Click here to load reader
Upload
michael
View
215
Download
0
Embed Size (px)
Citation preview
Journal of DocumentationThe effect of the internet on researcher motivations, behaviour and attitudesAdrian Mulligan Michael Mabe
Article information:To cite this document:Adrian Mulligan Michael Mabe, (2011),"The effect of the internet on researcher motivations, behaviour andattitudes", Journal of Documentation, Vol. 67 Iss 2 pp. 290 - 311Permanent link to this document:http://dx.doi.org/10.1108/00220411111109485
Downloaded on: 10 October 2014, At: 03:56 (PT)References: this document contains references to 25 other documents.To copy this document: [email protected] fulltext of this document has been downloaded 1356 times since 2011*
Users who downloaded this article also downloaded:Anna Wojciechowska, (2007),"Analysis of the use of open archives in the fields of mathematics andcomputer science", OCLC Systems & Services: International digital library perspectives, Vol. 23 Iss 1pp. 54-69Suzana Sukovic, (2011),"E-Texts in Research Projects in the Humanities", Advances in Librarianship, Vol.33 pp. 131-202
Access to this document was granted through an Emerald subscription provided by 550547 []
For AuthorsIf you would like to write for this, or any other Emerald publication, then please use our Emerald forAuthors service information about how to choose which publication to write for and submission guidelinesare available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.comEmerald is a global publisher linking research and practice to the benefit of society. The companymanages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well asproviding an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committeeon Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archivepreservation.
*Related content and download information correct at time of download.
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
The effect of the internet onresearcher motivations,behaviour and attitudes
Adrian MulliganElsevier Ltd, Kidlington, UK, and
Michael MabeDepartment of Information Studies, University College London, London, UK
Abstract
Purpose – The purpose of this paper is to understand how the migration from the print world to theelectronic environment has affected the motivations, attitudes and behaviours of researchers inscholarly communication.
Design/methodology/approach – The paper takes the form of an investigation that is bothquantitative and qualitative. The study was split into three phases: understand the issues affectingresearchers (focus groups and interviews); an online survey of 6,344 researchers measuring attitudesand digging deeper into issues: telephone interviews to understand differences between differentgroups change. Differences in opinions were examined across discipline.
Findings – While there has been some change in the behaviour of researchers, there has been littlechange in their motivations for publication. Researchers want other researchers’ data but are lessinclined to share their own. Researcher attitudes towards repositories are very mixed. Researchershighly value peer review. The pressure to over-publish at the expense of quality is exaggerated.
Research limitations/implications – Further research is required to measure the impact onresearcher motivations and attitudes of external pressures that were emerging at the time of thisstudy. This includes the growing influence of funding bodies, the economic downturn and its impacton institutional budgets, as well as subsequent advances in the digital revolution.
Practical implications – This research suggests that, while technology may have positivelyimpacted the efficiency of scholarly communication, the drivers behind scholarly informationexchange remain relatively unchanged. Moreover, changes to the scholarly information businessmodel will only be successful if they continue to satisfy the underlying motivations and needs ofresearchers.
Originality/value – This paper fulfils an identified need to measure the motivations of researcherstowards the core functions of scholarly communication on a global level.
Keywords Research, Attitudes, Motivation (psychology), Individual behaviour, Peer review,Information exchange
Paper type Research paper
IntroductionChanges in researcher behaviour, motivations or influences are likely to affect thewhole value chain of scholarly communication. Wide-ranging and key decisions arebeing made about the future of scholarly communication in an environment wherethere has been little systematic research. This large-scale study, which is based on a
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0022-0418.htm
This paper will be a part of the doctoral work by Michael Mabe at the Department of InformationStudies, University College London.
JDOC67,2
290
Received 20 November 2009Revised 21 June 2010Accepted 21 June 2010
Journal of DocumentationVol. 67 No. 2, 2011pp. 290-311q Emerald Group Publishing Limited0022-0418DOI 10.1108/00220411111109485
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
number of responses greater than any previous similar survey, provides evidence tohelp inform these decisions.
The intention of this study was to identify what effects, if any, the move from theprint form of scholarly communication to the electronic form, particularly the advent ofthe world wide web has had on researchers, notably changes that have occurred overthe last decade.
Furthermore, the study seeks to place these results in the context of possible futurechanges in scholarly communication, and examine how current attitudes are likely toaffect future scholarly communication.
Previous studies in this area were conducted prior to researchers realising the fullpotential of the internet and too UK centric (Swan and Brown, 1999), or were too smallin nature and not representative of all the major disciplines (Swan and Brown, 2003).
The authors collaborated with NOP, a leading research agency in the UK, to conducta survey of over 6,000 researchers. This formed the principal stage of the research uponwhich the main findings of this paper are based. Researchers were asked to respond toa number of provocative statements, which concerned an attitude or motivation,affecting one of the four fundamental functions of scholarly publishing: registration,certification, dissemination and the archive (Merton, 1973, Ravetz, 1973, Ziman, 1968).
MethodologyThe study was split into three phases:
(1) Understand the issues affecting researchers (focus groups and interviews).
(2) An online survey of 6,000 þ researchers measuring attitudes.
(3) Digging deeper into issues, telephone interviews to understand differencesbetween different groups.
Phase 1This stage of the project was qualitative in nature and conducted in October andNovember 2004. Three focus groups in the UK were conducted (28 individualsattended) and ten in-depth telephone interviews[1] all in the USA were conducted. Thisphase identified the issues currently affecting researchers. In order to minimise costs itwas necessary to restrict the research during this qualitative phase to Englishspeaking regions (see Table I). In the groups and interviews researchers were allowedto articulate in their own terms, how they approached work, why they worked incertain ways, what was affecting them in their day-to-day lives, how their workenvironments had changed, and identify the influences acting upon them. Authors forthese interviews were randomly selected for each geographic location from a range ofjournals published by Elsevier and across different subject areas. Interviewees wererecruited via by the third party research agency, who were tasked with ensuringrepresentation across subject area, and a balance of seniority. In order to avoid bias onegroup was made up entirely of young researchers. The results of this first stage of theresearch were used to design the quantitative questionnaire.
Phase 2This stage of the research was conducted in April 2005, 6,344 publishing authorscompleted an online survey, in which they were asked to rate a series of statements.
The effect of theinternet
291
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
Foc
us
gro
up
1F
ocu
sg
rou
p2
Foc
us
gro
up
310
tele
dep
ths
Lon
don
,U
KL
ond
on,
UK
Lon
don
,U
KB
osto
n,
US
AJu
nio
rac
adem
ics
Sen
ior
acad
emic
sS
enio
rac
adem
ics
Jun
ior
and
sen
ior
acad
emic
sS
ub
ject
Cou
nt
Su
bje
ctC
oun
tS
ub
ject
Cou
nt
Su
bje
ctC
oun
t
Lif
e/M
edic
al4
Lif
e/M
edic
al6
Ph
ysi
cal
Sci
ence
s(P
hy
sics
,C
hem
istr
y)
8L
ife/
Med
ical
6
Soc
ial
Sci
ence
s2
Bu
sin
ess/
Eco
nom
ics
1P
hy
sica
lS
cien
ces
(Ph
ysi
cs,
Ch
emis
try
)4
Bu
sin
ess/
Eco
nom
ics
3P
hy
sica
lS
cien
ces
(Ph
ysi
cs,
Ch
emis
try
)3
Ph
ysi
cal
Sci
ence
s(P
hy
sics
,C
hem
istr
y)
1T
otal
10T
otal
10T
otal
8T
otal
10
Table I.Composition of focusgroup and teledepths
JDOC67,2
292
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
Audience. Approximately 30,000 authors were approached via e-mail to complete thesurvey. The authors were from variety of sources, they had published in journals fromdifferent publishers and across a range of disciplines. The types of scholarly journalalso varied; a number of authors had published in open access journals.
They were also diverse in terms of their age, country of origin (see Tables II-IV) andthe type of institute or organization they were associated with ranging from hospitalsand commercial R&D facilities to research institutes and universities. As can be seen inTable IV, the geographic distribution of authors broadly reflects the geographicdistribution of the world’s research output in terms of articles as reported in ResearchTrends (2008).
Research tool. The researchers were asked to complete a short survey that wasdesigned by the agency NOP, and took approximately ten minutes to complete. Theresearchers were asked to specify their level of agreement with a number of statements.The scale which was used to determine agreement was a five-point Likert scale (Likert,1932). In order to avoid the “halo” effect; the bias that occurs when a series of answersare influenced by responses to former statements, known as “carryover” (Churchill,1995, p. 482), some of the statements were semantically inverted. For example “Readersdo not really need refereed journals” as opposed to, “Readers do really need refereed
Field Frequency Percent
Arts and Hum and Social Sciences 724 11.4Chemistry 490 7.7Computer Sciences/IT 227 3.6Earth and Planetary Sciences 347 5.5Economics 341 5.4Engineering 565 8.9Environmental Sciences 310 4.9Fundamental Life Sciences 1,122 17.7Materials Science 405 6.4Mathematics 348 5.5Medicine and Allied Health 976 15.4Physics and Astronomy 404 6.4Other 84 1.3Total 6,343 100.0
Table II.Respondents according to
subject
Position Frequency Percent
Head of dept/professor/senior manager 2,464 38.8Senior researcher/lecturer/middle manager 1,849 29.2Researcher/staff member 1,409 22.2Graduate/postgraduate student 379 6.0Consultant 121 1.9Other 120 1.9No answer 1 0.0Total 6,343 100.0
Table III.Respondents according to
position
The effect of theinternet
293
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
journals”. Additionally, for different respondents, the order in which statementsappeared was rotated in order to avoid bias in response.
Response rate. It is difficult to be precise about response rates, as a number ofe-mails would have been stopped by spam filters, but it was good at approximately 15per cent.
Analysis: The responses were analysed according to demography, subject, age andseniority. All findings presented here have been tested for significance and arestatistically significant at 95 per cent confidence levels (p , 0:05). T-tests were used toidentify differences between populations. For convenience of understanding a red staris recorded against any sub-group in the charts, when the percentage of respondentsagreeing (or disagreeing, depending upon the question and the context) with thestatement is significantly different to the overall average for the total population(p , 0:05).
Phase 3After analysis of the quantitative findings, the third and final stage of research wascompleted. Altogether additional 70 in-depth telephone interviews were carried out inAugust and September 2005. This stage was completed by the agency in order toexplore in more detail the quantitative results. This phase was not meant to berepresentative of the broader population, but to gain qualitative insight into anyunusual figures in the quantitative analysis.
ResultsMotivations for publishingA key study into author motivations at the end of the paper-only publishing era waspublished 17 years ago (Coles, 1993). In the Coles study, which was UK-based, a number
Top 15 countries Frequency Percent Percent of research outputa
USA 1,388 21.9 25.9UK 422 6.7 7.3China 366 5.8 5.7Italy 322 5.1 3.5Germany 310 4.9 6.7Japan 252 4.0 7.4Spain 226 3.6 2.5France 211 3.3 4.8Canada 205 3.2 4.1India 188 3.0 2.2Australia 157 2.5 2.2The Netherlands 156 2.5 2.0Taiwan 149 2.3 1.2Turkey 149 2.3 n/aBrazil 131 2.1 1.2All other countries 1,711 26.9 23.3Total 6,343 100.0
Note: aGeographic distribution of research articles as reported by Research Trends in 2008 (years1997 to 2007)
Table IV.Respondents according tocountry
JDOC67,2
294
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
of motivations were identified and respondents invited to indicate which two were themost important. Figure 1 shows the results from this study in comparison with theresults from our 2005 survey.
First stated motivations can be misleading, respondents often choose a responsethat is “top of mind”, which can conveniently convey a more altruistic position thanmight truly exist, especially when the respondents are sophisticated and know thattheir answers are likely to be seen by their peers. Dissemination is the most importantmotivation in the Coles study with 57 per cent of the respondents indicating this is thereason why they publish. Analysis of secondary motivations can get beyond thissomewhat “obvious” response to “covert” motivations that are more likely to beinfluential in driving behaviour. In this context “future funding” (47 per cent) and“furthering my career” (27 per cent) in the Coles study are key differentiators. We haverepeated this approach.
In our study, respondents were invited to indicate which two motivations were themost important. The most important motivation is “dissemination” (73 per cent).“furthering my career” and “future funding” are the key secondary motivations.
By comparing the two studies we can determine whether there has beenmotivational shift in the intervening 12 years between the surveys. The two studies aregeographically distinct, so any differences identified are indicative[2]. The differencesbetween the two studies are presented in Figure 2. In terms of change to secondarymotivations, “recognition” and “establishing precedent” have clearly increased,especially the latter.
FundingAttracting funding is a key motivation for publishing (Figure 1). The study sought toexamine attitudes of researchers towards funding bodies through the questions posed
Figure 1.Motivations for publishing
The effect of theinternet
295
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
and answered in Figure 3. A total of 63 per cent (CI ^ 1:21 p , 0:05) of respondentsbelieve “Funding bodies have too much power over what research is conducted”. Thisis most acutely felt by those in Life Sciences and Chemistry (see Figure 4). However, incontrast researchers do not believe that funding bodies dominate the choice of journal,only 23 per cent (CI ^ 1:06 p , 0:05) agreed that “Research funding bodies dictatewhere research is published”.
Quality versus quantityThere is a commonly held view among researchers that too many research articles arebeing published in scholarly journals. In this study we sought to identify if researchersbelieved it was more important to publish a large number of research articles ratherthan a smaller number of higher quality papers. Figure 5 shows the results. Theseindicate that the majority (70 per cent) disagree (CI ^ 1:13 p , 0:05) with the idea thatquantity is more important than quality. This varies little by subject area, with nostatistically significant differences. Previous work (Mabe and Amin, 2002) supports the
Figure 3.Attitudes to fundingbodies
Figure 2.Change in motivationsover time
JDOC67,2
296
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
results here, and further suggests that the ratio of authors to the number of paperspublished is roughly constant and has been for the last several decades, so thatresearchers views of over-publication are more perceptual than actual.
Peer reviewAn essential mechanism that maintains the quality of research articles published in ascholarly journal is the peer review process. It has long been the corner-stone of formalscholarly communication, but in recent years has attracted considerable adverse pressas a small number of cases of fraud have been uncovered in the literature. Such is thedisquiet that it has raised concerns in a number of sectors, among them the UKgovernment (Parliamentary Office of Science and Technology, 2002) who have openly
Figure 5.Publication pressure
Figure 4.Funding bodies have too
much power over whatresearch is conducted:
subject variation
The effect of theinternet
297
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
questioned the validity of peer review. This study sought to determine whetherresearchers believed that peer review was necessary, how many thought it was usefuland how many were committed to the process.
The results (Figure 6) show that there is universal agreement that refereed journalsare important, 88 per cent (CI ^ 0:8 p , 0:05) disagreed with the test statement thatthey were not needed. There is little variation by subject. The majority believe thatpeer review improves the quality of a research article, but we do notice some subjectvariation with those in Engineering and Material Sciences being a little bit morecynical (Figure 7). The status of peer review appears secure as a large proportion, 85
Figure 7.Peer review does notimprove the quality of anarticle – total agreeing bysubject
Figure 6.Attitudes to peer review
JDOC67,2
298
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
per cent (CI ^ 0:88 p , 0:05), are actually willing to commit to peer review (Figure 6).Qualitative findings from phase 3 of the research, indicated that researchers are willingto review articles at the level of between two and 30 articles per annum.
Informal sourcesFormal sources of scholarly communication are dependent upon peer review and arethe bedrock of scholarly discourse, but it has long been recognised that the informalexchange of information plays an important role. Informal sources of communicationare typically bulletin boards or conferences and might include discussions withcolleagues in person, via e-mail or via the telephone. Our study attempted to quantifythe importance of this informal communication exchange (Figure 8). The findings showthat in an electronic environment these informal exchanges are valuable to 61 per cent(CI ^ 1:21 p , 0:05) of respondents. However, there is some subject differentiation(Figure 9). Conferences are most valued by Computer Scientists and Physicists, whouse them to gather new insights, and least valued by those in Earth Sciences andChemistry: the qualitative research indicated in these areas researchers are more boundto formal sources out of a concern that their research remains confidential.
Figure 8.Informal sources
communication
Figure 9.Informal sources of
communication such asconferences, bulletin
boards are not importantin scholarly publishing –total agreeing by subject
The effect of theinternet
299
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
CollaborationAttendance at conferences often leads to new friendships and collaborations forscholars. Traditionally collaborations over distances even of a few miles, let aloneoceans and continents, have been problematic. The advent of the internet and e-mailexchange, as well as the ability to exchange large data sets online, has facilitatedresearcher interaction – in this context, it is not surprising to learn that collaborationbetween researchers has increased (Figure 10); 53 per cent (CI ^ 1:28 p , 0:05) believeresearchers are more likely to collaborate on research projects now than they were tenyears ago. Figure 11 shows that the subject most likely to collaborate is Chemistry,while least likely is Physics (this is probably a consequence of being early adopters ofthe internet as a collaborative tool through the establishment of the pre-print serverarXiv in 1991). Qualitative research suggests that greater collaboration has largelybeen facilitated, as might be expected, by the adoption of technology.
AccessThe introduction of the electronic communication has not only facilitated greatercollaboration between researchers but also affected the way they access information.Traditionally, researchers had to go to the physical library in order to interrogate thescholarly literature. Electronic access has meant that those visits have becomeincreasingly redundant, and other studies (King and Montgomery, 2002) show a shiftin behaviour among researchers. These studies largely reflect a shift within the context
Figure 10.Collaboration in research
Figure 11.Researchers are morelikely to collaborate onresearch projects now thanfive years ago: subjectvariation
JDOC67,2
300
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
of the institute or organization, from library to faculty, department laboratory or office.An overlooked aspect of the shift to electronic publishing is the movement of access tooutside the organization/institute. Increasingly scholarly literature is being madeavailable to researchers at home, so far there is little data on how much access is fromthe home. Our results (Figure 12) show that a proportion of researchers, 22 per cent(CI ^ 1:03 p , 0:05), prefer to access the scholarly literature from their home. Thisfigure is highest among those in Medicine (Figure 13)
While it is clear that the electronic version is important our research suggests that ithas not yet taken over and a substantial proportion of respondents, 47 per cent(CI ^ 1:23 p , 0:05), disagree that an article will be read if it is only availableelectronically (Figure 12). However, those in Computer Sciences, Life Sciences and
Figure 12.Accessing journal content
Figure 13.I prefer to do my e-journal
browsing at home ratherthan at work: subject
variation
The effect of theinternet
301
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
Physics and Astronomy are more dependent upon the electronic version of the article(Figure 14).
A clear advantage of the electronic form over the print form of the research article isthe ability to publish and provide access to the supplementary data that formed part ofthe original research. Overwhelmingly, researchers indicated (Figure 15) that theywanted access to this supplementary data with 67 per cent (CI ^ 1:17 p , 0:05)agreeing that it was vital to access supplementary data such as tables, images andvideo. Those in Medicine (71 per cent) and Chemistry (72 per cent) are most keen toaccess supplementary data (Figure 16).
However, attitudes are contradictory when it comes to sharing data. Almost allresearchers want to have access to other researchers’ raw data (Figure 15), 75 per cent
Figure 14.An article will only be readif it is availableelectronically: subjectvariation
Figure 15.Author/reader dichotomy
JDOC67,2
302
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
(CI ^ 1:08 p , 0:05) a significant majority, and this is significantly higher for those inComputer Sciences (88 per cent), Maths (83 per cent), Earth Sciences (83 per cent),Engineering (83 per cent) and Materials Science (81 per cent) (Figure 17). However,fewer researchers are willing to share their own data, 52 per cent (CI ^ 1:23 p , 0:05),apart from those in Materials Science, those subjects who wish to have greater accessto raw data tend to be more likely to share their own data, albeit at a lower rate. The
Figure 16.It is vital that all
supplementary data suchas extra tables, images
and videos are publishedelectronically: subject
variation
Figure 17.Having greater access to
other researchers’ datawould benefit my own
research: subject variation
The effect of theinternet
303
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
exception to the rule are those in Economics (Figure 18) The qualitative researchrevealed a number of very tangible reasons why individuals are less likely to sharetheir own data. Principal amongst them is the desire to capitalise on their own work asmuch as possible; researchers did not want to give away their competitive advantage.Researchers also expressed concerns about how other researchers might use their rawdata, possibly misinterpreting the data. Some thought they did not own the data andwere not in a position to make it publicly available.
Versions, self-publication and repositoriesThe advent of electronic access has also meant that versions of an article notpreviously accessible in the print world are now available for viewing. To add to thecomplexity, these versions are often available from a multitude of different locations,authors can self-publish on their own web site or deposit articles in repositories.Typically issues of trust emerge with articles when there is some doubt over theirauthenticity, as the work by Nicholas et al. (2006) suggests. In this context we wishedto quantify which versions of an article were used, which were most important and,critically, how they are used.
Self-publication is not yet common place among those surveyed, 11 per cent(CI ^ 0:76 p , 0:05) claim to have published an early version of one of their own articleon their own web site, and 14 per cent (CI ^ 0:85 p , 0:05) have published the finalversion of their article on their web site. When we followed these results up in the finalqualitative stage, the in-depth telephone interviews suggest that these figures may infact be lower. Quite often researchers were self-publishing lists of articles that hadappeared in a journal, but not a version of the actual article itself.
The proportion of respondents that access articles via other researchers’ web sites ishigher at 26 per cent (CI ^ 1:03 p , 0:05). The qualitative research indicated that
Figure 18.I am willing to allow otherresearchers to access myraw research data: subjectvariation
JDOC67,2
304
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
respondents accessed their peers’ web site in order to track an individual author intheir field, and to keep abreast of the latest developments. Respondents indicated thatrarely do these visits lead to a download of information or a request for an article.
Publication on an author’s own web site is just one form of self-publication: authorscan also deposit versions in a variety of repositories. Repositories can be either an“institutional repository”, which will include a collection of scholarly materials indigital form that is typically managed by the library at a university or a “subjectrepository” which is similar, but stores materials specific to a discipline and isnormally managed by a funding body (examples include Pubmedcentral, ArXiv).
Proportionally, the number of respondents that have published in a repository, isrelatively small, for institutional repositories the figure is 8.2 per cent (CI ^ 0:68p , 0:05) and for subject repository even less 7.6 per cent (CI ^ 0:65 p , 0:05).
This study has also allowed us to estimate the extent of knowledge of repositoriesamong researchers. Overall knowledge of repositories globally is quite low (Figure 19).Only 5 per cent (CI ^ 0:53 p , 0:05) know a lot, an additional 28 per cent (CI ^ 1:1p , 0:05), indicate that they know a little. Subject repositories are slightly betterknown at 9 per cent (CI ^ 0:69 p , 0:05) with another 29 per cent (CI ^ 1:12 p , 0:05)indicating that they know a little about them.
In the qualitative follow-up stage, attitudes towards repositories were explored.While a proportion certainly felt that repositories were a good idea, as the respondentsthought they provided free access to current research, allowed them to find informationquickly, and raised the profile of the institution, a number of researchers expressedconcerns. Some questioned the purpose of repositories specifically in the context of thecurrent journal system, others worried about how they would receive credit for workpublished in such a system, and that such systems would fail to deliver the recognitionassociated with publication in a scholarly journal. Others worried about theirsustainability and speculated on whether the funding would be in place in years tocome. Finally, and perhaps most importantly for a number of researchers, they doubted
Figure 19.Awareness of repositories
The effect of theinternet
305
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
the quality control mechanisms that might be in place, and questioned who wouldmaintain the repositories.
As indicated above there are now various versions of an article availableelectronically. Authors can not only post in various locations, but also different formsof the article. We have defined essentially five versions of an article for the purposes ofresearch. The various stages and how they map onto the recently published NISOjournal article versions (NISO, 2008) are shown in Table V.
Critical to understanding how the various versions of articles function is discerninghow often they are used and identifying their relative importance among researchers.Figure 20 shows the usage and the importance of these different versions of the articlefrom our survey results. The final published article is still the most used at 80 per cent(CI ^ 0:98 p , 0:05). Critically it is also the most important version for 62 per cent(CI ^ 1:2 p , 0:05). The qualitative research showed that the main reason researchersvalue the final version so much is that they need to be able to cite it. It is the versionthey know is not going to change. Many researchers felt they would be wasting theirtime using earlier versions of the article if it is possible that it might be modified insome way.
Earlier versions of the article are still used, but to a lesser extent. The most widelyused earlier version is the corrected proof at 58 per cent (CI ^ 1:2 p , 0:05), next theauthor’s final accepted manuscript with 45 per cent (CI ^ 1:2 p , 0:05). In terms of theimportance that scholars assign to the earlier versions, the most important after thefinal version is the corrected proof at 24 per cent (CI ^ 1:05 p , 0:05). The qualitativeresearch indicated that more often than not the earlier versions are used to obtain earlyintelligence about a paper. The prepublication version of the article, which has beenpeer reviewed and often appears in a repository, is quite useful to many individuals forexchanging and generating ideas. Often respondents used them in order to avoidreplicating research that they know is already underway.
ArchiveThere is a perception in the scholarly community that researchers do not refer to olderarticles, and moreover that this attitude is more prevalent among younger researchers(Mulligan, 2005). Our research shows that this is not the case.
In order to measure the prevalence of this attitude, researchers were asked torespond to the provocative statement, “It is not important to have access to researcharticles that were published more than 10 years ago”. Universally 5 per cent ofresearchers (CI ^ 0:7 p , 0:05) agreed with the statement. The qualitative follow-upresearch revealed that researchers refer to articles that are up to 40-50 years old, oftenbecause these are the seminal works. Critically, being able to refer to older articles inthe archive ensured respondents did not repeat research unnecessarily. Economists areleast likely to agree with this statement (Figure 21).
DiscussionThis study was undertaken to investigate whether, and how, the advent of the worldwide web and wide-spread use and availability of electronic journals might change thebehaviours and attitudes of scholars. Many commentators have stated, not alwayswith strong backing, that “the internet changes everything” and that even in the
JDOC67,2
306
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
Art
icle
typ
eD
escr
ipti
onS
equ
ence
inp
ub
lica
tion
cycl
eN
ISO
jou
rnal
arti
cle
ver
sion
s(2
008)
Pre
-pu
bli
cati
ond
raft
(pre
pri
nts
)N
onp
eer-
rev
iew
ed1s
tA
uth
or’s
ver
sion
Au
thor
’sac
cep
ted
man
usc
rip
tv
ersi
onT
he
auth
or’s
ver
sion
ofth
em
anu
scri
pt
that
has
bee
nre
vis
edin
lig
ht
ofco
mm
ents
rece
ived
du
rin
gp
eer
rev
iew
,
2nd
Acc
epte
dm
anu
scri
pt
Un
corr
ecte
dp
roof
ofth
ear
ticl
eT
he
arti
cle
afte
rit
has
bee
nfo
rmat
ted
for
pu
bli
cati
on3r
dP
roof
In-p
ress
ver
sion
ofth
ear
ticl
eT
he
arti
cle
afte
rit
has
bee
nfo
rmat
ted
for
pu
bli
cati
onan
db
een
pro
ofed
4th
Pro
of
Fin
alp
ub
lish
edar
ticl
eT
he
arti
cle
afte
rit
has
bee
nfo
rmat
ted
for
pu
bli
cati
on,
pro
ofed
,al
loca
ted
toan
issu
ean
dp
agin
ated
5th
Ver
sion
ofre
cord
Table V.Different versions of theresearch article defined
The effect of theinternet
307
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
conservative world of scholarly communication huge shifts of behaviour and practicecould be expected (Gates, 1999; Arnold, 1995).
The world wide web-based electronic journal debuted in 1993 (Keyhani, 1993). By2005 when our study took place, full text electronic journal platforms such asScienceDirect, Synergy, SpringerLink and Wiley Interscience were accessed by 92 per
Figure 21.It is not important to haveaccess to research articlesthat were published morethan ten years ago: subjectvariation
Figure 20.Publication status ofdifferent versions of theresearch article
JDOC67,2
308
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
cent of the research institutions in the world (PCG, 2005, p. 17), and for many keyinstitutes had been a fixture of scholarly literature for nearly ten years.
In 2005, easy-to-use electronic journal systems were the norm and electronicworkflow solutions for researchers becoming commonplace, certainly for the scientific,technical and medical communities. In the light of this, we believe it is reasonable toposit that the responses to our survey instrument are representative of a community atearly stage digital maturity with experience in using and getting the best out ofelectronic publishing systems, and that the views expressed are more representative ofmature opinion than of initial experience.
A key aspect of our study was to compare and contrast the results obtained at earlydigital maturity with the results obtained from a major, highly respected studypublished in the last year of the paper-based scholarly communication era (Coles, 1993).One particular aspect of the Coles study, a series of questions relating to publishingmotivations of authors particularly lend themselves to analysing the extent to whichauthors still attach importance and value to the traditional drivers of publication. Inessence there has been no substantive shift in secondary motivations (which we argueabove are more likely to represent the covert or “underlying” motivators) as thecommunity moved from a fully paper-based environment to a virtually fully electronicone.
An examination of primary motivations shows that not only has “dissemination”remained the most important primary motivation but its standing has actuallyincreased. When combined with increases in the importance of the secondarymotivations “establishing precedence” and “recognition”, this suggests thatresearchers are feeling overwhelmed with the increasing quantities of researchavailable through electronic sources, and the perceived extra amount of competition,and are focusing more on maintaining their own position.
Overall, our results indicate that the fundamentals of formal scholarlycommunication steadfastly remain the same. Researchers need to publish researcharticles in order to further their own career. There is no overarching motivation toover-publish at the expense of quality. Moreover, researchers need to assure the valueof their work to secure “future funding” via the platform of peer reviewed journals.Peer review, will continue to be the lynchpin of formal scholarly communication.Articles will need to be disseminated to the appropriate people and date-stamped toensure the researcher can protect discovery rights. Finally, researchers want theirresearch to be in a permanent archive, so it can be referenced, and become the buildingblock for future science. All the intrinsic elements of the scholarly journal:dissemination, certification, registration and the archive remain present and important,even in this radically changed world of digital scholarship. Although there have notbeen any similar studies with such a large and international sample set, there havebeen some reports that echo our results. In 2007 the University of California (2007)surveyed its faculty and found very conservative behaviour patterns matching theresults we describe here. Most recently, the studies of RIN (2009) and of Harley et al.(2010) have found very similar results.
However, our study also shows that many less fundamental behaviours andopinions are in a state of flux. Clearly, in recent years there have been a number ofimprovements in the manner in which scholarly information is exchanged. The adventof the world wide web has facilitated a range of new processes, ultimately making the
The effect of theinternet
309
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
submission of manuscripts and access to research articles much easier and more timeefficient. As one of us has written elsewhere (Mabe, 2009), there are new tools but theyare for the same old purposes.
Interestingly, though, none of the advances just mentioned have yet to affect thefundamental form of the formal scholarly article. Why this should be the case iscomplex and outside the scope of this article, but is discussed by one of us in a recentpaper (Mabe, 2010).
The majority of researchers now submit their manuscripts online. Most reviewerslog-on to review a peer’s research rather than wait for it to arrive by post. Behaviour atthis level will likely continue to change and develop as novel tool sets becomeavailable. The electronic version of the journal will become even more pervasive in allfields. Researchers will continue to make more use of technology, accessing theircolleagues’ work from wherever they are: home or office. They will find novel ways ofcollaborating and sharing data, but likely only share data that will not disadvantagetheir own careers.
While the mechanisms that drive research behaviour are fundamentally unchanged,the conclusions presented here do not mean that scholarly publishing will beun-affected by technological development. New economic models are and will continueto emerge. New technologies will find a place if they allow scholars to conduct theirbusiness easily and quickly. At the fundamental level of drivers, though, there is littlediscernable change. It is clear that however the scholarly communication systemdevelops in the coming years, it will only be successful if it satisfies the needs of theresearcher and minimises the time spent on preparing materials for publication as wellas the time spent finding and retrieving articles.
Notes
1. In-depth telephone interviews typically last 20-30 minutes in which a number of complexissues are discussed in full.
2. It should be noted that there was little variation between the UK motivational data and therest of the world data in our study.
References
Arnold, K. (1995), “The body in the virtual library: rethinking scholarly communication”, Journalof Electronic Publishing, Vol. 1 Nos 1 and 2, January.
Churchill, G. (1995), Marketing Research Methodological Foundations, 6th ed., Dryden Press,Lansing, MI.
Coles, B. (1993), The Scientific, Technical and Medical Information System in the UK, BLR&DDReport 6123, Royal Society, The British Library, and the Association of Learned andProfessional Society Publishers, London.
Gates, W. (1999), Business@The Speed of Thought, Penguin, Harmondsworth.
Harley, D., Krzys Acord, S., Earl-Novell, S., Lawrence, S. and Judson, C. (2010), Final Report:Assessing the Future Landscape of Scholarly Communication: An Exploration of FacultyValues and Needs in Seven Disciplines, CSHE 1.10, Center for Studies in Higher Education,UC Berkeley, Berkeley, CA.
Keyhani, A. (1993), “The online journal of current clinical trials: an innovation in electronicjournal publishing”, Database, Vol. 16, February, pp. 14-23.
JDOC67,2
310
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
King, D.W. and Montgomery, C.H. (2002), “After migration to an electronic journal collection:impact on faculty and doctoral students”, D-Lib Magazine, Vol. 8 No. 12, p. 11.
Likert, R. (1932), “A technique for the measurement of attitudes”, Archives of Psychology, Vol. 20No. 140, pp. 1-55.
Mabe, M.A. (2009), “Scholarly publishing”, European Review, Vol. 17 No. 1, pp. 3-22.
Mabe, M.A. (2010), “Scholarly communication: the long view”, New Review of AcademicLibrarianship, Vol. 16 No. 1, pp. 132-44.
Mabe, M.A. and Amin, M. (2002), “Dr Jekyll and Mr Hyde: author-reader asymmetries inscholarly publishing”, Aslib Proceedings, Vol. 54 No. 3, pp. 149-57.
Merton, R.K. (1973), Sociology of Science, University of Chicago, Chicago, IL.
Mulligan, A. (2005), “Is peer review in crisis?”, Oral Oncology, Vol. 41 No. 2, pp. 135-41.
Nicholas, D., Jamali, H.R. and Rowlands, I. (2006), “On the tips of their tongues: authors and theirviews on scholarly publishing”, Learned Publishing, Vol. 19 No. 3, pp. 193-203.
NISO (2008), “NISO journal article versions”, available at: www.niso.org/publications/rp/RP-8-2008.pdf
Parliamentary Office of Science and Technology (2002), Postnote: Peer Review, No. 182,Parliamentary Office of Science and Technology, London, September.
PCG (2005), “PCG library budget survey for 2005, available at: www.pcgplus.com/Resources/LibraryBudgetSurvey2005.pdf
Ravetz, J.R. (1973), Scientific Knowledge, Penguin, Harmondsworth.
Research Information Network (RIN) (2009), Communicating Knowledge: How and Why UKResearchers Publish and Disseminate Their Findings, RIN, London.
Research Trends (2008), No. 8, November, available at: www.info.scopus.com/researchtrends/archive/RT8/cou_tre_8.html
Swan, A. and Brown, S. (1999), What Authors Want: The ALPSP Research Study on theMotivations and Concerns of Contributors to Learned Journals, ALPSP, Worthing, pp. 1-41.
Swan, A. and Brown, S. (2003), “Authors and electronic publishing: what authors want from thenew technology”, Learned Publishing, Vol. 16 No. 1, pp. 28-33.
University of California (2007), “Faculty attitudes and behaviors regarding scholarlycommunication: survey findings from the University of California”, available at: http://osc.universityofcalifornia.edu/responses/materials/OSC-survey-full-20070828.pdf
Ziman, J.M. (1968), Public Knowledge, Cambridge University Press, Cambridge.
Further reading
Wheeler, A.G. (1989), “The pressure to publish promotes disreputable science”, The Scientist,Vol. 3 No. 14, p. 11.
Corresponding authorAdrian Mulligan can be contacted at: [email protected]
The effect of theinternet
311
To purchase reprints of this article please e-mail: [email protected] visit our web site for further details: www.emeraldinsight.com/reprints
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)
This article has been cited by:
1. Xianjin Zha, Jinchao Zhang, Yalan Yan. 2014. Exploring the effect of individual differences on userperceptions of print and electronic resources. Library Hi Tech 32:2, 346-367. [Abstract] [Full Text] [PDF]
2. Tony Volpe, Joachim Schopfel. 2013. Dissemination of knowledge and copyright: an historical case study.Journal of Information, Communication and Ethics in Society 11:3, 144-155. [Abstract] [Full Text] [PDF]
3. Chèrifa Boukacem-Zeghmouri, Joachim SchöpfelBeyond the Google generation: towards community-specific usage patterns of scientific information 137-151. [CrossRef]
4. Michael MabeDoes journal publishing have a future? 413-440. [CrossRef]
Dow
nloa
ded
by U
NIV
ER
SIT
Y O
F H
ON
G K
ON
G A
t 03:
56 1
0 O
ctob
er 2
014
(PT
)