesNow let’s do this for repositories that contain Now let’s do this for repositories that contain

ETDs (electronic theses and dissertations), ETDs (electronic theses and dissertations), the subject of the the subject of the D-LibD-Lib article. article.

There are many more ETD-capable universities, There are many more ETD-capable universities, and more of them have ‘requirement and more of them have ‘requirement policies’.policies’.

The following chart shows The following chart shows 2005 theses as a 2005 theses as a percentage of the graduations reported to percentage of the graduations reported to DEST (black bars),DEST (black bars), and a ‘requirement and a ‘requirement effectiveness’ or ‘mandate potential’ (effectiveness’ or ‘mandate potential’ (greengreen).).

APSR Workshop: 4 May 2006 20

EDT chartEDT chart

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Adelai

de U

niver

sity

Austra

lian

Catho

lic U

niver

sity

Austra

lian

Natio

nal U

nivers

ity

Centra

l Que

ensla

nd U

nive

rsity

Curtin

Unive

rsity

of T

echno

logy

Deakin

Uni

vers

ity

Flinde

rs U

nive

rsity

Grif

fith U

nivers

ity

La T

robe

Uni

vers

ity

Mur

doch

Univ

ersit

y

Que

ensla

nd U

nivers

ity o

f Tec

hnolog

y

RMIT

Uni

vers

ity

South

ern

Cross

Univ

ersit

y

Swinbu

rne

Unive

rsity

of T

echno

logy

Unive

rsity

of B

allar

at

Unive

rsity

of C

anbe

rra

Unive

rsity

of M

elbou

rne

Unive

rsity

of N

ew S

outh

Wale

s

Unive

rsity

of N

ewca

stle

Unive

rsity

of Q

ueens

land

Unive

rsity

of S

outh

Aus

tralia

Unive

rsity

of S

outh

ern Q

ueen

sland

Unive

rsity

of S

ydne

y

Unive

rsity

of T

asman

ia

Unive

rsity

of T

echno

logy

Syd

ney

Unive

rsity

of W

este

rn A

ustra

lia

Unive

rsity

of W

este

rn S

ydne

y

Unive

rsity

of W

ollon

gong

Victor

ia U

niver

sity o

f Tec

hnol

ogy

Actual theses Mandate potential

APSR Workshop: 4 May 2006 21

ConclusionConclusionOnly one Australian university (QUT) has its act Only one Australian university (QUT) has its act

together in populating its OAR, and has a together in populating its OAR, and has a requirement policy, now approaching 80% requirement policy, now approaching 80% capture.capture.

Twelve Australian universities (of 39) have a Twelve Australian universities (of 39) have a requirement policy for theses. Nevertheless, requirement policy for theses. Nevertheless, the capture rate is still only 12% as (a) the capture rate is still only 12% as (a) dragged down by the rest and (b) several of dragged down by the rest and (b) several of the policies have still to bite seriously.the policies have still to bite seriously.

There is a long way to go in promoting and There is a long way to go in promoting and implementing requirement policies (aka implementing requirement policies (aka ‘mandatory policies’).‘mandatory policies’).

APSR Workshop: 4 May 2006 22

Research Quality FrameworkResearch Quality Framework• Minister has appointed a further Minister has appointed a further study study

committeecommittee (NOT an implementation committee) (NOT an implementation committee)

• Many important details remain Many important details remain unresolvedunresolved

• ProbabilityProbability: First RQF Round will be postponed : First RQF Round will be postponed to 2008 based on: 2007 census date and 2001-to 2008 based on: 2007 census date and 2001-2006 publications.2006 publications.

• At some time metrics will replace excessively At some time metrics will replace excessively expensive evaluation, maybe after 2008 or expensive evaluation, maybe after 2008 or 2012.2012.

APSR Workshop: 4 May 2006 23

RQF-OAR linkageRQF-OAR linkage– Research in OARs increase citations Research in OARs increase citations

(demonstrated) and increase generalized impact(demonstrated) and increase generalized impact

– RQF evaluation involves both citations and impact, RQF evaluation involves both citations and impact, now and in futurenow and in future

– ‘‘RQF Impact’ must be based on top four RQF Impact’ must be based on top four publicationspublications

– Panels will want to see some publicationsPanels will want to see some publications

Any Australian university concerned about RQF should Any Australian university concerned about RQF should be getting all its current research output into an OAR as be getting all its current research output into an OAR as soon as possible, and retrospectively back to 2001 as soon as possible, and retrospectively back to 2001 as possible. possible. The first RQF round will be critical. The RQF is The first RQF round will be critical. The RQF is about [rapid] change in the HE system.about [rapid] change in the HE system.

APSR Workshop: 4 May 2006 24

RAE (United Kingdom)RAE (United Kingdom)

• RQF is recent Australian equivalent of long-RQF is recent Australian equivalent of long-established RAE.established RAE.

• Next UK RAE assessment due in 2008.Next UK RAE assessment due in 2008.

• Will move to Will move to discipline-weighted metricsdiscipline-weighted metrics after after 2008 round to replace over-expensive 2008 round to replace over-expensive operation.operation.

• JISC has funded JISC has funded IRRAIRRA: RAE module add-ons : RAE module add-ons to EPrints and DSpace (2006 delivery).to EPrints and DSpace (2006 delivery).

APSR Workshop: 4 May 2006 25

UTas preparednessUTas preparedness

• Current WARP system provides IRRA-like Current WARP system provides IRRA-like functionalityfunctionality

• WARP records contain a WARP records contain a link to EPrintslink to EPrints full-text full-text record (have done for two years)record (have done for two years)

• UTas should have UTas should have ALLALL ‘four best publications’ ‘four best publications’ in EPrints (mandatory) and linked to WARP.in EPrints (mandatory) and linked to WARP.

• Since authors and best publications not known Since authors and best publications not known in advance, general mandatory system best.in advance, general mandatory system best.

APSR Workshop: 4 May 2006 26

RQF interfacingRQF interfacing

• Work in UTas has commenced on analysing Work in UTas has commenced on analysing RQF requirements against the OSS IRRA RQF requirements against the OSS IRRA module with a view to producing an EPrints-module with a view to producing an EPrints-compatible RQF module.compatible RQF module.

• Active development dependent on finalization Active development dependent on finalization of RQF requirements by Minister.of RQF requirements by Minister.

• Would provide Australia-wide RQF-compatible Would provide Australia-wide RQF-compatible repository software. UTas may or may not use.repository software. UTas may or may not use.

APSR Workshop: 4 May 2006 27

RQF SummaryRQF Summary

OAR provides for greater citations, therefore long-OAR provides for greater citations, therefore long-term better metrics.term better metrics.

OAR provides mechanism to deliver ‘four best OAR provides mechanism to deliver ‘four best publications’ to RQF panels on demand with publications’ to RQF panels on demand with minimal cost, perhaps with attached metrics in minimal cost, perhaps with attached metrics in future.future.

Add-on modules to OAR potentially provide Add-on modules to OAR potentially provide management information and other RQF input.management information and other RQF input.

APSR Workshop: 4 May 2006 28

DiscussionDiscussion

APSR Workshop: 4 May 2006 29

© Copyright 2006 Arthur Sale© Copyright 2006 Arthur Sale

All rights reservedAll rights reserved

Arthur Sale asserts the right to be Arthur Sale asserts the right to be recognized as author of this workrecognized as author of this work

Contact:Contact: Arthur.Sale@utas.edu.au Arthur.Sale@utas.edu.au