1

0

EARLY CAREER RESEARCHERS: THE

HARBINGERS OF CHANGE?

FinalreportfromCIBER

August2016

Yearone(2016)

2

Contents

1.0 Executivesummary 6

2.0 Background 14

3.0 WhoareECRs? 15

4.0 Aimsandhypotheses 16

5.0 Scope 17

6.0 Methodology 18

6.1 Researchmethodsandinstrument 18

6.2 ECRSample 19

6.3 Recruitment 21

6.4 Recordingandcoding 21

7.0 Literaturereview 22

8.0 Results 25

8.1 Careerandjobambitions 26

8.2 Characteristicsofscholarlycommunicationbehavior 27

8.2.1 ECRsasfollowers 27

8.2.2 Paper-drivenbehaviour 28

8.2.3 Publishingandauthorshippractices 28

8.2.3.1 ECRsasfirstauthors 29

8.2.3.2 JournalselectioninthehandsofECRs? 30

8.2.3.3 Criteriaforselection 31

8.2.3.4 Publishingstrategies 33

8.2.3.5 Experiencesofothercountries 33

8.3 Peerreview 35

8.4 Openaccess 37

8.5 Socialmediaandonlinecommunities 38

3

8.6 Discovering/findingpublications/information 39

8.7 Smartphones 39

8.8 Openscience 40

8.9 Sharingandcollaboration 40

8.10 Reputationandassessment 43

8.11 Metrics 43

8.12 Impact 44

8.13 Roleofpublishersandlibraries 45

8.14 Transformations 46

8.15 Countrycomparisons 48

8.16 Morediversity 50

9.0 Interimconclusionsandreflections 52

10.0 References 57

Appendix1:QuestionsforECRinterviews 61Appendix2:CodingtemplateforNationalReports 67Appendix3:DetailedsubjectrepresentationofECRs 70

4

TablesandFigures

Table1:Interviewersandpartnerinstitutions 17

Table2:Instrumentsusedforinterviews 19

Table3:NumbersandnationalitiesofECRsinterviewed 20

Table4:SubjectrepresentationofECRs 20

Table5:Gender,ageandstatusofECRs 21

Table6:NumberofarticlespublishedbyECRs 29

Table7:Firstauthorandmainchoice 30

Table8:Similaritiesbetweencountriesbasedonhypothesistest 48

Table9:CountryECRprofilesbroadlycompared 50

Figure1:Countryclustersbasedonhypothesistests 49

5

NoteTheprojectwas fundedbythePublishingResearchConsortiumandconductedduringtheperiodOctober2015toAugust2016byateamofresearchersfromtheUK,China,France,Malaysia,PolandandSpain.Subjecttoreview,thereportprovidestheresultsofyearoneofathree-yearproject.Inadditiontothisreport,anumberofother,moredetailedreportsareavailableontheCIBERwebsite(http://ciber-research.eu/harbingers.html):

HarbingerWorkingReport1:LiteratureReview

HarbingerWorkingReport2:HypothesesTests

HarbingerWorkingReport3:ComparativeNationalFindings

Theindividualnationalreportswillformasingledatabase,whichwillbemadeavailableonrequest.

The authors of these reports are David Nicholas, AnthonyWatkinson, Abrizah Abdullah,Chérifa Boukacem – Zeghmouri, Blanca Rodríguez Bravo,Marzena Świgoń, Jie Xu and EtiHerman.

6

1.0 Executivesummary

Context

Earlycareerresearchersareofgreatinterestnotjustbecausetheyarethenewwave,butbecausetheyarealsothebiggestwave– theyareby far the largestgroupof researchers(Jones,2014).Therefore,theymeritlong,detailedandcontinuousinvestigation.Towardsthisendthisreportprovidesthefirstyearfindings(thefoundationstone)foraproposed,novel,three-year, longitudinal study of 116 science and social science early career researchers(ECRs),whohavepublishednearly1200papersbetweenthem,comefromsevencountries(China,France,Malaysia,Poland,Spain,UKandUS)and81universities.The wide-ranging study, funded by the Publishing Research Consortium, focuses on theattitudesandbehavioursofECRsinrespecttoscholarlycommunicationsandtheextenttowhich they are adopting such potentially disruptive technologies as social media, onlinecommunities andOpen Science/Science 2.0 thatmight prove transformational. Uniquely,webelieve,ECRswereinterviewedintheirownlanguagesandbypeoplewhounderstoodthenationalcontexts,face-to-face,bySkypeortelephone,withastructuredscheduleandtypicallylasting90minutes.Thisresearchisverytimelyasthelastpieceofresearchtolookat this topic comprehensively, but from a largely UK perspective, was undertaken sevenyears ago by JISC in a very different scholarly world, one in which social media, onlinecommunities,reputationalplatformsandsmartphoneswereverymuchintheirinfancy.Qualitativemethodologies, as those deployed by the project, are best at providing deepconversations and understanding and personal insights and context, rather than makingrobust generalizations and comparisons, which is the territory of the ubiquitousquestionnaire.Nevertheless,thereisstillaneedtooffersummarizationandquantification,albeitwithqualificationandtothatendwehaveusedastructuredinterviewscheduleandcoded up the responses rigorously. Furthermore, our sample, despite our best efforts toobtainsomebalanceinselection,becauseoffundingrestraintsisonlyatinyfractionofthetotalpopulationandcannotberegardedasrepresentative,moresuggestive.Forall thesereasons we have used tables, percentages and statistical calculations selectively and,whereverpossible,allowedthe‘voices’oftheECRstocomethrough.[Sections3-6fordetail]

Mainfindings

• Careerandmotivation.ThevastmajorityofECRsinterviewedwantedtocontinuetodoresearchandhopedtomovetoapositionwheretheyhadajob(tenure)and,inthecaseofthesciences,usuallyintheirowngroup,butintheUSinparticularthereweresomewhowereeitherprovidingaresearchfunctionor(inthesocialsciences)weredoing-andintendedtocontinuedoing-atleastpartofthetimetheirownindividualresearch.Therewas clear evidence, too, of tactical thinking connected with career planning and oflookingaheadtowhattheymightdointhefutureiftheyobtainedtenureandwere,for

7

instance, inaposition tousesocialmedia to reachout toawideraudience. [Sections8.1;8.2.3.4;8.14]

• Followersorharbingers?TheanswertothemainquestiondrivingthestudyisthatECRs

donotinvariablyfollowthescholarlypracticesoftheirmentorsandseniors.True,inthecrucially important area of publishing ECRs still have little choice but to abide by theestablishedrules,atleastuntilthesearechanged,whichcouldhappenyet,withanudgefromthefunders(NicholasandHerman,2016).Indeed,itishardlysurprisingtofindthatECRs are evenmoredriven to publish in highly ranked Journal Impact Factor journalsbecause of their precarious positions and their belief that this leads to careeradvancement.However, inotherareasof their scholarlyundertakings theyareplainlymoreadventurous,ifnotalwaysintheirpractices,thenatleastintheirattitudes.Thus,forexample,theymaynotpreferpublishinginajournalwithinnovativefeatures,suchasvideoarticles,buttheyareawareoftheconceptorevenexcitedbyit.Bythesametoken,quiteafewoftheECRsusesocialmedia,ifmostlyforgettingPDFs,connectingwiththeircolleagues and, increasingly, encouraged by their institutions, to maximize researchimpact. Even those who do not tell us that they should make more use of theopportunitiespresentedandmightdosointhefuture,especially forbuildingresearchcollaborations.[Section8.2]

• Paper-drivenbehaviour.ThefourfunctionsofOldenburg’sjournal–registrationofnew

research, dissemination, peer review [certification] and archival record – are sofundamental to empirical scholarship that even in these digital times all the journalspublishedconformtoOldenburg’smodeland thenewwaveof researchers (ECRs)arestill fixated by them. Publishers today see themselves as investing in and organizingjournals to provide these functions for researchers andwe can find little evidence tosuggest they aremisguided.ECRsdance to the same reputational tune as researchershavedonefora very long time.SomeECRsdoponderonnovel researchoutputsandacknowledgetheunfairnessoftheexisting,unbalancedrewardsystem,butnotenough,orinsufficientnumbers,tofundamentallychallengethistraditionalpictureandthustoundermine the role of publication in peer reviewed journals in the short or mediumterm. However, we would be much more confident saying this after three years oflongitudinalresearch.

• Publishing practices. ECRs are more productive than is sometimes assumed, having

published around 10 papers each (and at least double that if conference proceedingsandbookchaptersareincluded)andasmentionedtheyareverydriventopublishinginhighlyrankedJIFjournals.Publishingoutletsinsomecountriestendtobeveryprescribed,withECRshavingtorefertolistsofacceptablejournals.Inmostcases,itisaproprietarylist(normally,WebofScience),butsometimesit isagovernmentlist,asinthecaseofPoland,althoughthese listsarebuiltonthefoundationsofproprietaryones.China, infact,operatesbothlists,withtheGovernmentlistusefulforfieldswhereitisdifficultfor

8

researcherstopublish inWebofScience journalsandasameanstopromoteChinesejournals.ThedominantinfluenceoftheWebofScienceisparticularlymarked,butnotwithmedicalresearcherswherePubMedinclusion,arguablyalowerbar,isimportant.ForECRstobeacknowledgedasfirstauthoris,onthewhole,notthatdifficultandtheyare typically first author in one-third to one-half of all the papers to which theycontribute.WheretopublishisgenerallyagroupdecisionandECRsdohaveaninfluencewith ECRs in the UK and US claiming to have considerable influence on the decisionwheretosubmitin,respectively,25%and30%ofallinstances.Ofcourse,iftheresearchcannotgetpublishedinatopjournal-andthereisalwaysatensionbetweenthewishtoget into a very top journal and thewish tobemorepragmatic forwhatever reason -therehavetobeother criteriaand these include submitting to journals: a)where thechancesofacceptancearehigher;b)wheretheyhavehadgoodexperiencesinthepast;c)whichprovidearapidturnaround,referredtoas‘quickjournals’;d)thatarethoroughand efficient (and give lots of helpful feedback); e)which have themost appropriateaudiences;andf)whichcaterforopenaccess(seeundertheOpenaccessSection8.4formoreonthis).

ECRswereaskedwhethertheyhadalongtermpublishingstrategy.Notsurprisingly,formany this was publishing in high impact factor journals. US researchers have thispressure, but seemingly less so than their colleagues inother countries, although thismight be partly explained by the fact that the US sample wasmore established andconsequently were less driven to publish in high impact factor journals. There was afeelinginsomecountriesthatJIFsweregoingtobemoreratherthanlessimportantinthefuture.[Section8.2.3]

• Peer review. Another interesting fact, surprising perhaps, is that there is no strong

evidence to support thewidely held belief, which is by now almost a truism, that theexistingpeerreviewsystemis'aclosedclub',fromwhichECRs(interalia)arelockedout.Somehaveheardof such things, but not really experienced them. Indeed, unawareorperhaps simply unimpressed by the perceived shortcomings of the system, at leastaccording to the literature, most express an overall, if hedged, satisfaction with peerreviewasitis.Infact,theyblameitsinadequacies(badlychosen,bad/biasedreviewers)ontheeditors,whoinanycasearethoughttohavetoomuchpower.Nevertheless,ECRsare of twomindsabout openpeer reviewas apossible alternative. Theyappreciate itstransparency, but do not believe it can work in practice and worry about lettingundesirablepeople into the system– something FrenchECRs are anxious about–as itwas thought this it would make itmore difficult to reject papers, which will be moredetrimentalforECRs.[Section8.3]

• Socialmediaandonlinecommunities.Asmentionedalready,therearepatchesofsocialmediaandonlinecommunityuseandthesepatchesarebiggerthanwehavewitnessedinourprevious investigations. ResearchGate (possibly the fastest grower in the field),

9

LinkedIn and Twitter are the tools of choice. Finding information, communicatinginformation,sharing,buildingadigitalprofile/presence,obtainingPDFsandengaginginoutreachactivitiesarethemainusestowhichtheseplatformsareput.Thisconstitutesquiteascholarlylist,butactivecollaborationisanotableabsentee.Socialmediahaveafirmfoothold,especiallyinChinaandMalaysia.[Section8.5]

• Smartphones.Givensociety’swidespreaduseofsmartphones,thefactthatsmartphones

are the main platform for connecting to the Internet, and that, like the rest of thepopulation,manyECRs themselveshave smartphones, itmight comeasa surprise forourresearchtoreporttohowlittleECRs(admit)toutilizingthemforscholarlypurposes.Evenwhentheyaresaidtobeused, it ismostlyoccasionally,forcommunicatingwhileawayfromtheoffice,travellingandatconferencesorforalerts,ratherthanforreadingandmarkingpapers.Ofcourse,socialmediaandsmartphonesgohandinhand,sotheincreases in the former that we have already detected look likely to result in similarincreases in the latter.Unsuitability for reading is claimed tobe themain reason,andthat of course is a valid reason, but the samewas said about laptops and desktops.Resolution is getting better all the time and screens larger, and research publishedelsewhereindicatesthatacademicsnowreadmoreHTMLwebpagesontheirtablets,e-readersorsmartphonescomparedtotwo yearsago (Haleviet al., 2015). The Chineseseemtobeleadingthesmartphonechange/charge,willothersfollow?Thisissomethingtowatchforthefuture.[Section8.7]

• Openaccess.Goldopenaccess(OAjournals)isuniversallythoughttobeagoodthing,but

ECRsarewellawareoftheproblemsassociatedwithopenaccessjournals.Openaccess(OA) is not really an issue, never mind a big issue, although there is some disquietregardingarticlepublicationcharges(APCs)whicharethoughttobetoohighandunfairbecause they are making the playing field uneven between those researchers withaccesstofundsthatcanpayforAPCsandthosethatdonot.Thereisalotlessdistrustofopenaccessthanwasencounteredinearlierstudiesontrust,butfewECRsarequeuinguptobepublishedinOAjournals.PublishinginOAjournalsisgenerallynotpartofanypublishingstrategy,despitethemandates(mainlyintheUKandUS)thatECRsare(onlyvaguely)awareof.[Sections8.2.3.3;8.4]

• Repositories.Somewhatsurprisingly,sinceECRsmighttobethoughttobeinterestedin

taking every opportunity to showcase their achievements, they regard archiving theirresearchworkinrepositoriesasanon-priority;ifundertakenatallitisthoughttobeamatterforlibrariansorresearchadministration.Archivingisdonewhenobligatory,butwithoutmuch enthusiasm; somuch so, that there is a general absenceof knowledgeaboutandinterestinrepositories,totheextentthatasignificantnumberofECRsdonotevenknowthattheirinstitutionhasaninstitutionalrepository.Thisisunlikelytochangequickly unless archiving in repositories obtains reputational credit. ECRs are howevermorelikelytodeposittothematicrepositoriessuchasarXiv.org,whicharesupportedby

10

the wider scientific community. The community is much more important than theinstitutioninthisregard.ECRsarealsowiseenoughtoknowthatresearchersingeneraldo not think of searching in repositories and that community networks such asResearchGateofferaneasierwayoffindingcontent,sowhynotdepositthere?[Section8.4]

• Openscience.Thereismuchtalkabouttheopenagendaintheprofessionalpressandat

conferences;however,ECRsdisplaylittleunderstandingoforinterestinOpenScience,Web2.0etal.and its technologiesaspossibleagents forchange inscholarlypractice.Indeed, French researchers are antagonistic to the concept, seeing it as a furtherrestraint on their scholarly freedoms. But related questions about open data andsoftware(componentsofOpenScience)didstirsomeinterestamongafewUKandUSresearchers. The sort of "pain free publishing" (https://elifesciences.org/about) thateLifepromisesalsomeritedsome interest,becauseof thepreoccupationofECRswithpublication.Theyarenotsointerested insharingdata(an importantelementofOpenScience)becausemanywanttoexploitthedatatheyhavegatheredtothefull(fortheirpublications),andnotgive itaway.TheOpenagenda includesblogsasnon-traditionalscholarlyoutputs,butnoone,certainlyintheUK/US,isreallyinterestedinblogsasanalternativetopublications.Again, thegamechangermightbegivingECRsreputationalcreditforsuchactivities.Tenureandpromotioncommitteeshaveasmuch, ifnotmore,influenceonresearcherpracticesthanfundermandates.[Section8.8]

• Sharingandcollaboration.SharingiseasiertodothaneversincetheemergenceofsocialmediaandonlinescholarlycommunityplatformssuchasResearchGate,andsharing isthoughttobeareputationaccruingactivity,whichmightbeexpectedtotakesciencetogreaterheights(Nicholasetal.,2015).WhilethelargemajorityofECRsshareideasandinterimdata,muchof thisactually takesplaceat the researchgroup level, at internalmeetingsandwithinlocalnetworks.WhilesharingismuchmentionedbyECRsascentraltothewaytheywanttolivetheirscholarlylives,and,perhaps,theyarealittleconflictedwhentheyhavetoactbytheacademic'rules',thesharingof ideasandinterimresultsusing socialmedia is littleundertaken. Sharing researchoutputs ‘afterpublication’ viaResearchGate in particular is a different matter and is a popular activity, especiallyamongUKresearchers.

Collaboration is clearly aweightier issue and the keyhypothesiswe testedwas:Earlycareer researchers share and collaborate extensively even at the risk of losing theircompetitiveedge. In fact, therewasnocountryconsensusherewith justonecountry,France, fully confirming that this is the case and three other countries partiallyconfirming it. For French ECRs, despite eschewing social media for this purpose,collaborationisclearlyking.Besidespublications,collaboration isaconstantobjective.ThestrategiesofECRsforgettinga jobandpublishingmoreandbetterpapersrelyoncollaboration.Conferencesandmeetingsarekeymoments,dedicatedtosearching forcollaborations. ECRs believe that they can be hired for their CV, but also for the

11

potentialoftheircollaborations.AdifferentpictureemergesfromtheUKandUSwhereECRshavetheirnetworkswithwhomtheyinteractoutsidetheirgroups,buttheredoesnotseemtobemuchevidenceofformal research collaboration. It tends to be ratherbasic and piecemeal, providing help among friends, giving feedback, linksandadvice.Nevertheless,while there is nobroad consensus as to the (presumed) value of socialmedia inbuilding research collaborations, there is still activity and interest. This is anarea where things are a little confused and unclear, and possibly waters are gettingmuddier as a consequence of socialmedia and scholarly collaboration platforms. Thislandscapewillneedclosemonitoringoverthenexttwoyears.[Section8.9]

• Metrics.Despitetheimportanceaccordedtometricsasa(future)fundamentalelementofreputationalassessments,ECRsdemonstratelittleinterestinusagemetrics(knownas'altmetrics'), although some do check their publication downloads. This is only to beexpected, of course, for altmetrics are not yet widely used and accepted either byresearchers or by the university system. However, some ECRs, but not in the UK/US,tendtoagreethataltmetricsisapotentialnewmethodtoevaluateresearchers’outputandinfluence.Nevertheless,ourstudyshowsthatthehypeassociatedwithaltmetricsisnotmatchedbyECRinterest.Itisstillveryearlydaysforaltmetricsandthisisanotherlandscapetowatch.[Section8.11]

• Impact. Most ECRs see that conducting good research and getting it published in

prestigiousjournalsisthewaytoinfluenceothersandtohaveanimpact,butUKECRs,undoubtedly influencedbythemetricsof theResearchExcellenceFrameworkusedbyHEFCEtoassessprioritiesfordispersingfunds,demonstrateawiderinterestinreachingouttothegeneralpublicandusinginnovativemeans(includingsocialmedia) todoso.[Section8.12]

• Publishers and libraries. There are mixedmessages for publishers and bad news for

libraries,thetwomainpillarsofthetraditionalscholarlycommunicationsystem.Despitepossessionofthereputationaldiamonds inthemine,thehighlyranked journals,mostECR views about “commercial” publishers are negative, although notmany ECRs hadviews on particular publishers, and in general they demonstrated a lack ofunderstanding of what publishers (or libraries) do. This, taken together with the factthatECRsdonotchoosetopublishonthebasisofthepublisherbutoftheirjournal,anddonotappeartousepublisherwebsites(oftenpreferringfreeandopenservices,suchasarXiv.organdResearchGate)thereisthechallengeforpublishersoflookingworryinglyanonymous and unpopular in transitional times. On the positive side, ECRs arecomfortable,butnotnecessarilyhappy,withpublishersmanagingthepeerreviewprocessand nor are they happy with learned societies doing it, largely because societies arethoughtnottobesufficiently independent.Perceivedindependenceseemstobearealpositiveforpublishers.[Section8.13]

12

Muchmoreworryingly, librariesseemtohave lostall theirvisibility.LotsofECRshavenot gone to the library for years. Libraries are mainly considered as places forundergraduates to sit and work. Their discovery systems have been bypassed byGoogle to alargeextentandtomakemattersworsetheirinstitutionalrepositoriesarenot popular either. Libraries appear to have little to offer to the big new wave ofresearchers, so down the line there have to beworries for their long term future asresourcesforpostdocs.Andthis,ofcourse,posessomechallengesforpublishersastheyhavelongworkedhandinglovewithlibraries.[Section8.14]

• Diversity.Clearlywehavetobecarefulinmakingcomparisonsatsuchanearlystagein

the project’s life and in dealing with quite a heterogeneous dataset. However, it ispossibletopointtoareasthatneedtobemonitored:

a. Country. The fact that the project is organized by country means that there is a

premisethattherewouldbenationaldifferences.Infact,wefoundrealdifferencesand similarities. There is clearly a UK/US special relationship, EU countries sort-ofcluster, but there is no 'Asian' cluster, with China closer to EU than toMalaysia.Malaysia,infact,seemstobeacontrarian.Someofthesedifferencescouldbeputdowntothedifferentmake-upofthenationalECRsamples.

b. Scholarly communication experience. ECRs who have reviewing experience holddifferent scholarly views from thosewho do not, perhaps because they aremorefamiliarwiththesystemandcantalkaboutitmorefully.Theyarealsomoredefensiveofasystemofwhichtheyfeeltheyareapart.

c. Loneresearchers.Thereisadifferencebetweenthosewhoworkmoreorlessontheirown, usually doing a doctorate after preliminary experience, and those who areembeddedingroups.Theformertendtobesocialscientistsandasageneralizationtheyprovidedfeweranswerstothequestionsaskedandarelessproductive.Indeed,mostofthem(thoughnotall)arebasicallyuninterestedinscholarlycommunicationandmoreofthemareprobablynotgoingtocontinueinacademiclife.

d. Prestigious research groups. Those who work in prestigious research groups feelmore secure about their prospects and tend to be happier with the academiccommunicationprocess,perhaps,justbecausetheyaremoreoptimisticabouttheirfuture.

e. Subjectofresearch.Someresearchtopicscanbemoreorless'bankable'thanothers.Sometopicsaremoretransient.Theconsequenceisthatthosewhohave'bankable'researchsubjectsaremorevisible,theirresultsaremorelikelytobepublished,aremorecontactedbycolleaguesintheircountriesandabroad.Itisakindof‘Matthew’effect(Merton,1968).

f. Age and experience are clearly correlated and are added values for ECRs, whichcontribute towards a deeper understanding of the system and knowing how tobehaveandwhattodoinmanysituationsandcontexts.

g. Gender.Generally,thereislittleevidenceofdifferencesbetweengendersintheway

13

ECRs see career progression (or anything else for that matter), which might besurprisinggiventheviewsofsomecommentatorsabouttheproblemsofwomen(not)breaking through the glass ceiling. Thus, Sugimoto et al. (2013) find genderimbalancespersistinresearchoutputworldwide:mendominatescientificproductioninnearlyeverycountry;globally,womenaccountforfewerthan30%ofauthorshipsof collaborativepapers,whereasmen represent slightlymore than70%; for everyarticlewithafemalefirstauthor,therearenearlytwoarticlesfirst-authoredbymen;and,arguablymostimportantly,whenawomanwasintheprominentauthorposition(sole, first or last authorship), thepaper attracted fewer citations than in cases inwhichamanwasinoneoftheseroles.Givenallthis,itisquitesurprisingtofindthatChinesewomenaretheonlyonestoadmittoatleastsomegenderdisparityamongECRs, saying that they are driven more by pure subject interest than their male,promotion-drivencounterparts.

h. Serviceandappliedresearchers.Thoseresearcherswhowork inaservicecapacity,usuallyinmedicine,offerexpertiseintechniquesandmethods.Theirattitudesshowdifferenceswiththosewhoseresearchispurerandlessapplied.ThoseECRswhoworkeither in industry or in government or medical laboratories where the nature ofresearch is different are cut off from someof the concernsof theAcademy: theirattitudesareinevitablydifferent,too.[Sections8.15;8.16]

• Transformations and transitions. On the broad front, independent of discipline or

nationality,our resultsshowclearlythetensionsthatoccur inaworld intransition. Inthis transition, there are signs that scholarly ‘things’ (practices, behaviours,representations, wishes, objectives) are moving in many directions while the formalframeofevaluationandcompetition isstrengthening,almostunbending.Someoftheapparentcontradictoryresultsweseeintheresearcharedowntothesetensions.ECRsseetheopportunitiestochange,butdonottaketheopportunitytodosobecausetheyjust do not have the time and space in an insecure and busy environment. They ofcourse also have limited scope to change as they (and their tenured colleagues) areconstrainedbyareputationalsystemthatpromotes,aboveallelse,publicationrecordandcitationscores.

Nevertheless,we seem to havemoved on from the situationwe found in a previousresearch (Watkinson et al., 2015), where no one had any ideas at all about change,never mind transformation, and those who disliked the present situation just railedagainst it. Three years on we do find ideas for change and even some fortransformation, mainly moving away from the current preoccupation with publishedpapers. Researchers, who happen to be ECRs, are thinking about change andtransformationand theseare topyoung researchers. Someevenaccept the idea thattheymightchangethingswheninapositiontodosomethingaboutit.Socialmediauseisclearlyupand,ifnotquiteatthetippingpoint,itiscreatingwaves.[Section8.14]

14

2.0Background

ECRsasthenewwave,notsurprisingly,attractalotofattentionandmuchhotair.Theytendtobeafavouritetopicofconferences,seminarsandblogs(see,forexample,Jones,2014;Poli,2016), where ECRs are often produced as scholarly ‘exhibits’ to pour over. However,surprisingly,thisinteresthasnottranslatedintomuchinthewayofrobustresearchprojects.Thus,anextensiveliteraturereviewshowsthatnothingsubstantialhasbeenundertakenforsevenyears,anaeoninthedigitalageinwhichwefindourselves.Theclosestanyonehascometowhat this researchproposes todowasa JISCstudyconducted in2009,The livesandtechnologiesofearlycareerresearchers(Jamesetal.,2009).However,itwasverymuchtechnology/toolsusagefocused,settingoutasitdid“toexaminethewaysinwhichcurrentor recent doctoral-level researchers use (or do not use) ICT to support their researchactivities”. It is obviously verymuch dated as ICTs are now endemic and really predatesmuchoftheScience2.0developmentsandtherealgrowthinsocialscholarlynetworksandemergingreputationalplatforms.Thusitfound,forexample,that72%ofECRsdidnotuseWeb2.0orsocialmediatosharetheirresearch.Giventherapidpaceofchangeinthisarea(e.g.,smartphones,openscienceandonlinecommunities)itishighlyunlikelythat the JISCfindingsstillholdtrue.

Therehavebeensomemajorstudieswhichhaveinvestigatedyoungresearchers,aspartofabroaderstudylookingattheresearchpopulationasawhole,toseehowdifferentorsimilartheyare.CIBERstudiesonsocialmediause(NicholasandRowlands,2011;Rowlandsetal.,2011) and trustworthiness (Nicholas et al., 2014; Nicholas et al., 2015a; Nicholas et al.,2015b;Tenopiretal.,2015)areinthiscamp.

Itseemsfromtheresearchthattherearecurrentlytwocontrastingassumptionsaboutthebehaviour of ECRs. On the one hand, they are carrying through the new attitudescharacteristic of digital natives into their research careers, which may eventually bringaboutfundamentalchangesintheirbehaviour,too.These,intheirturn,couldresultinthecollapseofthewholecurrentjournalsystem(Laine,2015).Others,ontheotherhand,haveobserved the way in which early career researchers have recognized their position asapprenticesandtheirrelianceontheguidanceofmentors,whichtendstomakethemmoreconservativeandlessadventurousthanestablishedresearchers(Jones,2014;Harleyetal.,2010).

TheCIBERresearch,whichwehavereferredto,hasnotconcentratedwhollyontheattitudesand practices of ECRs. ECRs have only been part, and sometimes a small part, of thepopulationsstudied.However,theresearchsuggeststhatthetruthliessomewherebetweenthesetwosuppositions.Yes,traditionalbehavioursdominate,buttheseedsofchangearethere. It is alsomuchmore complicated than this and younger researchers differ in theirbeliefs/behavioursaccordingto,forinstance,discipline,nationality,roleintheresearchgroupandgender.

15

3.0 WhoareECRs(andwhyaretheysoimportant)?

Therearedifferentand conflictingdefinitionsofearly career researchers circulating (Poli,2016) and they vary from country to country, which is very important of course for aninternational survey such as this one. After an extensive trawl of the literature andconsultationswithafocusgroupformedofpublishersandourinternationalpartners(senioracademics),thefollowingdefinitionwasagreedon:

Researcherswhoaregenerallynotolderthan35,whoeitherhavereceivedtheirdoctorate and are currently in a research position or have been in researchpositionsbutarecurrentlydoingadoctorate.Inneithercasearetheyresearchersin established or tenured positions. In the case of academics, they are non-facultyresearchemployeesoftheuniversity.

TheageoftheECRandwhethertoincludePhDstudentswerethemainissuesthathadtoberesolved in establishing a definition. Discussions with national partners resulted in theadoptionofahigheragelimitthaninitiallyenvisagedbecausewewereinformedthatECRsaregettingolderasa resultof thedecline in jobopportunities (becauseof theeconomicrecession),greatercompetitionandtheraisingofthebarfortenuredposts.Initially,doctoralstudentsweregoingtobeexcludedonthebasisthattheyareadifferentscholarlyanimal(studentsratherthanstaff),howeveritsoonbecameclearthattherewasaneedformoreflexibilityasagoodnumberofECRsarehybridsandaredoingaPhDatthesametimeorhaveundertakenresearchbeforedoingtheirPhD.Why,then,areECRssuchan interestingand importantcommunitytoresearch?Forthesereasons:

1. Mostimportantly,perhaps,becausetheyaremostlyrelativelyyoung.TheyconstitutetheGenerationY/Millennialgeneration(born1982-1994)andtheyrepresentthe‘newwave’ of researchers, born digital or long conditioned by living in a digitalenvironment.Theyrepresentthefutureandweareallfascinatedbythefuture.Theyconstitutethebreedinggroundfortomorrow’sestablishedresearchers.Theycouldbe the harbingers of change, ushering all things new. However, if our sample isanything to goby, theyarenot that young,more likely tobe in their thirties thantwenties(Table5).

2. Thereareagreatmanyofthem.ItisclaimedthatECRsconstitutethebiggestgroupofresearchers (Jones,2014)andhence theyarenot just thenewwave,butalso the‘big’ wave. They are also growing rapidly in number. There are more and moreresearchers who are in some respects (economically and in status terms) earlycareer researchers,becausetheyarenotestablished in tenured positions.ThishasalwaysbeenthecaseinEurope(notsomuchintheUK),butitismoreandmorethecaseelsewhere,too.Thus,forexample,asabroadguidetheproportionofpart-time

16

staffamountsto50percentintheUK,84percentinBrazil(ShinandCummings,2013)and51percent intheUS (Weir, 2011). It is worth noting however that there is abottleneckinplacescausedbytheoverproductionofPhDsandtherelativescarcityoffacultyposts,Polandbeingagoodexample.

3. Theyareresearchersstillmakingtheirway,mostoftenintheAcademy.Theyareatastatus passage from the apprentice to the colleague state of their career in theirscientific communities, which hinges on the crucially important transition fromdependenttoindependentresearch(LaudelandGlaser,2008).

4. They are non-established researchers at the start of their careers, seeking apermanent job. Ina fewcountries, themajorityof thoseemployedatauniversitysubsequenttotheawardofadoctoraldegreearelikelytobepromotedeventuallyto a senior academicposition,while inother countries thismightbe true foronlyaboutonetenth(TeichlerandCummings,2015).

5. Theyworkinverycompetitive,selectiveandprecariousenvironments(Belluzetal.,2016). Early career employment is characterizedbymovesbetween institutions, astate of affairs prevalent in Latin and North American and continental Europeancountries(BennionandLocke,2010).Thisishardlysurprising,forthepart-timeand/orcontract-based, non-tenure track is becoming widely adopted in many countries'highereducationsystems(TeichlerandCummings,2015).TheseshorttermcontractsmeanthatECRsareunderconsiderablepressuretofast-tracktheirdevelopmentinorder to obtain scholarly reputation, whilst maintaining the delicate balancebetween mutual support from peers, and competition for funding, jobs, andpublications(Jamesetal.,2009;Müller,2014a;2014b).

6. Theyareanunstablecommunity.Commentatorstellusthattheyoungergenerationsarenot‘stable’andtheynaturallywanttousenewanddifferentplatformstotheirparents,andmaybethisextendstotheirmanagersandprofessorsaswell(Chudziak,2015).

7. ECRsareatargetformanypublishers,whodevelopdedicatedservicesforthem.Fortheseservicestobeeffective,theyneedtobebasedonrobustevidenceastoECRs'idiosyncraticpracticesandspecificneeds.

4.0Aimsandhypotheses

The principal aim of the project is to study the evolving scholarly communicationbehaviours andattitudesof early career researchers inorder todeterminewhether theyare the harbingers of change, utilizing for instance Science 2.0 developments. From thisstandpoint, the study concentrates on the key scholarly activities of information use,information seeking, citing, publishing, peer review, sharing/collaborating and reputationbuilding.Allofthiswithaspecialfocusontheimpactofopenaccesspublishing,thesocialmedia, online social networks and emerging reputation mechanisms on these activities.Additionally,wewereespeciallytaskedtoinvestigatehowECRsgoaboutselectingthejournal

17

inwhichtheypublish,somethingwhichpublishersareveryinterestedin.

Adirectcomparativeelementwith tenured/mature researcherscannotbemadebecausethey are not included in the study. However, indirect comparisons will be made by: a)askingECRsthemselveswhethertheirbehavioursandattitudesdifferfromtheirseniors;b)evaluatingthepublishedresearchonsenior,tenuredresearchers.

A secondary aim is to investigatewhether there is any evidence of differences betweenECRs,especiallyinrespectofcountry,discipline,gender,age/experience,typeofinstitution(e.g., research intensive) andaccording to their role in the research group towhich theybelong.Becauseof thequalitativenatureof the study this canonlybeundertaken in anexploratory, albeit, considered manner and only after ‘following’ researchers for a fewyears.

Toprovidetheprojectwithshapeanddirectionandtoavoidproducingtiredoldplatitudes,a series of hypotheses were generated about how ECRs are thought to utilize scholarlyresearchandcommunicationandinterviewquestionswereshapedaroundthem.TheycanbefoundlistedinAppendix1.

5.0Scope

ThemainfocusofthestudyisECRsinthesciencesandsocialsciences,whichiswherethefunders' (PRC)main priorities lie and alsowhere the vastmajority of ECRs operate. Thestudyaimedalso toobtain awide geographical reach, as the funderswanted to supportresearch on issues facing the publishing industry globally. Balancing the need forrepresentativeness (with regard to size, importance, level of development and language)withfunderinterestsandtheavailabilityofinterviewersontheground,thefollowingsevencountrieswereselected:UK,USA,China,Malaysia,Poland,SpainandFrance.SeeTable1foralistoftheinstitutionsandindividualsthatcollaborated

Table1:Interviewersandpartnerinstitutions

Country Interviewer InstitutionChina JieXu WuhanUniversityFrance ChérifaBoukacem-Zeghmouri UniversitéClaudeBernardLyon1Malaysia AbrizahAbdullah UniversityofMalayaPoland MarzenaŚwigoń UniversityofWarmiaandMazuryin

OlsztynSpain BlancaRodríguezBravo UniversityofLeonUK AnthonyWatkinson CIBERResearchUSA AnthonyWatkinson CIBERResearch

18

6.0Methodology

6.1 Researchmethodandinstrument

A key feature of the study is that, uniquely, it aims to be a longitudinal, three-yearinvestigation,askingthesameECRsthesamequestionseachyear(duringJanuarytoMarch)inordertomapattitudesandbehaviourandidentifyanychangestothem.Thiswasthoughttobeessentialastheproject isfundamentallyaboutchange,whetherthingsarechangingandatwhatspeed.Structuredinterviewswereused,ratherthanquestionnaires,giventhecomplexities, nuances and uncertainties of the subject being studied and the need toestablishapersonallinkwithECRsinordertoobtaintheirfullco-operation.Questionswereessentially compulsory and many required a yes, no, don’t know response. Nationalinterviewersweregiventhechoiceofdoingtheinterviewsface-to-faceorremotely(Skypeortelephone). Travel distances and fitting in with the busy schedules of ECRs meant thatinterviewscouldnotbeallface-to-face.ThevariousinstrumentsusedareshowninTable2.Interviewerswhousedmorethanonemethoddidnotreportonanydifferencesbetweenmethodsintermsofdatayield,butpersonalinterviewstendedtorunonforlonger.Adetailedinterviewschedulewascompiledandsenttointervieweesaheadoftheinterview(Appendix1).Thestructureandscopeoftheinterviewandthenatureofthequestionstobeusedwereinformedbytwofocusgroupsheldpriortothestartofinterviewing,onewithpublishersandtheotherwithECRsrecruitedthroughtheaidoftheaforesaidpublishers.Theinterviewschedulecovered12mainsubjectsandforeachsubjecttherewascommonlyfiveormorequestions,meaning thereweremore than60questions inall. Youcouldnotaskthatmany inaquestionnaire.Questionswerequitedetailed inorder tomake iteasier tomakeyear-by-yearcomparisons.Notsurprisingly,theinterviewsgenerallytookbetween60–90minutes.Finally,astatistical(correlation)analysiswasconductedinordertoobtainabetterideaofhowsimilarordifferentcountriesareandwhetherclusterscouldbeidentified,whichwouldmeritcloserinspection.Thiswaslargelyundertakenforvisualizationpurposesandahealthwarninghastobeposted.Thisisbecause,whileallcountrieswereprovidedwithquotasinrespect to the subject and demographic backgrounds of interviewees, there were theinevitabledifferences,mostnotablyinrespecttoage,subjectspecializationandnumberofECRsinterviewed.Thepowerofthisanalysisshouldbecomemoreobviousaswerepeatinitensuingyears.

19

Table2:Instrumentsusedforinterviews

Country MethodPoland Facetoface;TelephoneSpain Telephone;SkypeFrance FacetofaceMalaysia FacetofaceChina Facetoface;SkypeUK Telephone;SkypeUSA Telephone;Skype

6.2 ECRsampleTheprojectwassufficientlyresourcedtofollow100ECRs.Anticipatingwastageastheprojectproceededoverthethreeyears,morethanthatnumberofECRswererecruitedand,asaresult,weendedupwith116ECRs(Table3).Inreachingthisnumber,interviewersforthevariouscountriesweregivenaroughrecruitmentquotaof20-29fortheUKandUS(thelargernumberbeingareflectionoftheimportanceofthesecommunitiestopublishers)and10-15forthe other countries.Within this broadnumber the general guidancewas to build thesample in these ways: a) two-thirds science and one-third social sciences (to reflect thelargernumbers of ECRs in science); b) a representative balance ofmenandwomen; c) arange of ages within the twenties and thirties age groups; d) and, if possible, to includeresearchersfromamixtureofuniversitiesandresearchgrouptypesinregardtostanding.Obviously, with the relatively small numbers involved, the prescribed balance could notalwaysbeachieved.Tables4and5givetheoutcomes.ECRscomefrom81institutionsandthe variation between countries is explained by themethod of recruitment (see below).Generally, publisher-based recruitment generated a more heterogeneous sample. WhatTables4and5show is that therearemoremen in thesample (mainlybecausetherearejustmoreofthem,especiallyinthesciences),thesampleisgenerallyskewedtowardsthesciences (seeAppendix3formoredetails)andthatthereisabigcountryvariationinECRsstudyingforaPhD.Perhapsmost surprisingly, the number of ECRs in their thirties (generally 35 or under) isrelatively high. There are two reasons for this. First, tough economic circumstances andcompetitionmeanthatresearchersareECRsformuchlonger,becausetherearejustnotthetenuredjobsoutthere.Ittakesmuchlongertoclimbtheladder.ThereisanassumptioninmostoftheparticipatingcountriesthatanECRwillhavetodoatleasttwopostdocsbeforeeven being considered for a tenured job. Second, younger researchers did not putthemselves forward in numbers because they might have felt they lacked sufficientexperienceorfeltinsecurebecauseofthenatureofthequestioning(e.g.howmanyarticleshaveyoupublished?).InthecaseofMalaysia,wherealltheECRsareintheirthirties,ECRswill have had to have completed their PhDs first and only thosewith PhDs are hired byresearch-intensiveuniversities(asimilarsituationexists inChina).Theywouldhavebegun

20

their PhD at the age of 28-30 (eligible for government scholarship because they haveworkingexperience),spentthreetofiveyearsonit,socompletingattheageof31-35,andonlythenstarttheircareerasresearchers.

Table3:NumbersandnationalitiesofECRsinterviewed

Country No.

Poland 10(8.6%)

Spain 18(15.5%)

France 14(12.1%)

Malaysia 12(10.3%)

China 13(11.2%)

UK 21(18.1%)

USA 28(24.1%)

Total 116(100%)

Table4:SubjectrepresentationofECRs

Subject Total %

Biology&agriculture 25 21.6

Medicineandhealth 15 12.9

Engineeringandtechnology 12 10.3

Chemistry 10 8.6

Computerscience 10 8.6

Physics 8 6.9

Psychology 6 5.2

Othersocialsciences 22 19.0

Othersciences 8 6.9

Sciences 88 75.8

Socialsciences 28 24.2

21

Table5:Gender,ageandstatusofECRs

Females Males

N % N %

Gender 49 42 67 58

Twenties Thirties

Age 36 31 80 69

Doctoral Postdoc

Doctoral/Postdoc 28 24 88 76

6.3 Recruitment

Recruitment was undertaken in a number of different ways and this was because ofconvenienceandnationalpreferencesastowhatwasthebestwaytoensuremaximumco-operationandcompliance.Thebasicmethodsweretoenlistpublisherhelpingettingintouchwiththeirauthorsresidentinthecountriescovered(UK,USA,Spain)andtouseuniversityandresearchernetworks (Poland,France,Malaysia,China). Insomecases, thesemethodsweresupplementedbypersonalcontacts,workshopattendancesandbytheECRsthemselves(the invitationsgoingviral).Usingpublishersnaturallyattractedahigher institutioncount.Given the length and complexity of the questions it was, surprisingly, not as difficult asanticipatedtorecruitECRs.Avoucherworth£50ortheforeigncurrencyequivalentprovedanaddedattraction.

BecauseofthecomplexitiesofECRidentificationandtheneedtoensurethatvolunteersmetourdefinitionalrequirements,weaskedthosewhocameforwardtosendustheirCV.TheCVwasalsousefulinsupplementingandprovidingcontextfortheinterviewquestions.

6.4 Recordingandcoding

Interviewsweregenerallyconductedbynationalinterviewersintheirlocallanguage.Thiswasinordertoobtainmaximumco-operationandcomplianceandbuildarelationshipthatcouldlastthreeyears.Theproceedingsoftheinterviewsweretakendowninnoteform,asitwasfelt that the subjectsof thequestionswere toopersonal to recordduring thesessions.Atranscriptoftheinterviewwasreturnedtotheintervieweeforvalidationandfurtherdatacollectingpurposes,whichwasnecessarytoplugtheinevitablegapsintheinterviewrecord.TherecordwasthentranslatedintoEnglishforallnon-Englishspeakingcountries(butnotMalaysiawherethiswasnotnecessarybecauseofwidespreadproficiencyinEnglish)andthenmanually coded up using a heuristic approach and a standardized thematic framework.

22

Becauseoflanguagedifferencesandpossiblemisunderstandingscodingwasveryprescriptiveanddetailed(Appendix2).Finally,forECRs,yesornoresponsesdonotalwayswork.Theirresponses tend tobeusually 'yes,but'or 'no,but'.That is theadvantageof interviewing,becausequiteanumberaddsomethingotherthanyesornoandthiswouldnototherwisebecaptured.

7.0Literaturereview

Ashasalreadybeennoted,nooneappearstohaveundertakenthetypeofstudyconductedhere,theclosestanybodyhascometodoingsobeingtheaforementioned,nowdatedJISCstudy(Jamesetal.,2009).Piecingtogetherdatafrommorerecentstudies,ourknowledgeofECRsisasfollows(ourfullliteraturereviewcanbefoundinHarbingerWorkingReport1,availableat:http://ciber-research.eu/harbingers.html).

ECRsareconservativeintheirscholarlybehaviour

ECRs have been repeatedly found to be particularly conservative in their attitudes andbehaviours, tending to toe the line and foregoing the possibility of acting upon anyrevolutionary thoughts that theymighthaveabout the current systemat leastuntil theirpositionstabilizes(Fransman,2014;Housewrightetal.,2013;Jamesetal.,2009;Jones,2014;Nicholasetal.,2015a;Watkinsonetal.,2016).Theirsafestcareerbetisoptingfor'thetriedand true' in their scholarly undertakings. Indeed, although today's novice researchers areplainlycognizantoftheneedfor,andeventheadvantagesofalternativeoratleastadditionalways andmeansof conducting research (Nicholas et al., 2015d), they tend to steadfastlyadheretothelong-establishedscholarlystandardsandprinciplesofresearchwork,modelingtheirbehaviouronthoseoftheirmentors(Harley,2010;Housewrightetal.,2013;Jamesetal.,2009;Nicholasetal.,2015c;Tenopiretal.,2010;2011;Watkinsonetal.,2016).

As longas thedictatesof theacademic rewardsystem in relationtoemployment, tenureandpromotionfocusexclusivelyonthevolumeofpaperspublishedinhigh-rankingjournalsand the number of citations they obtain (Housewright et al., 2013;Mulligan andMabe,2011;Mulliganetal.,2013;Nicholasetal.,2015b,2015c;VanDalenandHenkens,2012),itis only prudent for ECRs to abide by traditional values, principles and practices. Theirposition as apprentices, coupled with their understandable reliance on the help andguidance of their mentors on the way to becoming fully independent scholars(Brechelmacheretal.,2015;Cusick,2015;Foote,2010;FriesenhahnandBeaudry,2014;Guetal.,2011),alsospeakagainsttheirstrayingfromthewell-troddenacademicpaths.

ECRshaveaone-trackmind

Müller's(2014a,2014b)findingssuggest,underscoringtheearlierevidenceaccumulatedonthesubject,thatECRsperceiveinvestinginthereproductiveaspectsofacademiclabour,suchas education-oriented activities, i.e. teaching, supervising and mentoring students, as

23

hinderingratherthanpropellingforwardtheircareers.Indeed,howelsecanayoungpersonaspiring toanacademiccareerbehave in the faceof theadviceconsistentlygiventopre-tenurescholars,ascitedbyHarleyetal.(2010):"...focusonpublishingintherightvenuesandavoidspendingtoomuchtimeonpublicengagement,committeework,writingop-edpieces,developingwebsites,blogging,andothernon-traditionalformsofelectronicdissemination(includingcourseware)".

However,asnotedelsewhere(Nicholasetal.,2015d),thisisoneaspectoftheacademicworldthatmightchangeyet,asitcanbearguedthatitrunscountertotoday’schangingsocietalpriorities,whichseethefuture intheglobalizedknowledgesocietyashingingnotonlyonresearch and innovation, but also on education for all. Also, the emerging paradigms ofScience2.0,withitscollaboration-centred,web-basedsocio-technicalsystems(Shneiderman,2008) and open, increasingly democratized, practices of scholarship (Veletsianos andKimmons,2012),bothcall forandenabletakingamuchmorewide-ranging, inclusiveandrepresentativeviewofscholarlyachievement.

Authorshipiscomingearlier

Withresearchuniversallyheldtobetheprincipalprofessionalendeavourandfocalpointofthescholarlyenterpriseandtheyardstickbywhichscholarlysuccessismeasured(Nicholasetal.,2015c),itscentralityisconveyedearlyonthewayaspartandparcelofthesocializationofnewcomerstotheworldofscholarship.Infact,asSinclairetal.(2014)concludefromtheirreviewofanumberofpertinentstudies,producingpublicationsisincreasinglyexpectedasearlyasduringdoctoralcandidatureandcompletingdoctorateswithsomepublicationsarebetterplacedforfutureemployment,includingresearchemployment.Nosurprisinglythen,in their study of the stability and longevity of the publication careers of US doctoraterecipientsWaaijeretal.(2016)findthatthetimeofdoctoralrecipients’firstpublicationhasshiftedfromafterthePhDtoseveralyearsbeforethePhDinfourofthefivefieldstheylookedat.Withgood reason, too,as the findingsofHortaandSantos (2015) indicate:publishingduringPhDstudiesleadstogreaterresearchproductivityandvisibilityinthelongrun.

Socialmediacentrictheyarenot

ECRs conservativeness is perhaps best exemplified by young researchers' uptake ofinnovative, social media based platforms, techniques and metrics for publishing andevaluationpurposes.Astheyoungarecommonlyheldtobe 'tech-savvy'andpreoccupiedwiththesocialmedia,ECRsmightbeexpectedtobeamongthemoreenthusiasticproponentsofparticipatoryandsocialwaysofresearchwork.However,youngacademicsdonotseemtobe any keener to employ novel, socialmedia basedmethods and tools than their seniorcounterparts;rathertothecontraryattimes.Indeed,asHarleyetal.(2010)pointout,acrosstheboarditisinfactpost-tenurescholarsthatarepushingtheboundaries,muchmorethantheiryoungercolleagues,sincetheyhavealreadyearnedtenureandarethereforelessrisk-averseintheirresearchandpublishingpractices.

24

Thus,forexample,astudyintoresearchers'perceptionsanduseofWeb2.0.(Procteretal.,2010;RIN,2010)showedthathighusageforproducing,sharingandcommentingonscholarlycontentwaspositivelyassociatedwitholderagegroupsandthoseinmoreseniorpositions,butthedifferencesbetweentheage-groupswererelativelysmall.ThesefindingsareborneoutbyCIBER’sstudyintosocialmediauseintheresearchworkflow,conductedayearlater(NicholasandRowlands,2011;Rowlandsetal.,2011).Theagedistributionofresearchusersofeachoftheeightsocialmediatoolsexaminedfailedtoindicateanygeneraloverallpatternanda crystal cleardistinctionbetween juniorand senior researchers.By the same token,Tenopiretal.(2013)foundnorelationshipbetweenageandcreationoruseofsocialmediaother thanblogs,RSS feeds,andTwitter; in thecaseof the latterhigh-frequencyusersorcreatorsweremorelikelytobeage50oryounger.

Thingsmightbechangingthough

There are some reasons to think ECRs will introduce the attitudes and technical facilitycharacteristicofdigitalnativesintotheirresearchcareersandthismayeventuallybringaboutchangesintheirbehaviour.AsGrahametal.(2014)contend,today'sECRisanewbreedofscholar:nolongertheindividualizedresearcher,butratheraconnectedandcommunicativeknowledgebroker,translatingbetweendifferentworldsofacademy,communityandoftenalsopolicyorgeneralpublic.Withscholarsnotonlyincreasinglyvisibleonthewebandsocialmedia(Bar-Ilanetal.,2012),butalsousingsocialmediaatallpointsoftheresearchlifecycle,fromidentifyingresearchopportunitiestodisseminatingfindingsat theend(NicholasandRowlands, 2011; Rowlands et al., 2011), novel, real time, social web based methods ofworkingshowpotentialforbecominganecessarycomplementtothetraditionalones.Also,theirmorepositive viewsof openaccesspublications (Jameset al., 2009;Nicholas et al.,2015a;Watkinsonetal.,2016)alsoseemtoindicatethattheyarebasicallymoreliberalintheir professional choices, as long as thesedonotharm their futureprospects. True, therealities of ECR life, as they emerge from the literature, seem to indicate that theirconservativenesswillpersistuntilhiring,tenureandpromotionrequirementsinacademearechanged and expanded to include novel ways of disseminating and measuring scholarlyachievement.However,asalreadynoted,developmentsinthesedirectionsseemtobequiteconceivable,ifnotnecessarilyimminent.

25

8.0Results

WithonehundredandsixteenECRsfromsevencountriesbeinginterviewedforuptotwohourseachalargeamountofqualitativedata(around170hoursoftranscripts)wasproduced.Addintheextradatathatwasobtainedasaresultofreturningtranscriptstointervieweesforcommentandclarification,thecontextualdataobtainedfromtheCVseachECRfurnished,andtheprojecthasproducedaveritablemountainofqualitativedata.Itisnotjustthevolumeofdatathatisimpressive,thereisalsoitsoriginalityastherehasbeennomajorstudylikeitinrecenttimes.

Becauseofitsqualitativenaturetheevidencepresentedisrichindepth,detail,explanationanddiversityand,assuch,notsuitableforstatisticalanalysisatthisstage.Therefore,wehavebeen careful about generalizing and cumulating the data, and expressing results inpercentages,whichwould give datamore precision than deserved. There is also anotherreason for being careful - preliminary analysis shows sizeable differences between thenationalresultsandhencewehaveoftenexpressedfindingsintermsofcountryconsensus.Given that theUKandUS accounted forwell over 40%of ECRs and that there aremanysimilarities between the two countries, this meant we had a reasonably large andhomogenous sub-set about which we could generalize and, as a consequence, we haveundertakenmoreanalyseswiththeUK/USdataset.

Designedasalongitudinalstudy,withtheinvestigationofchangeatitsheart,thefullfruitsoftheprojectcanonlybedeliveredafteryear-by-yearcomparisonshavebeenundertaken,twoyearsdowntheline.Unlessyouaregoingtosecond-guesschange(assomanystudiesdo),thisistheonlywayyoucanmakerobuststatementsaboutchange.Thefirstyearfindings,then,laydownthefoundationstonesforthefullstudy,providingaframework,anoverview,snapshotsandhighlightsofthefirstyear’sdata.Itidentifiesareaswherethereisconsensus,differencesandopportunitiesforgeneralizingthedataandidentifiestopicstowatch.

Thisreportispartofafamilyofreports,andtheotherreportsshouldbeconsultedformoredetail,specifically:

a) HarbingerWorking Report 2, which provides the results of a hypotheses test foreach partner country. It identifieswhich of the two dozen hypotheses driving thestudywereconfirmedandobtainedthegreatestconsensus.

b) HarbingerWorkingReport3,whichprovides theresultsofacanvasof thepartnercountries regarding theirmain findings.The interview teamwereaskedwhat theyconsideredtobethemainfindingsfortheircountriesandthenoncewehadthese,theywerepresentedtothewholegrouptoseeiftheyobtainedwidersupport.

Theresultsofthesevennationalstudieswillbemadeavailableinautumn2016availableasHarbinger Working Report 4 and entered on a database in order to make year by yearanalysesshouldtheprojectproceedtothesecondstage.

26

The report focuses on 16 key areas of scholarly communication: 1) Careers and jobambitions; 2) Characteristicsof scholarly communicationbehaviour, featuring an in-depthinvestigationofauthorshipand publishing;3)Peerreview;4)Openaccess;5)Socialmediaandonlinecommunities;6)Discovering/findingpublications/information;7)Smartphones;8)Openscience;9)Sharingandcollaboration;10)Reputationandassessment;11)Metrics;12) Impact; 13) Role of publishers and libraries; 14) Transformations; 15) Nationalcomparisons;16)Diversity.

8.1 Careersandjobambitions

MostECRsaredrivenbytheirownsubject interests.Theyregardanacademiclifeastheirambitionandarewhollyfocuseduponbecomingtenuredacademics.Atthesametime,theycomplainaboutlowincome,lowstatusjobsandheavyburdens.Inspiteoftheircomplaints,no one really wants to quit. The strong impression conveyed is that they love research;indeed,theyopenlyprofessthis.Moneyisclearlynotasimportantasreputationandprestige.For Chinese and Spanish ECRs, flexibility of working is a big attraction, too. Freedom todeveloptheircareer,albeitonewithoutsecurity,isanotherattractionformanyresearchers.WhileMalaysian ECRs complain that they are forced to think about survival and buildingacademic reputation, instead of focusing on contributing to knowledge/science, for themmeaningful progress in research, nevertheless, has to come frompeoplewho care aboutscienceratherthanthosewhocareaboutsuccessintheircareers.

Generally,UKandUSdoctoralstudentswanttogeta job inacademeandmostpostdocsnearingtheendoftheirtimewanttogettenureideally,buttheyareallveryrealisticandeven the best see that theymight have to spend time in industry. There are, of course,some who are in industry already and mostly they want to continue in industry. MostbelievethataPhDisapassporttojobmobility,althoughforChineseresearchers,aPhDisnot sufficient to start an academic career, a post-doc or overseas degree is also needed.Thereisevidence,too,thatnotonlyisthereisahighbartogettinganECRjobinChina(andelsewhere),butthatbarisgettinghigher.Thus,topuniversitiesinChinarecruitpeoplewithinternationally top university doctorates. Although evaluation policies are thought to berelativelyfairandgenerallyclear,ECRsbelievethatthestandardsarechallenginglyhighanditisdifficulttomeettherequirements.

ThereisabeliefamongECRsthattheirfutureisverydependentontheimportanceoftheresearchgrouptowhichtheybelong.Itiswidelyfeltthattherearenodifferencesbetweengendersinthewaytheyseecareerprogression,althoughChinesefemaleECRsappeartobedrivenmorebypuresubjectinterestthantheirmale,promotion-drivencounterparts.Thisisatopicweshallreturnto.

Sevenhypothesesweretestedinrespecttocareers.Inorderofstrengthofagreement,theywere:

• The environment in which they work is precarious. This was conclusively anduniversally agreed on. There are, then, few doubts that the ECR environment isprecarious.

27

• ECRsdomanyjobsforshortperiodsoftime.ThishypothesiswasconfirmedorpartlyconfirmedbyECRsinmostcountries,withtheexceptionoftheUKandMalaysia.

• ECRs are not very happy with their lot as research ‘apprentices’ or ‘slaves’. Thishypothesis was widely supported, with only Malaysian and American ECRs notsupportingit.SpanishECRs,forinstance,seethemselvesasscholarlyslavesandthattheir status will only change if more positions for young researchers are offered,which,inturn,hingesonasorelyneededincreaseinresearchinvestment.

• ECRshave littlepersonalfreedomandsecurity.There isageneralfeelingthatECRshaveplentyofpersonalfreedominregardtocareerdevelopment,butnosecurity.Hence, fivecountriespartly confirmed thehypothesis.Theonecountry thatmadean unqualified confirmation, Spain, is a place where conditions are particularlyprecarious.

• ECRsdomanythingsonaproject(multi-taskers).Thiswasconfirmedbyjustoverhalfthe(four)countrieswiththeexceptionsbeingtheUK,USandMalaysia.

• There isabigdrop-out rateamongECRs.Thishypothesisobtainedarelatively lowlevel of agreement,with only China and France the exceptions. This supports theearlierfindingthatECRsareverycommittedtotheirjobsandwilldoanythingtokeepit.

• Getting a good job is themajor motivation, not changing the world/science. Fivecountrieseitherconfirmedorpartlyconfirmedthehypothesis,whichprobably isareflectiononECRsprecariouspositions.ChinaandtheUKrejectedthehypothesis.

MalaysianECRsareverymuchtheoddonesoutandthiscanbepartlyascribedtothefactthattheytendtobeolderandoperateinamuchmorestableenvironment.

8.2 Characteristicsofscholarlycommunicationbehaviour

8.2.1 ECRsasfollowersThe literature review found that ECRs follow the scholarly practicesof theirmentors andseniorsandweranwiththisasahypothesis.However,weonlyfoundthistobepartlythecase,withjusttwocountries(PolandandtheUSA)fullyconfirmingthepractice.ItissaidofPolishECRsthattheyhavegreatrespectfortheirseniorsandtheirpracticesandsofollowthem. Regarding the US, while not all ECRs knew about the scholarly communicationpracticesoftheirmentors,advisorsandsupervisors,theirassumptionisthatthepracticesoftheirseniorcounterpartsaremuchthesameastheirown,except,possibly,inregardtosocial media and sharing, something which we shall return to later. In France, as in anumberofcountries,thesituation ismuchmorenuancedthanthis,with ‘colleagues’whohaveequalCVs,ageandstanding,butwhoaretenuredandfulfillsupervisorrolesorserveasmembersofrecruitmentcommissions,beingabiginfluence.Chinaisattheotherextreme,withECRsbelievingthatwhilethegroupswhichtheyusedtoworkforhadagreatinfluenceon their currentwork and determined their research orientation, in general they did notadoptthepracticesoftheirmentorsinscholarlycommunication.

28

However,anotherhypothesistested,thatECRstoetheline(dowhattheyaretold)whenitcomes to publishing, obtainedmore universal agreement, being either fully supportedorpartlysupportedbysevencountries.Wecanprobablyascribethemorewidespreadsupportobtained by this hypothesis compared to the previously cited one that ECRs adopt thepracticesoftheirmentorsandheadsofgroupstowhichtheybelongtothefactthatthereislessroomformanoeuvrewhenitcomestothecriticaltaskofpublishing.

8.2.2 Paper-drivenbehaviour

Publishingandcitingarealmostexclusivelyfocusedonjournalarticles.Reading,albeitstilldominatedbypapers,issomewhatlessprescribedwithmorethanadozentypesofmaterialmentioned, includingsocialmediaandpolicydocuments.Disseminationisalso lesspaper-centricwithconferencesvyingforpopularityandsocialmediaandonlinecommunitiesalsobeing frequently mentioned. Conferences do count more in certain disciplines, such asComputerScienceandPhysics,butevenherepapersarestillverymuchking.ThisisallsimplyexplainedbythemethodbywhichECRsareevaluated,whichfocuseslargelyonpublishinginjournals.

8.2.3 Publishingandauthorshippractices

Thefunders (PRC)requestedustopayspecialattentiontothe journalchoicesECRsmakewhenpublishing their research.Clearly this is a topic very close topublishers’hearts andeconomichealth,andifthereareanychangesinpracticesthismightchallengepublishersthemost, and theynaturallywouldwant tobe the first ones to know. Thequestionwasnotaddressed directly, instead information was pieced together from a number of relatedquestions.Mostof thequestionsusedcame fromSection5 (Authorship)of the interviewschedule,butdatahavealsobeensourcedfromelsewhereintheinterviewschedule.TheCVtoohasbeeninvaluablehere,especiallybecauseECRs,understandablybecausethatishowtheyarejudged,tendtoexaggeratetheirindependenceandreputationasauthors.

ThetopiciscomplicatedbythefactthatECRsdonothaveanunfetteredchoice,becausetheyarenotalwaysthemainauthorand/orcorresponding author,areboxed inbyassessmentproceduresthatstronglyfavourpublishinginhighimpactfactorjournalsandinfluencedbyother members of the group to which they belong and, of course, theirsupervisors/mentors.ThequestforauthorshipdatamakestheassumptionthatECRspublishinjournalsandthatthisistheirmainmethodfordisseminatingresearchresults;this,aswehaveheard, is indeedmainly, if not exclusively, true. Thus, taking our ECR sample, theypublishedanimpressive

1,178 journalarticles.Therearebigvariationsbetween individualECRsandcountries(seeTable 6),with a few individuals publishingmore than 50, butmost publishingmore thaneightarticles(anddoublethatinthecaseofMalaysia).TherelativelylownumbersreportedforPolishECRs isexplainedby the fact thatmonographpublishing ispopularandequallyrewarded.Therefore,overall,thehypothesisthatECRsarenotveryproductive(intermsofpublications)wasnotproven.

29

Table6:NumberofarticlespublishedbyECRs

No.ofarticles China France Malaysia Poland Spain UK USA0–5 5 8 3 6 5 10 116–10 4 5 3 2 5 3 1011–15 3 1 2 2 2 5 416+ 1 - 4 - 6 3 3Totalno.ofarticles

107 133 193 48 248 191 258

No.ofarticlesperECR

8.2 9.5 16.1 4.8 13.8 9.1 8.9

ECRs' publishing practices are also constrained by the fact that in a number of countriestheyhavetorefertolistsofacceptablejournals.AcaseinpointisPoland,whereECRsareformally directed towards a Government list and appear to have limited room formanoeuvre.ThisisthecasetooinChina,France,MalaysiaandSpain,buttheirliststendtobetheproprietaryones,suchasthejournalsindexedbytheWebofScience.IntheUK/USECRs appear to have more freedom, but that is still only relative.Of course, our primeinterestistoseeifthisisallchangingandweshallnotknowthatuntilwequestiontheECRsagain,thenextyearandtheyearafter.

Overall,thefindingspointtothepossibilitythatECRsareevenmoredriventopublishinhighlyrankedJIFjournalsbecauseoftheirprecariouspositionsandthebeliefthatitisthisthatleadstocareeradvancementandsecurity.

8.2.3.1 ECRsasfirstauthors

It seems, perhaps surprisingly, that for ECRs to be first author is, on thewhole, not thatdifficult.Itisnormalforthemtobefirstauthoronanyscholarlyoutputs,includingpapers,based on their dissertation. Most Principal Investigators (PIs) allow postdocs to be firstauthorswhentheyaretheoneswhohaveundertakenmostoftheresearchbehindapaper.Somepostdocshaveexperiencederivedfromtheirpastwork,whichmaybeatechniqueoranareaofexpertisethatprovidesthemwithknowledgeofjournalsinaparticularfieldandputstheminagoodpositiontochoose.Theymakethedecisions,butitverymuch“dependsonthesubject”.Thereareotherspecialcircumstances–“Idecidewheretosubmit,exceptwhereaspecialissueisinvolved”.ItcanalsodependoncountryandChinaisaninterestingand complex case. Thus, in China, most universities acknowledge that graduatestudents/postdocs/PhDcandidatescanbefirstauthors,butonlyiftheyputtheirsupervisors’namebeforetheirs.So,ECRsmightbethefirstauthorwhodidthemainresearch,buttheirnameisplacedsecond.Inmostcases,ECRswilldothistopleasetheirmentorsortomakethepapereasiertobeacceptedbecausetheirmentorsaremoreinfluentialandwellknown,thuswillhelpthemtogetpublished.

30

Table7showsthatECRsare,typically,firstauthorinaroundone-thirdtoone-halfofallthepaperstowhichtheycontribute,butitcanvarybetween0-100%.InFrancetheproportionisgenerallyat thehigherend.WhileECRsare firstauthor in thecaseswheretheydothemostworkonanarticle theydonothave themainchoiceas towhere topublishas thatdecisionisverymuchthatoftheirsupervisororheadofresearchteam.FrenchECRsappearnottomind,astheyseetheirmentors/supervisorshelpingthemtopublishintopjournals,andmainlyasafirstauthor. Indeed,theyregardthepracticeasakindof“help”thatmaycompensateforthefactthatthereisnootherspecialtreatmentfromtheinstitutionregardingtheirprecarioussituationintheuniversity.

Table7:Firstauthorandmainchoice

CH FR ML PO SP UK US%asfirstauthorforallpaperscontributedto

26-50% 51-75% 26-50% 26-50% 26-50% 26-75% 26-50%

Journalchoiceforpapersmainlyresponsiblefor

Yes No Yes Yes Yes Yes Yes

ECRshaveproblemswith co-authorship, sometimes seriousones, concerninguncongenialpractices.Interviewersfromfiveofthesevencountriesfeltthistobethecase.FranceandPoland,wherepracticesarewelldefined,aretheoddonesout.InChinathereareparticularco-authorship difficulties, with sharing credit among authors being a particularly thornyproblemindomesticcollaborations.ThisisbecausemanyChineseECRswanttobethefirstauthor,orthecorrespondingauthor,sothattheirworkcanberecognizedbytheirinstitutesandfunders.

8.2.3.2 JournalselectioninthehandsofECRs?

In order to providemore details onwhose choice it really iswhen it comes towhere topublish,thecriteriaareusedinselectionandwhetherECRshaveapublishingstrategy,weshall concentrateespeciallyon theUK/USAdata. This is becauseof the largenumbersofECRsthatcomefromthesecountries(43%ofallECRs)andbecausetheirsimilaritiesprovideuswith a relatively homogeneous group, so that it is easier tomake generalizations. Thequestionwasasked:Whatinfluence(ifany)haveyouhadonthechoiceofjournal?IntheUKoneECRdidnotanswerthequestioninarelevantwayandintheUSAoneintervieweehadnothingtosayastheyareonlyjustwritingtheirfirstpaper,soourevidencerefersto20and27ECRs,respectively.AnimportantdeterminantastohowECRsansweredthequestioniswhethertheyarepartofagroupornotandforthegreatmajorityintheUSAandthemajorityin the UK ECRs are party to group decisions. In the group they do have some influence,whichseemstovaryagreatdeal,onwhatjournaltheysubmittointhefirstplaceandhowtheychoose thebackup,althoughwehave lessevidenceon thebackup.WhereECRsarenotinagroup,forexampleinthecaseofmanysocialscientists,theydomakedecisionson

31

theirown,butnotalwaysas therearesupervisorsandmentors toconsult. In thecaseofmultidisciplinaryresearch,thechoicegoesmoretothepersoninwhosedisciplinetheywishtopublishtheresearch.AlargegroupofECRsinbothcountriesdoclaimtohaveconsiderableinfluenceonadecisionwheretosubmit:25%intheUKand30%intheUS.Attheotherendofthescale,aquarterofUKECRssaytheyhaveverylittleinfluenceandthesameistrueforUSECRs.AroundthesameproportionareECRswhohavesomeinfluence,but(forexample)thePIalwaysdecides:thefiguresare30%intheUKand30%intheUS.Thisalllooksverydefinite,butinpracticethereisacontinuumandthejudgementsmadebytheinterviewerareofnecessitysubjective.Itisalsodifficult todisentangle the influence ingroupdecisions,evenwhen theyare the firstauthor(muchgreater influence),rathermoresowhentheyare justoneoftheauthors.AUK zoologist told us, “Always take advice from people who know best, even when firstauthor”.TheabovestatisticsarebasedontheECRswhosaidtheywerefirstauthor.

8.2.3.3 Criteriaforselection

Whenmakingaselection,areECRsoptingforimpactfactor,prestigeorthesame,trustworthyjournals? Impact factor is, by far,mentioned themost and seemingly this is set in stone.However,itisnotassimpleasthat,becausetherealwayshastobeaplanBiftheresearchcannotgetpublishedintopjournals,forwhateverreason,andaplanforpaperswhicharealreadyacknowledgedasbeinglessimportant(andtherearealwaysplentyofthese).Infact,thereisoftenatensionbetweenawishtogetintotheverytopjournalandtheneedtobemorepragmatic.Eveninprestigiousresearchgroups,ECRsareonlyexpected,forinstance,topublishonepaperoutofeverythreeorfourinatopjournal.Muchoftheresearch,therefore,inevitablygoestolessimportantjournals.ManyECRsemphasizethatitistheresearch itselfandwhatittellsusthatisimportantandnotanyonepublication.

Some researchgroups in theUKandUS (ChinaandSpainaswell) try to get into the toprankedjournalsandthengodowntherankedlistifrejected,whichastheycommonlypointoutleadstodelaysandthisisbadfortheall-importantcareerprogressionandgettingnewjobs.Sometimes,too,asmanyofthemnote,thereisatrade-offbetweenimportanceandspeed,butthisisinthecaseofthesocialsciences–ineconomics,forexample,itisnotsoimportant.Othercriteria (andpossibleplanBs)mentionedare:goodchancesofacceptance, familiarterritory,a‘quickjournal’andefficientjournal.ThereareasmallnumberofECRs–lessthanthree inboththeUSandUK-whothink largely intermsofnumberofpublicationsandasimilarsmallnumberwho(still)aimforsecondorthirdrankjournalsthatarejustrightfortheiraudience.Thesearepeoplewhoare,tosomeextent,outoftheratraceandcandowhattheylike.Theyseemtobehappyeithertobeina'service'capacityortoenjoyworkinginaless demanding environment in a less important university, where teaching is a bigcomponent.

32

It is interestingtofindthat, forECRs,ranking intheUK/USWebofScience(WoS) ismoreimportantthanScopus(asitisinPolandandChina,whereScopuscountsverylittle),butthereisnotmuchdiscussionintheUKorUSaboutthis.CertainlynotasmuchasinMalaysia,whereMalaysianECRsinthesciencesdonothavecomplaintsabouttherequirementthattheyhavetopublishinWoSindexedjournals,butsocialscienceECRsdo.ForbothUSandUKECRsinthemedicalsciences,beingindexedbyPubMedisquiteimportant.

Whataboutopenaccess(OA)asacriterionusedbyECRswhendecidingwheretopublish?This,ofcourse,mightverywellbeadecisionimposedonthem,althoughintheUSitisonlynowthatfundershavebeguntoadoptmandates,whereasthishasbeenthenormintheUKforawhile.Also,universitiesintheUSmayhavemandates,butthesearenotverysuccessfulonthewhole,aremoregreenOAthangold,andtheydonothavetheResearchExcellenceFrameworktodealwith. IntheUK,11(52%)ECRssaidtheyareawareofapolicypositivetowardsOA,usuallyatuniversitylevel,andsomeofthese(3)areawareoftheneedtoputpapers into institutional repositories as (part of) this policy (one mentioned “green”).However, theunawaregroupmaskedanumber (6)whoare very aware that the fundersrequiredOApublishingandtwooftheotherssaiditdidnotaffectthemastheydidnothaveResearchCouncilgrants.Twomentionedthat“chemistsdidnotlikethispolicy”.Twosaidthattheyprefer topublishOAanyway. So,whatwas the impactondecisionsaboutwhere topublish?Twomentionedthattheypublishin'hybrid'journalsasaresult.TherewasmentionofgrouppreferencesinthreeresponsesandtheywerepositivetowardsOA.

In the USA the picture looks very different with 23 (82%) of the ECRs saying that theiruniversityhasnopolicyonopenaccesspublishing. Thereare three 'don’t knows', oneofwhomisatauniversitywithaverywell-knownpositivepolicy.Hereisaquotefromsomeonewho isawareofoneof thesepolicies, and it representsoneof thevery smallnumberofmentionsofinstitutionalrepositories/greenOA:“TheUniversityhasstartedaHarvard-typesystemandgetcopyfortheIR”.HepersonallylooksforanOAjournal.Onealsohastobearinmind thatmaybe50%of the23ECRs,whoseuniversitiesdonothavepolicieshavenoviewstoexpressunderthisheadingandprobablyareagnostictowardsOAanditdoesnotenterintotheirdecisionmakingprocess.

Finally,thehypothesisthatECRspublishinOAjournalsbecausetheyareeasiertogetinto,posedtoECRsinallcountries,wasroundlyrejected.Clearly,thereisalotlessdistrustofopenaccessthanthereoncewas,whichisinlinewithCIBERfindingsin2013intheAlfredP.SloanFoundationfundedstudyoftrustworthinessinthedigitalage(Watkinsonetal.,2016).

Whatthenofchoosingajournalhavinginnovativefeatures,suchasvideoarticles(e.g.,Jove),to ECRs when placing their research? In the US nine (32%) answered yes when queriedabout theirpossiblepreference forsuch journals,and19 (68%)answeredno.Butat leasthalfareeitherawareofjournalswithinnovativefeatures,ortheconcept,orareexcitedbyit. Onewouldexpect aproportionof interviewees tobe indisciplineswhere video isnotused,sothisisaninterestingresult.Negativesincluded:“itistheprestigethatcounts”and“ithastohaveagoodIF”and“IlooktobalanceofreadershipandIF”,but“toobusy”was

33

really the big excuse. The UK responses are similar, if a little more muted. Six (29%)answered yes and15 (71%)answeredno,buttherearepositiveremarksaboutinnovationamongthelatter.Severalhavebeenaskedtoproducevideosbypublishers,buthaddeclinedbecauseoflackoftime.

8.2.3.4 Publishingstrategies

ECRswere askedwhether they had a publishing strategy, for instance publishing in highimpact factor journals, or whether they operated in an ad hocway. In the UK only four(19%)deniedthereispressuretopublishinhighimpactfactorjournalsandoneofthemwasimpelledtoexplainthathewasnot“goingforlowquality”.Ofthe17(81%)agreeing,several(threeatleast)saidthepressureisself-induced.Asocialscientiststatedthatthe“balancebetweenhighIFandhighreadershipworriedher”,anassistantprofessorontenuretrackinanappliedsciencedepartmentworriedthatIFwasbecomingmoreimportantinhisresearchlifetime, and an experienced neuroscientist pointed out that the most importantrequirementistogetapublicationout.

In theUSonly six (21%) ECRs denied anypressure and thiswas surprising in viewof thewider range of universities involved. Onewho answered 'no' suggested that you did notneedanypublicationstogetadoctorateinhercurrentuniversity.Anothertookadifferentposition:“No,butitisanegothingandIwillgoforitanyway”.Someofthe79%(22)whosaid'yes'werekeentomakethesamepoint.Pressurewasself-induced.“Ihaveto,inordertogetajob.Idothepressure”wasatypicalreply.Otherssaidthatthepressureisnotstrongandmightevenbedescribedasencouragement.Onepost-doctoralmadethepointthatthebigpressureistogetfunding.

8.2.3.5 Experiencesothercountries

China

Asamatterofcourse,ChineseECRscheckifthejournalisindexedinSCI,SSCI,A&HCI,andEI,orChineseindexes,suchasCSSCIandCSCI.ECRsneedtoreachtheparticularrequirementsoftheinstitutes,whichareverydemanding,sothestandingandrankofthejournalisveryimportant.Listsand indexesarevery important.Second, theywilldeterminewhether thejournalisrelatedtotheirresearchfieldandwithinthistheygenerallypreferjournalswhichare really specific to their “small field". Such journals bring their researchto their targetaudience,sothattheycanmakemaximumimpact.However,havingsaidallthis,afewdooptforthetopjournalwithhighestJIFandabigrejectionrate(thisisadesirablefactor),becausetheywantto"challenge"themselvesor,maybe,thesystem.Third,theyconsiderarangeofotherfactors,suchasprocessingduration,thejournal'sbroaderreputationinacademiaandpublishingcost(OApublishingfee,forinstance).

ChineseECRsdohaveaclear, fixedpublishingstratagem.Mostwill followthesteps listedabove,somewilloptforthetopjournalsand, if theygetrejected,theywillsubmittothesecondrankedones.Oneevensaidthattheywillstudythejournalsbeforewriting,aslongas

34

theyhavethe"targetjournal",theywill"tailor"theirpaperforthatparticularjournal.

France

FrenchECRsareverypragmaticwhenitcomestomakingtheirjournalchoices.Whetheritistheirchoiceorthatoftheirsupervisors,theyselectthejournaltowhichtheypublishonthebasisofrelevancetothetopicanditsIF.Thus,withinthegroupofthemostrelevantjournalstotheirresearchtopic,ECRstrytotargetthejournalwiththehighestIF.Insomecases,thefactthatajournalisrankedinthefirstquartile(WoS)isanimportantcriterionforECRs,theirsupervisorsandheadofdepartment.

Whenthechoiceofajournalistheresultofadiscussion,theconsensusbetweentheECRs,thesupervisorsandheadofdepartmentisbasedontheappropriateIFlevel(nottoohigh,nottoolow).However,inthecaseofarticleshavingthenameoftheECRasfirstauthor,the“high”IFargumenttakesprominenceasitimprovestheirchancesofgettingajob.FourECRsalsomentionedthat,whenchoosingatoprankedjournal,theyfactoredintoaccounthowlikelytheyweretobeaccepted.Abovealltheydonotwanttowastetimeasitisacommoditytheydonothave inabundance.Rejectionswastevaluabletime(butnot, interestingly, forChineseECRswhosaidrejectionsproducevaluableadvice).Thenameofthepublisher,theeditorinchief,theopenaccesspolicy(embargos)oranyothercriteriadonotseemtoplayanyroleintheselectionprocess.

It is interesting to note that publishing in “indexed journals”, an important reputationalrequisite, is implicitly understood as indexed in Web of Science, which has a uniquereputationalpositioninFrance,especiallyforevaluators.AjournalindexedinScopus(andnotindexedinWoS)islesshighlyregarded,ornotconsideredatall(asinthecaseofPhysics).

Malaysia

MalaysianECRsareunanimousinstatingthattheyfirstchoosejournalsonthebasisoftheirrelevance to their research field and this, probably, can be taken as true for all ECRseverywhere, even if not made implicit. Three quarters of ECRs say this. After that,interestingly,comeschoosingmultidisciplinaryjournalsrelatedtotheirdiscipline(nearlyhalfsaythat).Then, inorderofpopularity,comesuchfactorsas: IF journals (WebofScience);journalslistedbyScopus;journalswithnopagechargesorsubmissionfees;journalswithanearly view online function (a sure sign of rapid publication); journals approved by theuniversityandtheMinistry;andjournalsthatreviewquickly.AsECRsprogressthroughtheiracademiccareer,attitudestowardsdisseminationchange.InthecaseofMalaysianECRs,thismeanstheyareevenmoreoccupiedwithjournalqualityandpublishingimpact.Nothingnewhere,buttheyhavealsotoadoptamoreopenapproachtodisseminatetheirresearchworksonsocialmediaplatformsandindigitalrepositories.

Poland

InPolandthingsarehighlyprescribedandseeminglyroutine.Firstly,andmostimportantly,ECRstakeaccountoftheListprovidedbytheMinistryofHigherEducation,whichisupdated

35

annually.TheListcomprisesthreeparts:partA(thebestjournals,withJIFsandindexedbyWoS,mostlyinternational),partB(mainlyPolishjournalswhichdonothaveanIF,buthaveaPolish score/points) andpart C (Polish and international journals without JIFs, but with

Polish scores). ListAisthebest,thenCandlastofallisB.Thelistisveryimportant,notonlyforindividualresearchers,butfortheuniversities/departmentsaswell.Everyyearthesumofpoints foreveryuniversity iscountedandgrantsareprovidedaccording to thissumofpoints.Afterthelisttherelevancetothetopicandappropriateaudienceisaconsideration.

It is alsoworth pointing out that in Poland books and book chapters contribute towardsofficialpointscoresandcanattractsimilarpointsaspublishinginListBjournals.

Spain

As with their Chinese, Malaysian and Polish counterparts, Spanish ECRs will conduct anauthenticationcheckoftherankingsandwhetherthejournalisindexedinWoSorScopus,butmainlyWoS.TheywilllookforjournalsinthefirstorsecondJCRQuartile.Metricsarethemaincriteriaindeterminingwheretopublish.Theychoosetoprankedjournalsandiftheirworkisnotacceptedbythefirstchoiceinthetimehonouredwaytheywilltryanothertoptitleandsoonuntilthearticlefinallyfindsitslevel.Theyknowthatthepublishingprocesscanbea longone,buttheyprefertotrymorethanonehigh impact journalbeforegoingdown in the list.Theyalso try to select the journalmoreclosely related to their researchfield.As in China, Spanish ECRs prefer specialist journals because such journalswill bringtheir research to their target audience. The quality of reviews and the duration of thepublicationprocessaretwootherfactorstotakeintoaccount.Afewsaythatiftheyhadthefunding,theywouldchooseOAjournals,butfewdo.8.3 Peerreview

ThemajorityofECRshaveexperienceofbeingreviewedandbeingareviewer,althoughjust30%inPolandandasweshallseethisimpactsontheiropinionofpeerreview.Thequestionasked,whichwasdeliberatelyprobing,was: Doyoufeelthatpeerreviewformostjournalsisin the hands of established researcherswho are not always sympathetic to new ideas? IttranspiredthatECRsappearpositiveaboutpeerreviewing,althoughnotwithoutreservation,andthemajorityhavegoodexperiencesinrespondingtopeerreview,eventhoughitcouldbealongandpainfulexperience.However,beyondthegeneralwarmfeelingtowardspeerreview,attitudestoitarecomplex.Theretendstobeacontinuumofresponsestopeerreviewquestions from positive, positive with qualifications, negative with qualifications andcompletelynegative,withmostinthemiddle.SomeECRsdidfeelthatestablishedresearchersare not always receptive. They usually said other people they knew had had badexperiencesandonlysomeofthemhadhadsuchexperiencesthemselvesorthoughtthatthiswasthereasonforarejection.RegardingthelattersomeECRssuggestedthatitismoredifficult to get papers published (in top journals) if you are not established or in anestablishedandwellknowngroup.USresearchersweremorelikelytobelievethismightbethecase.Overall,though,thereisnotalotofevidencetosupportthehypothesisthatECRs

36

feelalienated/lockedoutbytheexistingpeerreviewsystem,whichtheythinkofasaclosedclub.Thus,fourcountriesrejectedthehypothesisandtherestonlypartlyconfirmedit.ThosethathaveundertakenpeerreviewingarethemostpositiveabouttheprocessandthatisverymuchthecasewithFrenchECRs.Amongthepositives,themostimportantaresaidtobe:a)itimprovesthequalityofapaper;b)itstopsthembeingsloppy.Thefewnegativeonestendtofocuson:a)Editors.Theyarethoughttohavetoomuchpowerandtheydonotcontrol bad/biased reviewers sufficiently. ECRs have suggestions for improvement, butfundamentally, are fairly satisfiedwith peer review. Thiswas found to be the case in allseven participating countries; b) Reviewers. There are complaints that peer review fallsdowninrespecttoreviewerselection;frequentlytheyarethoughtnottobespecialistsinthepapersubject(thiswasthecaseforfourcountries).

French ECRs,mainly bio-scientists, were very specific regarding their criticisms of editorsandreviewers.Theysaid:

§ Itishardtopublishanoriginalresearch;newideasarenotappreciatedbyreviewers.§ Therearetoomanyauthorsandtoofewexpertswhocanhandlethereviewing

process.§ There are toomany articles for the reviewers to dealwith,who therefore do not

spendenoughtimeonthereviewingprocess.Asaresult,theydonotreallydothejobwellanddonotalwaysprovideevidenceforrejection.

§ Theprocessgetsdumbeddownbythepeerreviewers,whoaskECRsandPhDstudentstodothejob,becausetheydonothaveenoughtimetodoitthemselves.

§ Reviewersarethecompetitorsandcolleagues,whichcanintroduceanunfairpersonalelementtotheprocess.

§ Thereneedstobemoreturnoverofreviewersinordertoreducetheirinfluence.Suggestionsforimprovingpeerreview,whichhadreasonablywidespreadsupport(fromECRsin four countries) include: authors and reviewers should be in contact to solve doubts ormisunderstandingsduringtheprocessandreviewersshouldbeidentified.ASpanishECRevenwanted to know thenameof reviewerswhen theprocesswas completedandaPolishECRsuggesting having three reviewers, which is notuncommon. There is universal support fordoubleblindpeerreview,withallcountriessupportingthehypothesisthatECRspreferdoubleblindpeer reviewbecause itprovides fairerappraisal,butwiththeprovisothat itshouldbeproperlyblind.The latterwouldprovedifficult toachieve in Poland because in very narrowspecializationsallexperts/reviewers’namesarewellknown.

Inthescholarlyworldmuchiscurrentlyclaimedofopenpeerreviewwithadvocatesarguing,for instance, that: "by adopting a more transparent process of research evaluation, wemoveonestepcloser towardsa faireranddemocratic researchprocess" (Tennant,2016).WhatthendoourECRsthinkofaprocessthatappearstodealwithsomeoftheirconcerns?Thequestionputtothemwas:wouldyoulikeallpeerreviewtobeopen?Infact,itturnedoutthat they are not sure because there seems to be a conflict for them. They believe that

37

transparencyisagoodthing,butthatitdoesnotworkinpractice.Thus,oftheUKECRs,just10outofthe18whoansweredthequestionsaidtheywouldlikeallpeerreviewtobeopen;this compared to just eight out of 23 for US ECRs. Their specific criticisms reflect theuncertainty: ‘toorisky’, ‘moreofaworry forECRsbutOKforseniors’, ‘unwantedeffects’,‘dangerous’and‘itwouldbemoredifficulttoreject'.Ingeneral,FrenchECRsarealsoverysuspiciousofanythinglabeledas‘open’ortransparent.Ironically,theEuropeanCommissionwanttomovescholarsfurtherdownthisroad.

No surprise, then, that the hypothesis Early career researchers are worried by toomuchtransparency in peer review because it will make it difficult for them to criticise thesubmissionsoftheirseniorsdidnotobtainuniversalsupport.Viewsareverymuchsplitonopenpeer review,withECRs in threecountries insupport (UK,USandFrance), threenotconfirmingandoneonlypartlyconfirming.

Researcherswerealsoaskedwhetherpublishers,thepresentincumbents,shouldorganizepeer review. In the case of theUK, nearly three quarters of ECRs said they should do it,althoughsomenotwithagreatdealofconviction.Therewasasensethattherereallyisnoalternativeand,asoneresearchersaid,“publishersareindependent”.

Finally, while writing this report, a PRC funded report was published on peer review(PublishingResearchConsortium,2016),sowehavetakentheopportunitytocompareourresultswith thoseof the report. The report covers all researchers andnot just ECRs, andalthougharoundaquarterofrespondentswereaged35orless,thedatahavenotbeencutby age.Nevertheless, the general results are very similar toours,with around two-thirdssayingtheyaresatisfiedwithpeerreviewandthreequarterssayingitimprovedthequalityofthepaper.ItmightbeprofitabletoidentifyECRsinthePRCsurveyandthenenhancetheanalysiswithdatainthisreporttoprovidesomethingreallysubstantial.

8.4 Openaccess

Openaccesshasbeendiscussedalreadyintermsofpublishingandwediscussitinbroadertermshere.OAisgenerallyunderstoodbyECRstobegoldopenaccess.Thegreenrouteisnot really considered, or confused with social media depositing (on ResearchGate, forinstance).Goldopenaccessisuniversallythoughttobeagoodthing,butthisisnotarguedbyECRswithanyrealpassionorcommitment.ThereissomedisquietamongSpanishECRsthatOA ismakingtheplayingfieldunevenbetweenthoseresearchersthathaveaccesstofundsthatcanpayforitandthosethatdonot.Thisisagoodpointthatneedsfurtherinvestigation.

Archivingtheirresearchworkinrepositoriesisanon-priorityforECRs,theyseethisasthejobforlibrariansorresearchadministrationofficers.Researchersfromeverycountryagreeonthis.Thereisageneralabsenceofknowledgeofand interest inrepositories,totheextentthat a significant number do not even know if their institution has an institutionalrepository. Depositing is not undertaken with any warmth, it is considered obligatory.SpanishECRsareespeciallynotcommittedtoarchivingtheirresearchoutput.

38

8.5 Socialmediaandonlinecommunities

All countries reported that their ECRs do not cite social media or release their researchfindingsordataonsocialmedia.However,theydousesocialmediaforcommunicationandfindings andpassing informationaround. The caseofUKECRs is illustrative,where socialmediaarewidelyused for finding information,withall 20 researcherswhoanswered thequestionsayingtheydidso.ResearchGate(14mentions)andTwitter(8)areclearlythetoolsofchoiceandtheycomeintotheirownwhenlookingfordifficulttofindthings,andtheall-importantserendipity.

ExceptinthecaseofChina,ECRsuseonlinecommunitiespassivelyandmainlyResearchGate.ManytalkedaboutsharingandmostknewaboutandareatleastregisteredonResearchGate,butfewuseditssharingorcollaborativemechanisms.ObtainingPDFsandconnectingwiththeircolleaguesarethemainactivitiesundertakenbyECRsonsocialnetworkplatforms.

Thereare, though, signs that there couldbe changes in thepipeline. Thus, in theUK/US,where although by no means all researchers used social media in their scholarlycommunications,alotofthemdid.However,thosewhodidnotoftenfeltthattheyshouldmakemoreuseoftheopportunitiespresentedandmightdosointhefuture.Significantly,thedrivetousesocialmedia,especiallytoreachpractitionersandpolicymakers,appearstobecoming fromuniversitymanagement,marketingdepartmentsand, less so, their seniorcolleagues. Spanish and to a certain extent British ECRs consider outreach anddisseminationtoindustryandsocietyasveryimportantscholarlyactivities,butdonothaveenoughtimetodoasmuchastheywouldlike.Theywouldliketoincreasetheirpresenceinscholarlysocialmediaasameanstoachievethisaim.

Becauseofthehype(andpromise)associatedwiththetopic,fourhypothesespoweredthesocialmediaanalysis:1) ECRswouldliketousesocialmediamore,buttraditionalnormsthatdominatescholarlybehaviourpreventthemfromdoingso.Evidencefromfiveofthesevencountriesconfirmedorpartlyconfirmedthis.ThehypothesiswasunsupportedinFranceandtheUK,althoughinthecaseofthelatter,aswehaveheard,researchersarebeingencouragedbytheiruniversities,mainlyforoutreachandshowcasingpurposes,withwhatseemstobeagrowingimpact.

2) ECRsdonotseesocialmediaasbeingscholarly‘noise’,butusefulforresearch purposes.Countriesweresplitaboutthis,withChina,MalaysiaandSpainbelievingthistobetrue.

3) Socialscientistsaremorefavourabletowardsthescholarlyuseofsocialmedia.There isverylittlesupportforthishypothesis,withjustSpainpartlyconfirmingit.Sosocialscientists,despitehaving(potentially)moredegreesoffreedomthantheirscientificcounterparts,donotfavoursocialmediamorethantheirscientificcolleagues.

4) ECRs are detached from institutions andmore closely networked/connected with theirpeers.ThisisthoughttobeanimpactofonlinecommunitiesandsomethingResearchGateencourages.Itreceivedareasonableamountofsupport,withjusttwocountries(UKandUSA)

39

notsupportingitwhollyorpartially.MaybetheUS/UKECRsarebetterconnectedandonlinecommunities actually provide a real advantage to other countries where internationalconnectionsarenotsoeasilyestablishedand,hence,aremoreprized.

As alreadymentioned,more informationon socialmedia andonline communities canbefoundintheSharingandCollaborationandMetricssections.

8.6 Discovering/findingpublications/information

GoogleScholar(GS)holdsavirtualmonopolyforfindingscholarlycontent. InallcountriesbarChina,GSisverymuchthetooloffirstchoice.Forinstance,inthecaseofMalaysia11outof the 12 ECRs interviewed rated it first as a discovery tool. Library-based and publishersystemsverymuchtakeabackseat.Itwouldbethesame,too,forChina,butthereGShasbeenblockedbyGovernmentauthoritiesand soawholevarietyofdatabaseshave tobeusedinstead-PubMed,WoS,CNKIandBaidu,forinstance.However,evenso,10outofthe13ChineseECRsfindwaysofaccessingGS,viaaproxyserver,forinstance.Establishedjustover a decade ago, GS has moved from being a figure of ridicule to a dominant andrespected member of the scholarly communications community. It has transformed thediscovery(andreputational)landscapeandiscertainlyachangeagenttowatchinthenextdecade.8.7 Smartphones

Of course, smartphones are associated with the social media, and like the rest of thepopulation most ECRs (but not in Poland) have smartphones, too. However, with theexceptionoftheChineseECRs,theytendtousesmartphonesonlyoccasionallyforscholarlypurposes.AndPolishECRsnotatall.InthecaseoftheChinese,smartphonesarewidelyusedforscholarlycommunicationpurposes,butnotforreadingpapers(“thescreenissmallandmarkingthemupisdifficult”).MalaysianECRs'mainuseofsmartphonesisforfindingandchecking information on the Internet, and only a small number acknowledge using it forscholarly reading and note-taking. ECRs in other countries said that they readmainly oncomputer screens, but hardly ever in print or on smartphones. UnitedStates and UKECRs do use smartphoneswhen away from their desk. Thus, amajority ofUK ECRs (12)usedtheminthisway,butonlyoccasionally,foralerts,whiletravellingandatconferences.Generally, smartphones arenot talkedaboutwith anypassionand certainlynotwith thedevotionthattheygetinpersonalandsocialenvironments.So,onlyalittleimpactherefromthis technology, but we need to monitor this activity closely as research publishedelsewhere indicates that smartphones are increasingly used in an academic sense forreading purposes (Halevi et al., 2015) and most frequently used for checking socialnetworkingsites(Madhusudhan,2015),andweknowfromourresearchthatECRsareusingthesesites.

40

8.8 Openscience

Intheinformationsciencecommunitythereismuchtalkabouttheopenagenda,whichissponsored by the EU and UNESCO inter alia. Despite this, and the associated hype andpromotionsurroundingthetopic,ECRsdisplayedhardlyanyunderstanding(or interest)ofopenscience,Web2.0etal.whatsoever.Notsurprisingly,then,ECRshavefewopinionsonthetopic,andthosewhoventuredsometendnottobeverywellinformed.Itwasnotjustthecaseof them notunderstandingthetopicbecauseoftheconfusingnomenclature,forevenafterpromptingECRswithexamplestheyoffered littleonthetopic.ECRsareclearlynotharbingers inthis regard.Take, for instance,ChineseECRs:eightof the13statedthatthey had never heard about open access and nine of them clearly said that they knewnothingaboutopenscienceorScience2.0.ItwasthesamewithBritishECRs,withjustthreeoutofthe21showinganyunderstandingoforinterestinthetopic.

EventhefewECRswhohavethoughtsaboutscholarlytransformationaremoreinterestedinthe sort of "pain free publishing" that eLife promises (https://elifesciences.org/about),because they are preoccupied with publications. Having said that, the related questionaboutopendataandsoftware (componentsofopenscience)didstirsomeinterestamongUK/USresearchers.Theyarekeenonobtaining credit for such activities, but not quite sointerested insharingdata (an importantelementof theopentechnologies),becausetheywanttoexploitthedatatheyhavegatheredandnotgiveitaway.

Theopenagendaincludesblogsasnon-traditionalscholarlyoutputs,butnoone,certainlyintheUK/US,arereally interested inblogsasanalternativetopublications.This isprobablyunlikelytochangebecausethecurrentevaluationsystemdoesnotrecognizenon-scholarlyoutputs.

ForthosefewFrenchECRswhoknewalittleabouttheopenconcept,theyseeitalsoasanewmeansof imposingcontrolandevaluation. It is seenmoreasaconstraint,somethingyourarelyhearmentionedelsewhere.

8.9 Sharingandcollaboration

Sharingandcollaborating, intheoryeasiertodonowwiththeemergenceofsocialmediaand online communities, are highly thought-of activities, which are expected to powerscience.Researcherswere,thus,askedinwhatwaystheyshare:a)ideasandinterimresearchresults; b) research findings, data andpublications.We shall first lookmainly atwhatUKandUSECRssaid.As justtwoUSECRssaidtheydonotshare ideasand interimdata,andone ECR clearly misunderstood the question, it seems that 25 (89%) do share suchinformation.IntheUKtherearealsoonlytwoECRswhodonotshare,and19(90%)whodo.IntheUSA13(52%)ofthe89%shareatinternalmeetings,workshopsorteamtalks.IntheUKonlyfive(26%)mentionedinternalsharing(maybeitisjustassumedbyothersandthefigurecouldbehigher?).16%oftheUSsamplewhodidshare,shareontheirnetwork,asdo21%oftheUKsample.Thiscouldmeanquiteawidenetwork(onementionedaprofessionalsite), or it could be a closed network of friends using email (the latter still being a very

41

important sharing instrument). It is not certain whether interviewees are talking aboutsharingamongpeerstoobtainfeedbackorifaretheyputtingitalloutthereforpurposesofoutreach.“Sharing”ismuchmentionedbyECRsascentraltothewaytheywanttolivetheirscholarly lives and, maybe, they are a little conflicted when they have to act by theacademic “rules”. Finally, there is sharingof ideas and interim results using socialmedia.Surprisingly, perhaps, only one US intervieweementioned socialmedia and they did notspecifywhich particular type. Six (31%)UK ECRsmentioned socialmedia andwere quitespecificmentioningResearchGate,AcademiaandLinkedIn,inthatorder.

Regardingthesharingofresearchfindings,dataandpublications,manyoftheintervieweesunderstoodthequestionasreferringnottosharingofpublicationsordata,butoffindings,sothe normal answer tended to be “publications” and “conferences”. Other questionselsewhere,aboutpublishingpracticesandaboutdisseminationusingsocialmedia,providedthe additional information needed to obtain a better idea of what goes on. Of course,disseminationisadifferentthingthansharing,buthowmuchitisconflatedinECRmindsissomething we shall have to probe next time around. All ECRs from both countries didmention publications and conferences and 19% of UK interviewees (4) and 11% of USinterviewees (3) appear from the evidence available to abstain from any other form ofsharing/dissemination.Someactuallysaidtheydonotusesocialmediainthisregard.

ThereareinterestingdifferencesbetweenUSandUKECRs.AmongthoseusingsocialmediainordertoinformothersabouttheirresearchoutputsintheUK,58%useTwitterand52%useResearchGate.IntheUSA,48%useResearchGateandonly8%Twitter.Numericaltotalscoveringbothcountriesare21usingResearchGateand12 tweeting.TheoverallnumbersusingResearchGatefordisseminationcouldbeevenhigherbecauseinbothcountriesthereareothermentionsofputtingup“profiles”,whichislikelytomeantheuseofResearchGate.Elsewhere,answerstoquestionssuggest that formanytheuseofResearchGate ismainlylimitedtoputtingupaprofile:veryfewmentioned“upload”inthecontextofthisquestion,but maybe they are less likely to admit this given the sponsors of the research project.Otheruseofprofilestoshowcasepublicationsetc.refertousingLinkedInforthepurpose.IntheUK29%(5)ofthosewhoattempttodrawattentiontopublicationsandconferencepresentationsmentionLinkedIn,whereasitisnotmentionedspecificallyintheUS.IntheUS,however,20%(5)mentionFacebookinthisconnection,whereasintheUKonlytwo(12%)didso.IntheUKAcademia.eduwasmentionedtwice,butsurprisingly(itisUSbased)notatall in theUS.Blogs werementioned once in the US and twice in theUK. Therewas alsomentionofpresentationswithin internal seminarsandalsoamongnetworks.Presumably,there aremoredisseminationof links, at least, than ismentioned– probably, itwas justassumed.Itisalsonotclearwhatthementionofnetworksreallymeant–closedoropen?IntheUKthere isonly onementionof outreach in this context, and that is anegativeone,whichisrathersurprisinginviewoftheREF.IntheUS,acoupleofintervieweesmentioned“outreach”andathird,whoalsomentionedoutreach,qualifieditbypointingoutthatthereis no encouragement to do this. One US interviewee mentioned that making researchavailablemeantonlyafterpublication–“asshehadbeentrainedtodo”-butwhatsheis

42

makingavailablewasnotclear?Noonesuggestedthattheymightbepointingtoversionsavailableinrepositories.

What then of collaboration? It is difficult to be sure because ECRs portray a mixed andconfusedpicture,justthekindofpictureyougetwhenevaluatingadisruptivetechnology.Ashas alreadybeennoted, there is notmuch collaboration conducted through socialmediaandonlinecommunities.Asfaraswecantell,andalthoughalmostallUK/USECRshavetheirnetworkswithwhomthey interactoutsidetheirgroups, theredoesnotseemtobemuchevidenceofformalresearchcollaboration.Itisratherpiecemealhelpamongfriends,givingfeedback, links and advice. Nevertheless, while there is no broad consensus as to the(presumed)valueofsocialmediainbuildingresearchcollaborations,thereisstillactivityandinterest.Thus,amajorityofUKECRs(12)thinkthatthereisabenefithere,withResearchGateandLinkedInthetoolsofchoice.Socialmediaparticularlyseemtowork,too,forChineseandMalaysianresearchers,inacceleratingacademiccollaboration,butthenontheotherhandnotforFrenchresearchers.FrenchECRsspendlittletimeusingsocialmediaanddonotthinkthatitcanbringthemanythingotherthanmore(digital)visibilityontheWeb,whichofcourseis important in itself. Spanish ECRs are not committed to building reputation throughsharingpapersonsocialmediaandacademicsocialnetwork.

Thehypothesestotestinthecontextofcollaborationwere:

a) Earlycareerresearchersshareandcollaborateextensivelyevenattheriskoflosingtheircompetitive edge. There is notmuch of a consensus here,with just one country, France,confirmingthatthisis,indeed,thecaseandthreeothercountriespartiallyconfirmingit.ForFrenchECRs,despiteeschewingsocialmediaforthispurpose,collaborationisclearlyking.Besidespublications,collaborationisaconstantobjective.ECRs'strategiesforgettingajobandpublishingmoreandbetterpapersrelyoncollaboration.Conferencesandmeetingsarekeymoments,dedicatedtosearchingforcollaborations.ECRsbelievethattheycanbehiredfortheirCV,butalsoforthepotentialoftheircollaborations.

b) ECRsmakeuseofsocialnetworkingsitesinordertobuilduptheirownnetworks,separatefromthenetworksalreadyestablishedbytheresearchgroupstheyworkinortheconnectionsoftheirmentors.Generally,thereisverylittlesupportforthis,althoughmattersareprobablyintoomuchastateoffluxtobereallysure.Justonecountry(Malaysia)supporteditanditpays to look at the situation in this country to seewhat we can learn. InMalaysia, ECRscientists(notsocialscientists)aresaidtobemorefamiliarwithacademicsocialnetworksand share their researchdataon them.While invisible colleges, in the formofdiscussionforums (online) andmeetings (face-to-face), are stillmostly used for the sharing of ideasand collaborating, online communities aremaking theirmark. International collaborationthrough academic social networking sites, such as ResearchGate, Mendeley andAcademia.edu,isthoughtbyMalaysianECRstobeafeasibleandeffectivemeanstoaddressimportant research challenges, by increasing opportunities for professional support andnetworking, problem- solving, discussion of data, and ultimately publishing. ECRs believethattheirsharing/collaboratingbehaviourisdifferentfromthatoftheirresearchmentorsin

43

current and previous jobs. They feel that their behaviour with respect tosharing/collaboratingischanging,astheyhavebecomemoreexperienced.Real-time,open,collaborative science tools, however, such as ThinkLab and F1000Workspace, werementionedbyonlyoneECR.AlltheMalaysianECRshavenofearsoflosingtheircompetitiveedgethroughsharingandcollaboratingextensively.

8.10Reputationandassessment

Almostallearlycareerresearchersrecognizethatpublicationsareessentialresearchoutputsinordertoclimbtheacademiccareerladder.Everythingwehaveheardsofarconfirmsthis.Nevertheless, ECRs do feel they are “slaves”, albeit willing ones, to a publishing-basedreputationalsystem.WhileECRsbelievethesystemofreputationalassessmentisimperfect,theyfinditdifficulttothinkofhowtochangeitforthebetter,exceptperhapsbyobtainingamorecomprehensiveevaluationofscholarshipandmaking it lesspaper-centric.Spanishresearchers areparticular vocal about this.Aminorityof researchers, especially from theUK/USA,arealsoworriedthattoomuchpressuretopublishintopjournalscoulddistortthewayresearch isdone,and indoingsocouldslowdownscholarlyprogress.AlthoughECRsadmittheimportanceofthesocialmediatofacilitatecommunicationandreachout,theyrelysolelyonpeerreviewedjournalstobuildtheirreputation.

FromECRsinfivecountriestherewerecomplaintsaboutdifferencesbetweenthedisciplineswithregardtothetimeanddifficultyofobtainingresultsandthenpublishingthem,whichtheyfeelputsthematareputationaldisadvantage.Somefeelthisneedstobefactoredintoscholarly reputational assessment, as it does in Malaysia. Data from five countries alsoprovidedsupportforaSpanishfindingthatECRswhoworkinbiggroupshaveareputationaladvantage.

PeriodicevaluationsarethenormforECRs.TaketheexampleofBritishECRs,forinstance,with17outofthe21researchersbeingsubjecttoformalevaluation.Evaluationsareheldtypically yearly, half-yearly and termly. Productivity, impact, outreach andpresence/reputation are among the factors contributing towards the evaluation of otherECRs.OneBritishresearcherisevaluatedpurelyonoutputs–publishedpapersandgrantswon.

There was one hypothesis to test relating to evaluation: ECRs are ‘slaves’ to a metric-based/journalfocusedsystem,whichtheyhavetoadheretoinordertoclimbtheacademicladder.Asmentionedpreviously,nocountrydisputedthis,althoughtwo(FranceandUK)onlypartiallyconfirmedit.8.11Metrics

Metrics,althoughnotuniversallylikedbyECRs,arethoughttobeafundamentalelementofreputational assessments. Metrics are important because of the highly competitiveenvironmentinwhichECRsfindthemselves.Afterall,inchasinghighimpactjournalsECRsarechasinghighlycitedones,ametricthathasbeenwithusaverylongtimeandonewhichis

44

verymuchstillking.Thus,datafromfivecountriessupportedafindingfirstemergingfromthestudyofPolishECRs,whichwasthatmetricsareimportantbecauseoftheassessmentsystemthatisembeddedinacademia–collectingpoints/scoresforpublishinginexcellent,internationaljournalsandotherprescribedacademicactivities.Theyareanecessity,althoughnotquitesomuchintheUKandUS,wheresystemsarelessprescriptive,sofaranyway.

Citations(inalltheirforms)areconsideredofgreatvaluetoECRsinbuildingreputation.Thereislittlediscussionaboutthemeritsofcitations,theyaresimplyregardedasafaitaccompli.Butitisaverydifferentstorywhenitcomestotheother(alt)metrics.Despitethefactthatmanycommentatorsarebusytalking-uptheprospectsofaltmetrics(WilliamsJune,2016),thereisnotmuchevidencethattheyarepopularorwidelyemployedbyanyscholars,nevermind ECRs. Our ECRs demonstrated little interest in social media and usage metrics asreputationalmeasurements.ThereisalittlemoreinterestinChinaandMalaysia,butstillnotalot.Thusonlyfourofthe13ChineseECRsandthreeofthe12MalaysianECRsknewwhataltmetrics are about. While altmetrics are not widely used and accepted by eitherresearchersortheuniversitysystem,ECRstendtoagreethatitisapotentialnewmethodtoevaluate researchers’ output and influence. Some of them do check their publicationsdownloads. But nothing, including citations, beats publishing in a good journal for theircareer,sowhybotherwithaltmetrics?

TheotherbarrierforChineseECRstoacceptingandusingaltmetricsisthelackofacrediblealtmetricstandardorintegratedplatform.Forinstance,nowadaysanauthorcanchecktheirbookaltmetricscoreonSpringerlink,buttheycannotcomparethescoretothatoftheirpeersbecausetheyhavepublishedtheirbookwithSage.AsoneoftheChineseECRssaid:"ifaninternationallyacceptedplatform isestablished, thenewmetrics, suchasusagedata,willwork."To conclude, the vastmajority of our 100+ researchers, boasting around 1200 publicationsbetweenthem,areatbestlukewarmwhenitcomestoaltmetrics.Notsurprisinglythen,thehypothesis which actually acknowledges one of altmetric’s alleged reputational attractions:ECRsareinterestedmoreinsocialmediaandusagemetricsbecausecitationstakesolongtocount,obtainednosupport,noteveninChina.

8.12Impact

WithpressureplacedonresearchersbythelikesoftheUK’sResearchExcellenceFrameworkto demonstrate that their research has an impact, and knowing how importantcommunicatingtothepublicistopolicymakersandindustryinordertohelpachievethatimpact, itmight be expected that impact is an areawhere things are happening,maybechanging. Thus, the investigation of impact led with the hypothesis that: ECRs seeconnectingtoawideraudienceasbeinganimportant[research]impact.Thisturnedouttobe thecase,withonlyChina rejecting thehypothesis,Malaysia confirming it fully, andalltheothercountriespartlyconfirmingit.

45

LookingattheChinesenaysayersfirst,ECRsthereconsiderscholarlycommunicationtobeanessentialcomponentpartofscientificresearch,andbelievethatthiscommunicationismainlyacademicoriented.Sevenoutofthe13ECRslisted“peers”asthemostimportantgroupthatresearchersshouldtarget,andmostsaidthattopublishpapersinthetopjournalsintheirfield is the bestway to influence peers. Unlike the research finding released by Springer(SpringerNature PublishingGroup, 2015), Chinese participants in our study showed littleinterestincommunicatingtothepublic.Onlyoneparticipantthoughtthatthepublicisthemost important group that her research should have an impact on, but that could beexplainedbyherfieldofresearch,whichisstronglyapplication-orientedandpublic-serving.

Turning toMalaysia, another country in Asia, the very opposite is true. Malaysian ECRsbelievethatitisimportantfortheresearchtheyareinvolvedintohaveanimpactoutsideacademia.Aftertheiracademicpeersitwasfeltthattheirworkshouldimpacton,inpriorityorder, the general public, industry and the government (policymakers). The bestway toinfluence the general public is through the social media, industry throughmeetings andconferences,andthegovernmentthroughthemainstreammedia.

InSpainECRswouldliketohavemoretimetoworkmorecloselywiththeirlocalcommunityorindustry.Inotherwords,inordertoobtain‘real’impact.Unfortunately,becausetheyarefocusedonpublishinginhighimpactjournalsforthesakeoftheircareers,theydonothavethetime.ThesamecanbesaidaboutPolishECRs.

8.13Roleoflibrariesandpublishers

Mostviewsabout“commercial”publishersandaboutlibrariesarenegative,thoughnotmanyECRs had views on particular publishers, and in both cases they demonstrate a lack ofunderstandingofwhatpublishersorlibrariesdo.Itisalmostasthoughtheyseethembothaspartofaninvisiblescholarlywallpaper,andnotashighprofileactors,astheymightliketothinktheyare.Inrespectofpublishers,thiscouldbepartlyexplainedbythefactthat,inthecase of the nearly 50 US/UK ECRs, only two had close connections with journalmanagement,suchasbeingmembersofaneditorialboard.Theotherssawjournaleditorsasthekeyfigureswhoshouldadoptparticularpolicies,policiesthatusuallyreflecttheirownexperiencesasapublishedauthor. Taken togetherwith the fact that ECRsdonot choosewheretopublishonthebasisofthepublisheranddonotappeartousepublisherwebsites(oftenpreferringfree,morevisibleandopenservices,suchasarXiv.organdResearchGate),publishersarethuslookingalittleanonymous.Fortunately,ashasalreadybeennoted,ECRsare satisfied enough (not necessarily happy) with publishers managing the peer reviewprocess.However,ECRsarenot so satisfiedwith learned societiesmanagingpeer review,largelybecausetheyarenotthoughttobesufficientlyindependent.

Librariesare,perhaps, inamuchworseplace.Whilepublishersmighthaveabrand issue,their journals have no such problems. Libraries have nothing similar to offer, unless youcount ‘their’ discovery systems which are hardly ‘theirs’, and they appear equallyanonymoustoECRs.Also,makingmattersworse,institutionalrepositoriesinwhichalotof

46

store has been set are not popular, and are certainly hardly searched. As our Frenchcolleaguesaidafterinterviewing14ECRs,theysimplydonot“see”librariesanymore.Someofthemhavenotgonetothelibraryforyears.Librariesaremainlyconsideredasaplaceforundergraduatestositandworkandthatobviouslymakesthemveryexpensiveassets.

8.14Transformations

Ingeneral,notmanyECRswereforthcomingaboutcurrentandfuturemodesofscholarlycommunicationandmostoftheideasweobtainedwereprosaicincharacter,suchasmoresharing, greater access and transparency, better/moreopenpeer review,more credit fornon-standardoutputsandlessmetrics.Thepossiblereasonsforthisareasfollows:

• ECRsdonothave the time/need to think about transformationsbecause they areoverloaded with many responsibilities: research, publications, teaching andadministration,aswellasthenecessityofearningmoney(fivecountriessupportedthisview).

• In general, ECRs know the importanceandnecessityof changing,butdue to theirhumbleandlowlypositionstheybelievetheycannotimpactonthecurrentsystem.Sotheyadoptanegativeornon-plusattitudeonshort-termchange.However,somearepositivetowards long-term,systemicreform(allsevencountriesbelievethistobethecase).Perhapsitiseasiertogazeintoadistantfuture?

• Newbehaviourscannotreallytakeholdwhileacademicsarerecruited,promotedandobtain funding on the basis of their publication record and citation scores. Fourcountries felt this was the case and the remaining three did not have enoughevidencetosayforcertain.

However,therearesomesignsofpossiblechangeinthepipeline:• There are a few ECRs who did have quite innovative although not revolutionary

ideas, which may not have been complete but which had the seeds of a newsystem, although most seem more resigned than positive. See below for instances inconnectionwithUSECRswhoseemmoreforthcoming.

• Thosewhodonot use socialmedia, particularly Twitter andResearchGate, at thesame time regularly used the words "not yet", often accompanied by the word"should".

• ThereissurprisinglyhighuseofLinkedInintheUKandUSA.UseofLinkedIn,likemuchoftheuseofResearchGate,istomaintainaprofile.ItissurprisingbecauseLinkedInisusuallynotthoughtofasaplacewhereacademicsgo.

• Similarly,theregularuseofthewords"transparency"and"sharing",asfuturedecisionactions,isanothersmallsign,perhaps.

• Manyacceptedtheideathattheymightchangesomethingwheninapositiontodoso.

• Aminorrebellionseemstobeintheairoverthedominantandoverbearingpositionthathighlyrankedjournalscurrentlypossess.

47

Looking at the issues raised in more detail, a few UK/US ECRs thought that there areaccessibilityproblems,withresearch locked in labbooksor theseswhichcannotbeeasilyaccessed,andthatthedatathatmakesitintopapersareoftenpresentedinthetraditionalway,whichdoesnotenableeasyre-use/reproduction.

Somemorewiderrangingcomplaintsweremadeaboutthepresentsystemandaminorityhadideastohelpdealwithsomeoftheircomplaints.Thus,itwassaidthattherearesomanypapers published each year that it is impossible to keep up with developments in theliterature, somaybe it is timetodosomethingabout thecurrentpreoccupationwith thisstellarresearchoutput. Itwasarguedbya fewthat itmightbebettertomovetowardsasystem inwhichmore results aremadeavailable throughdatabases (and credit given toPrincipal Investigators and students for doing this) and figures shown in journals aredepositedinsearchabledatabasesthatallowthedatatobeusedbyothers.Also,papersthatarepublishedintopjournalscouldbeadverts/interfacesthatbringtogetherlargerbodiesofresearch in a usefulway that adds value to thedata andexplains keydevelopments andtrends.Thiswouldbebetterthansalamislicingkeyresults intoseparatecommunications.OneECRgaveaverygoodexplanationwhyallthismightnotwork:“IfIwereto‘goitalone’andimplementsuchastrategy(givemyresultsandideasawayunrecognizedindatabasesandonlypublishafew,longerpapersthatwouldbetrickytogetintotopflightjournals)Iwouldnotexpecttogetpromotedanditwouldbedifficulttogetfunding”.

FrenchECRs, though, are verypessimistic about the future, especially regarding thepacerequiredandthepressuretopublishinhighrankedjournals.Theydonotbelievethatevenopenaccessisgoingtoimproveorchangethesystem.Theyarealsoworriedthatcreativityandoriginalityhavenoplaceanymoreandthattheyhavetopublishevenunconsolidatedresultstostayintherace.Theyarguefora‘slow’sciencethatwillsaveitfroma collapse.Theytinkerwiththerulesasbesttheycanandtrytoincorporatenewwaysofbehaviourinthetraditionalbehaviouralset,buttheydonotseethemselvesasthegeneration thatwillchangeanything.Harbingerstheythinktheyarenot.

FewMalaysianECRshaveanswerstothevariousproblemstheyperceiveincurrentmodesofscholarlycommunication,butafewdidarticulatethefoundationsofamodel,whichseessharingandtransparency(nottomentionsecurity)atitsheartandholdsthekeystoabetterscholarlyfuture.Chineseresearchersarenotveryforthcoming,astheyfeltthattheyarenotinapositiontocomment,beingtoo juniorfortheiropinionstobetaken intoaccount–asentimentfeltinothercountries,too.

Finally, as to the hypothesis being tested in respect to transformations: The system isunchangingandunbending,butthereislittleevidenceofthedesireforchangeamongECRs.Theonlycountriesthatdisagreedwiththehypothesis–thatis,providedplentifulevidenceofadesireforchange–weretheUKandUS.USECRsarevocalindesiringchange,with22sayingso;intheUK14arealsoofthisopinion.OnthewholetheUSECRsaremoredefiniteand revolutionary aboutwhat changes they require. For instance, one said: “Researchersarefundamentallyprovokingthatchange.Theyaresharingpapersillegallyanditischanging

48

the system, like Napster. It is a reality. Journals will have to evolve and change. It ishappening now”. Another said: “Currently, there are more ways to communicate withpeople thanhasbeenrecorded inourhistory,sowiththatwedohaveanopportunity tochangehowcommunicationofscholarlyworkscontinues”.Othersfeelthatthepresenceofsocialmedia is the tippingpointordisrupter.UK researchersusually answer yesorno inregardtochangeagents,althoughopenaccessissaidtobehavingtraction(seeOAsectionof report for more details), scientific rigour is not being upheld in the new informationorder,andchangewillbedrivenbythefunders.However,wehavetobecarefulnottoviewtheuseofsocialmediaand thewish topublishOA (quite common) as twoaspectsof anoverallwishtotransformthesystem.Forexample,anECRwhoiskeenontransformationwashostiletosocialmedia.

Overall though, certainly in the caseof theUK/US,we seem tohavemovedon from thesituationwefoundinapreviousresearchprojectthatweconductedafewyearsago,whichsought todiscoverwhether thedigital transitionhad led to changes in thewayacademicresearchersplacedtheirtrustinscholarlycommunications(Watkinsonetal.,2016).There,open comments showed no one had any ideas at all about change, never mindtransformation,and thosewhodisliked thepresent situation just railedagainst itwithnopositive intent. Three years later we do find ideas for change, and even some fortransformation. Admittedlyweare now talking about ECRs, but it is our hypothesis thatECRs might be where transformation may start. Our study shows that at least someresearchers,who happen to be ECRs, are thinking about change and transformation andthesearenottheembitteredminority–thosewhoarenevergoingtoprogress–butsometopyoungresearchers.

8.15Countrycomparisons

Wehavereferredthroughoutthereporttothespecificdifferencesbetweencountries;wenowlookatsimilaritiesanddifferencesinthebroad.Acorrelationanalysisofthehypothesistestsshows(Table8)that:theUKismuchliketheUS;SpainismuchlikePoland;Franceisquite likePoland;China isnotdissimilar to FranceandSpain,butdoesnothavemuch incommonwiththeUK/US.MalaysiaisverymuchtheoddoneoutandacontrarianinrelationtoFranceanditispossiblethatalittleoftheclosenessbetweentheUKandUScouldbeputdowntothefactthatthesamepersoninterviewedthemboth.

Table8:Similaritiesbetweencountriesbasedonhypothesistests

PO SP 0.7 MostAlikeUK US 0.7 FR PO 0.6 CH SP 0.5 CH FR 0.5 PO US 0.4 FR UK 0.4 CH PO 0.4

49

FR US 0.4 FR SP 0.4 SP US 0.4 ML UK 0.4 PO UK 0.3 ML SP 0.3 ML PO 0.3 SP UK 0.3 CH US 0.2 ML US 0.1 CH UK 0.1 CH ML 0.1 FR ML 0.0 LeastAlike

Asforclusters(Figure1),thereisaUK/US'specialrelationship';EUcountriessort-ofcluster;but there is no 'Asian' cluster, with China closer to EU than to Malaysia. Perhaps notsurprisingly,becauseoftheircommonlanguageandsimilarcultures/institutions,UKandUSECRsholdverysimilarviewsandarequitedifferentfromtheothercountries,whichmightbethevariationsinsizeanddifferentcompositionofECRpopulations,whichwillbethesubjectoffurtherinvestigationastheprojectproceeds–seeTable9.However,intheUKinsomecases theinfluenceoftheResearchExcellenceFramework(REF)wasparamount.Thus,forexample, in theUKECRsaremuchmore likely tobe interested in influencing thegeneralpublic, something encouragedby theREF. Theexplanations for other countrydifferencescouldalsobepartlydowntothedifferentprofilesoftheECRpopulationineachcountryandlocal factors.Thus, in thecaseofMalaysia, thedifferencescouldbeputdown to theagefactor(alltheirECRswereintheirthirtiesbecauseoflocalconditionsandthefactthattheyneedtohaveaPhDfirstinordertobecomeanECR).

Figure1:Countryclustersbasedonhypothesistests

50

It is interesting to see that the consensus (positive or negative) regarding thehypothesestendstoderive fromtheglobalizedaspectsofacademia.Thus, theveryhighdegreeofagreementobtainedforthestatement‘Newbehavioursarenotreallytakinghold,whileacademicsaretypicallyrecruited,promotedandobtainfundingonthebasisoftheirpublicationrecordandcitationscores’ istheresultof internationalevaluation policies. The lack of consensus seems to come from the variety anddifferencesof thenationalenvironments.Takethestatement ‘Theydomanythingson a project (multi-taskers)’. Typically, this a locality issue, it depends on howscientific activity is organized and managed from one country to another, oneinstitution to another. ECRs clearly have to jugglewith local and global factors, sotheirstruggleisadifficultonebothonthelocalandglobalfronts.

Table9:CountryECRprofilesbroadlycompared

Country No. Subject Gender Age PhD InstitutionsChina 13 Science:40%

Soc.Sci.:30%Female:46%Male:54%

Twenties:46%Thirties:54%

8%Doctoralstudents

6

France 14 Science:82%Soc.Sci.:18%

Female:35%Male:65%

Twenties:65%Thirties:35%

100%Postdocs 4

Malaysia 12 Science:58%Soc.Sci.:42%

Female:50%Male:50%

Thirties:100% 100%Postdocs

5

Poland 10 Science:80%Soc.Sci.:20%

Female:40%Male:60%

Twenties:40%:Thirties:60%:

50%Doctoralstudents

1

Spain 18 Science:78%Soc.Sci.:22%

Female:44%Male:56%

Twenties:40%Thirties:60%

28%Doctoralstudents

16

UK 21 Science:62%Soc.Sci.:38%

Female:38%Male:62%

Twenties:24%Thirties:76%

33%Doctoralstudents

20

USA 28 Science:79%Soc.Sci.:21%

Female:41%Male:59%

Twenties:27%Thirties:73%

34%Doctoralstudents

28

8.16Morediversity

ThereisevidencefromfourcountriesthatECRswhohavereviewingexperienceholddifferent scholarly views from those who do not, perhaps because they are morefamiliarwith theworkingof thesystem.Thus, inSpainolderandmoreexpertECRsnaturallyenougharemoreforthcomingin interviewsandaremorecriticalaboutanassessmentsystemfocusedonlyonpublicationsinalimitednumberofelitejournals.

51

However,whentheyparticipateinthepeerreviewmechanismthemselvestheyfeelalotlesscritical.Acase,maybe,ofbeingconditionedbythesystem.Therearealsosignificantdifferencesinfivecountriesbetweenthosewhoworkmoreor less on their own, usually undertaking a doctorate after preliminary experience,andthosewhoareembeddedingroups.Theformertendtobesocialscientistsandasa generalisation they provide fewer answers to the questions asked and are lessproductive in terms of papers published. Indeed, most of them (though not all ofthem)arebasicallydisinterested in scholarly communicationandmoreof themareprobably not going to continue in academic life. Also, Spanish ECRs who work instrongresearchgroupsfeelmoresecureabouttheirprospectsandtendtobehappierwith the academic communication process, perhaps just because they are moreoptimisticabouttheirfuture.

The French study, like the Malaysian one, found differences according to subject.Thus,manyECRsclearlyexplainedthatsomeresearchtopicscouldturnouttobemoreor less “bankable” than others. Some topics are more “in the spirit of time” andothers are less so. The consequence is that those who have “bankable” researchsubjectsaremorevisible,theirresultsaremorelikelytobepublished,andtheyaremorecontactedbycolleagues in their countriesandabroad. It’sakindofMattheweffect, whereby eminent scientists get disproportionately greater credit for theircontributionswhilerelativelyunknownscientiststendtogetdisproportionatelylittlecreditforcomparablecontributions(Merton,1968).In theFrenchstudyageandexperiencewereclearlycorrelated,andareaddedvalues forECRs,whichcontributetowardsagreaterunderstandingofthesystemandtheknowhowastohowtobehaveandwhattodecideinmanysituationsandcontexts.IntheUS,anumberofECRs(perhaps,15%)workedinwhatonemightcallservicecapacities,usually in medicine. They offered techniques and methods. Their attitudes showeddifferences from thosewhose researchwaspurerand lessapplied.Again in theUS therewasanothergroupofaboutthesamesizewhowereeitherinindustryoringovernmentormedicallaboratories.Thenatureoftheirresearchwasdifferentandagaintheywerecutofffrom some of the concerns of the Academy: their attitudes were different, too. A smallpercentageofthesetypesofresearcherswerepartoftheUKcohortalso.Withregardtogender,femalesrepresentedoverone-thirdofallinterviewees,buttheserialinequalitiesuncoveredbySugimotoetal.(2013),especiallyinregardtoauthorship,didnotcropupininterviews,althoughfurtheranalysiswillbeconductedinthisregard.Withrespectto careerprogression,no issueswere raised, althoughChinese femaleECRsappear tobedrivenmorebypuresubjectinterestthantheirmale,promotion-drivencounterparts.Mostofourinterviewerswerewomen.

52

9.0Interimconclusionsandreflections

Itwouldbeprematureandhastytomakefirmconclusionsbecauseweareatanearlystageinapotentially lengthystudy.Thestudy isallaboutchangeand,sofar,wehaveonly laiddownthefoundationstone.Themaininterestliesinmappingchangeintheyearstocome.Not wanting to duplicate the executive summary and because of the dangers inmakinggeneralizations across an immature, complex, variable, and possibly dynamic dataset,wehavechoseninsteadtoreportonthereflectionsofnationalleads,thepeoplewhodidtheinterviews, in regard towhat they think they have discovered in relation to the project’saimsandhypotheses.

Independentofdisciplineornationality,theseresultsshowclearlythetensionsthatoccurina context of transition. In this transition, we have signs that scholarly ‘things’ (practices,behaviours, representations, wishes, objectives) aremoving inmany directionswhile theformal frame of evaluation and competition is strengthening. Some of the apparentcontradictoryresultsarerelatedtothesetensions.ECRsseetheopportunitiesofchange,butdonottaketheopportunitytodosobecausetheyjustdonothavethetimeandspaceinaninsecureandbusyenvironment.Thisisalso,possibly,thereasonwhyharbingersofchangearethinontheground,forthemoment.

CHINA

ThereareseveralstrongleadscomingfromtheChinesestudy.Firstly,inregardtoreputationandevaluation, there isnoevidence thataltmetricswill replace traditionalmetrics in thescholarly communication system, unless things change radically. On the contrary, ECRsbelievethatthecurrentevaluationsystem,whichisbasedonhighIF journals,workswell,becauseitisopenandfair.Second,inrespecttothedominanceofpeerreviewedjournals,astheirmentorsdo,ECRsreadandcitepeer-reviewedjournalsandpublishtheirfinalfindingsin these journals, although they do use smartphones and social media to publicize theirarticles, follow scholars and topics that interest them, and to contact their friends andcolleagues.Third,ECRsknowlittleaboutOAandopenscienceor,indeed,showlittleinterestinthesetopics.TheywillnotpublishpapersingoldOAjournals,becausetheycouldnotaffordthe cost, nor will they upload their papers to institutional repositories, since it does notgenerateanycreditfortheirstaffassessments.Fourth,ECRsdofeelthatsomethingsneedtochangeinthecurrentsystem,suchasthefixationwithWebofScienceindexedjournals,buttheydonotthinktheyarethepeoplewhoarecapableofmakingthechanges.Theybelievethatreformneedstocomefromthetop,notthebottom.Lastly,comparedtotheirseniors,ECRs: a) feel muchmore career and financial pressures; b) they are more willing to co-operatewiththeirpeersandhavemoreinteractionwithinternationalresearchers.

53

FRANCE

ECRsinFrancearegettingolder,astheyareinmanyothercountries,asittakesmoretimetoberecruitedandobtaintenure.Inrespecttoscholarlybehaviour,theybehavestrategicallyandconservativelyallthetimewhateverthekindofscholarlyactivitytheyareinvolvedin.Theyareunlikelytomoveawayfromcitationmetricsbecauseofgreaterlevelsofcompetitionandevaluation.FortheseedsofchangeinFranceweneedtolookat:a)ResearchGateandGoogleScholar,whicharebecomingthemajorsourcesofinformation(discoveringarticlesandobtainingPDFs);b)Smartphones:althoughonlyafewECRshavesmartphonesandusethemforscholarlypurposes(forcurrentawarenessandreading),itispossibletoseethatitwillnotbelongwhenallECRswillwanttobeconnectedandlinkedallthetime,inrealtime;c) Globalization(mobilityandcollaboration,mostnotably),whichiswidelyadmittedtobeakeyarea,muchdesiredandtargetedbyECRs;d)libraries,becausetheyarelosingvisibility:e) thedesire,butnotyetpractice, indisseminating research findings in less formalways,thoughnotnecessarilyviasocialmediaoutlets,whicharecurrentlynotpopular;f)disruptioncaused by ECRs getting closer to their peers and more detached from their homeinstitutions,courtesyofonlinecommunitiesandincreasedopportunitiesforcollaboration.OpenaccessisseenpositivelybymostECRs,butitisnotconsideredasagamechanger.MALAYSIAAstrongdrivetogetpublished in indexed journalshasplayedamajorrole inshapingthescholarly communication landscape in Malaysia and especially so for research-intensiveuniversities. Ingeneral,then,MalaysianECRs'behaviour isstillverytraditionalandmainlypaper-drivenwithafocusonproductivityandimpactindicatorsgarneredfromWoSand/orScopusdatabases.Scientists (comparedtonon-sciences)arenotablymorestrategicaboutwheretheypublishandaremoreinterestedinself-promotion.Theymakeuseofscholarlymetrics, but are more concerned with traditional metrics than altmetrics. ECRs want topublishintheirpreferredjournals,whetherornottheyareopenaccessorsubscription-based,buttheydemonstrateanincreasingopennesstosharingandadesireforthesupporttomakethat possible. ECRswhopublish in eitheran open access journalor a hybridopen accessjournalmakesurethatthejournalsareindexedbyeitherWoSorScopus.ECRsare,though,concernedaboutarticleprocessingcharges(APCs)astheirinstitutionwillonlybearthecostofAPCs for“those journals thatare inQ1ofWoS”andtherearenot thatmanyof them.Althoughtheydonotseriouslyusesocialmediaplatformsforscholarlypurposes, theydoseesocialmediaasusefulforresearchpurposes.ECRsagreethatonlinescholarlynetworkslead to greatercollaborationand/orconnectivity,andhelpbuildreputation.Andtheyareconcerned with digital visibility and are encouraged tomake their academic profiles andresearchmetricsopenlyavailablethroughResearcherIDandORCID.Thosewhodidnotusethemfeltthattheyshouldmakemoreuseoftheopportunitiespresentedandmightdosoinfuture. They may want to use social media more, but traditional norms that dominatescholarlybehaviourperhapspreventthemfromdoingso.

54

POLAND

Itseemsthat,reflectingonallthedatacollected,PolishECRsarethemostunlikelytochange,that is they will not be the harbingers of change or anything remotely like that. Theirbehaviourisveryconservative,withPolishECRseschewingsocialmedia,onlinecommunities,altmetrics and smartphone use for scholarly communication purposes. Changing, eventinkeringwith,thescholarlysystemisfarfromtheirthoughts.Threeimportantfactorshelptoexplainthis:1)inPolandthereisnosuchthingasresearch-intensiveuniversities.Allstateuniversitiesandallfacultyhavetodohalfteachingandhalfresearchandthatmeanstheydoalotofjobsinadditiontoresearching.Itisalevelplayingfield;2)allacademicstaffhavetodo a Habilitation after their PhD if they want to keep their job (Habitation involvespublishingamonographandthispartlyexplainsthefollowingpoint);3)theyaremuchlessjournalfocusedintheirdisseminationactivities,embracinginparticularmonographs,bookchaptersandconferenceproceedings.While Polish ECRs seemunchanging, albeit because they areparticularly overworked andhavenotimetothinkaboutchange,theyandtheiremployersareawareofthefactthattheywillhavetofollow,andcannotignore,whatishappeningtotheUS/Westernscholarlymodel.Theyneedtokeepupwithinternationalrankingsifnothingelse.Indeed,thereisadesirethatshouldhappenandprovideafairerandmoreappropriatesystem.ThenecessitytopublishininternationaljournalswithhighIFsisarealproblembecauseofthelanguagebarrieranddeliversabigreputationaladvantagetonativeEnglishlanguagespeakers.Anotherconcernisthatifsocialmediaishavingsobiganinfluenceinourdailylife,whynotinscience?SPAIN

TheECRsinterviewedare,liketheirUS/UKcolleaguesaverymixedgroup,andtheyarealsogettingolderbecauseoftheseriouslackoftenuredpositionsinSpainoverthepastfiveyearsdue to theeconomicrecession.Thisshowsupinthefindings.ECRsinSpainare,likePolishECRs,probably,someof themostconservative.Theprominent findings forSpainarethatECRsareveryhardworkingpeople,veryinvolvedinthecompetitionforobtainingatenuredpositionorthefirst/nextpost-doc.Theyare,probably,moreworriedanddedicatedtothejobthanUS/UKECRswhohavegreater(other)jobopportunities.Theyknowthattheevaluationsystemisfocusedonpublishinginhighrankjournalsandplaythegame,eventhoughtheyconsideritunfair.ECRswouldmuchpreferamorecomprehensiveevaluationprocessthattakesintoaccounteveryachievementandthedifferencesbetweenscientificareas,butthiswillonlyhappenwhentheeconomicpressureisoffandmorejobsaregenerated.

They are very concerned about pushing science forwardandhaving real impact on theircommunitiesandindustries.ECRsconsiderknowledgetransferanddisseminationtoindustryandsocietytobeveryimportant,butdonothaveenoughtimetoworkonit.Theywouldliketoincreasetheirpresenceinscholarlysocialmediaasameanstoachievethisaim.Theirnextgoalistousereputationaltools.Openaccessisseenasapositiveinnovationandasameans

55

forchangingthesystem,butaspublishinginOAjournalsisexpensive,theyfeelthatitsuseismakingtheplayingfieldunevenbetweenthoseresearchgroupsthatcanpayforitandthosethatcannot.Theyarebigadvocatesof‘proper’doubleblindpeerreview,butwouldlikemoretransparencyand,aspartofthis,theyfeelthatreviewers'namesshouldbereleasedattheendoftheprocess.

UK/US

UKandUSECRshavemuchincommon.TheECRsinterviewedareamixedgroupandthusitis not easy to generalize about their behaviour and attitudes. Some have been trainedoriginally in another country andare aware that theyhave to learn to fit in. Someneverreallyadjustandyoucanseethis.OtherswhohavebeenthroughtheUKorUSprocessarealsoclearlynotgoingtomakeitasrealresearchersandwillremainasasortofassistantormainlyateacher,andsomearebasically interestedinindustryorworkinginmedicalbackup services. We would expect only those who are really keen on research to discussseriouslyscholarlybehaviourorpointtotransformations.

Clearly, researchers tendtobeconservative, forverygoodreasons (nobodywants ‘hippy’science),andsomeofthosewhoareseriousresearchersareconsciouslyconservativeandhappywithcurrentscholarlycommunicationsystems.Thisdoesnotmeantheydonotusesocialmedia,buttheydonotuseitinwaysthatareconnectedwithchange,moreenhancingexistingpractices.Theoneswearereallyinterestedinaspotentialharbingersare,wesuspect,notmuchmorethanhalfofalltheECRs.Itisthesewemightexpecttobechanging,towanttoseechanges,andwillhavechanged inayear'stimeastheygetmoreopportunitiesforcarryingthroughtheirideas/principles.AlotoftheseECRstendedtoanswerourquestions"not yet". There is a fairminority (10%)who do have new ideas, newer than any of theinterviewees in the aforementioned study of trustworthiness in the digital age had(Watkinsonetal.,2016).Theyare trying to implement thesewhentheycan,which isnotoften,becausetheyareboundtopublishintopjournals,etc.Whatarethesenewideasthattheyhave?Theyareprincipallybasedaroundnewwaysofsharing,greatertransparencyandbuilding reputation. They appear not to worry about the dangers of sharing. They arekeenerondoing‘proper’research,butarenotkeenonthewaytheyhavetospendsomuchtimeondeterminingwhereandhowtodeviseresearchoutputs.Finally,webelievewehaveidentifiedawishforchange,whichmayforthemomentnotbepossiblebutmightbesometimesoon,fueledbythegeneralpushforopenness,includingoutreach.

Onreflection, thestrangest thingabout theUK/US interviews ishowrarelyECRsmentionpublishers(manyofwhomarebasedintheUK/US),althoughourquestionsledthetalktoalotoftheirpublications,andofcoursetheyknewwewerefundedbythem.InthesectionsrelatingtopeerreviewECRsnevermentionedthefrictioninthesystemwhichisthefeatureof publisher meetings currently – the ways in which the user experience is diminishedbecause of all sorts of obstacles to seamlessness and accessibility. All editorial onlinesystemsareknownforbeingdifficulttoworkwith.ECRsdonotmentionthis.Whynot?Isitbecausetheactualsubmission isusuallydoneby someoneelse– thegroupPAormaybe

56

someassistant?Wehavenoevidenceonthisyet,butwewillhave inthefuture.Theydonotblameorpraisepublishersforpeerreviewpolicies.Itisalwaystheeditororreviewerorthesystem.Theyalsodonotseemtoexpecttogetmuchfeedbackfromthepublicationofpapers,thoughsomedomentionpostpublicationreview.ELifeandPeer1000ResearchdomakeabigdealaboutpostpublicationfeedbackasdidPLOS,butpublishersgenerallydonot.As we know from the Sloan Trust study and occasionally in the current study, gettingfeedbackcomesfrominternaltalks,etc.,conferencepresentationsandposters,butpapersare not mentioned in this context. As one ECR said in a blog following a recent ALPSPseminar“Ifindthatpublisherswanttoknowwhatwewant,andweshouldtellthem”:thisreportshouldhelpinthisregard.

57

10.0References

Bar-Ilan,J.,Haustein,S.,Peters,I.,Priem,J.,Shema,H.andTerliesner,J.(2012)."Beyondcitations:Scholars'visibilityontheSocialWeb".arXivpreprint.arXiv:1205.5611.

Bennion,A.andLocke,W.(2010).Theearlycareerpathsandemploymentconditionsoftheacademicprofessionin17countries.EuropeanReview,1These8(S1),S7-S33.

Brechelmacher,A.,Park,E.,Ates,G.,andCampbell,D.F.(2015).Therockyroadtotenure–careerpathsinacademia.InAcademicworkandcareersinEurope:Trends,challenges,perspectives,13-40.SpringerInternationalPublishing.

Belluz,J.,Plumer,B.,andResnick,B.(2016).The7biggestproblemsfacingscience,accordingto270scientists.Availableat:http://www.vox.com/2016/7/14/12016710/science-challeges-research-funding-peer-review-process

Chudziak,J.(2015).Digitaltechnologyvision2015anditshumandimension.The3rdScientificConferenceInformationScienceinanAgeofChange,11-12May2015.UniversityofWarsaw.

Cusick,A.(2015).Researchtrainingasoccupationalsocialization:Doingresearchandbecomingresearchers.AsianSocialScience,11(2),252.

Foote,K.E.(2010).Creatingacommunityofsupportforgraduatestudentsandearlycareeracademics.JournalofGeographyinHigherEducation,34(1):7-19.

Fransman,J.(2014).Becomingacademicinthedigitalage:NegotiationsofidentityinthedailypracticesofEarlyCareerResearchers.ConnectedCommunitiesandEarlyCareerResearchersworkshop,CityUniversity,May2014.Availableat:https://www.srhe.ac.uk/downloads/FRANSMAN_Final_Report.pdf

Friesenhahn,I.andBeaudry,C.(2014).Theglobalstateofyoungscientists–Projectreportandrecommendations.Berlin:AkademieVerlag.

Graham,H.,Hill,K.,Matthews,P.,O'Brien,D.andTaylor,M.(2014).Connectingepistemologies:MethodsandearlycareerresearchersintheConnectedCommunitiesProgramme.Availableat:https://earlycareerresearchers.files.wordpress.com/2014/10/connecting-epistemologies-report.pdf

Gu,J.,Lin,Y.,Vogel,D.andTian,W.(2011).WhatarethemajorimpactfactorsonresearchperformanceofyoungdoctorateholdersinscienceinChina:aUSTCsurvey.HigherEducation,62(4),483-502.

Halevi,G.,Moed,H.andBar-Ilan,J.(2015).Accessing,ReadingandInteractingwithScientificLiteratureasaFactorofAcademicRole.PublishingResearchQuarterly,31(2),102-121.

Harley,D.,Acord,S.K.,Earl-Novell,S.,Lawrence,S.,andKing,C.J.(2010).Assessingthefuturelandscapeofscholarlycommunication:Anexplorationoffacultyvaluesandneedsinsevendisciplines.UCBerkeley:CenterforStudiesinHigherEducation.Availableat:https://escholarship.org/uc/item/15x7385g

Housewright,R.,Schonfeld,R.C.andWulfson,K.(2013).IthakaS+RUSfacultysurvey2012.Availableat:http://lgdata.s3-website-us-east1.amazonaws.com/docs/923/721668/Ithaka_S_R_US_Faculty_Survey_2012_FINAL.pdf

Horta,H.,andSantos,J.M.(2015).TheimpactofpublishingduringPhDstudiesoncareerresearchpublication,visibility,andcollaborations.ResearchinHigherEducation,1-23.

58

James,L.,Norman,J.,DeBaets,A.S.,Burchell-Hughes,I.,Burchmore,H.,Philips,A.,Sheppard,D.,Wilks,L.andWolffe,J.(2009).Thelivesandtechnologiesofearlycareerresearchers.JISC;CARET,UniversityofCambridge;TheOpenUniversity,UK.Availableat:http://www.webarchive.org.uk/wayback/archive/20140614204612/http://www.jisc.ac.uk/publications/reports/2009/earlycareerresearchersstudy.aspx

Jones,P.(2014).PhillJonesonthechangingroleofthepostdocandwhypublishersshouldcare.Blogpost,October6,2014.TheScholarlyKitchen,Availableat:https://scholarlykitchen.sspnet.org/2014/10/06/guest-post-phill-jones-on-the-changing-role-of-the-postdoc-and-why-publishers-should-care/

Laine,H.(2015).Thecaseagainstthejournalarticle:Theageofpublisherauthorityisgoing,going,gone—andwe’llbejustfine.Blogpost,July14,2015.LSE-LondonSchoolofEconomicsandPoliticalScience.TheImpactBlog.Availableat:http://blogs.lse.ac.uk/impactofsocialsciences/2015/07/14/the-case-against-the-journal-article/

Laudel,G.andGläser,J.(2008).Fromapprenticetocolleague:Themetamorphosisofearlycareerresearchers.HigherEducation,55(3),387-406.

Madhusudhan,M.(2015).UseofMobileDevicesforImprovingtheResearchWorkbyResearchScholarsoftheUniversityofDelhiandUniversityofHyderabad:AStudy.WorldDigitalLibraries-Aninternationaljournal,8(2),127-144.

Merton,R.K.(1968).TheMattheweffectinscience.Science,159(3810),56-63.

Müller,R.(2014a).Racingforwhat?Anticipationandaccelerationintheworkandcareerpracticesofacademiclifesciencepostdocs.In:ForumQualitativeSozialforschung/ForumQualitativeSocialResearch,15(3).

Müller,R.(2014b).Postdoctorallifescientistsandsupervisionworkinthecontemporaryuniversity:Acasestudyofchangesintheculturalnormsofscience.Minerva,52(3),329-349.

Mulligan,A.andMabe,M.(2011).Theeffectoftheinternetonresearchermotivations,behaviourandattitudes.JournalofDocumentation,67(2),290-311.

Mulligan,A.,Hall,L.,&Raphael,E.(2013).Peerreviewinachangingworld:Aninternationalstudymeasuringtheattitudesofresearchers.JournaloftheAmericanSocietyforInformationScienceandTechnology,64(1),132-161.

Nicholas,D.andHerman,E.(2016).Digitalnetworkswilltransformhowacademicsbuildreputations.ResearchFortnight,11May2016,20-21.

NicholasD.andRowlands,I.(2011).Socialmediauseintheresearchworkflow.InformationServicesandUse,31(1-2),61-83.

Nicholas,D.,Watkinson,A.,Volentine,R.,Allard,S.,Levine,K.,Tenopir,C.andHerman,E.(2014).Trustandauthorityinscholarlycommunicationsinthelightofthedigitaltransition:Settingthesceneforamajorstudy.LearnedPublishing,27(2),121-134.

Nicholas,D.,Jamali,H.R.,Watkinson,A.,Herman,E.,Tenopir,C.,Volentine,R.,Allard,S.andLevine,K.(2015a).Doyoungerresearchersassesstrustworthinessdifferentlywhendecidingwhattoreadandciteandwheretopublish?InternationalJournalofKnowledgeContentDevelopmentandTechnology,5(2),46-53.

Nicholas,D.,Watkinson,A.,Jamali,H.R.,Herman.E.,Tenopir,C.,Volentine,R.,Allard,S.,&Levine,K.(2015b).Peerreview:Stillkinginthedigitalage.LearnedPublishing,28(1),15-21.Nicholas,D.,Herman,E.andJamali,H.R.(2015c).Emergingreputationmechanismsforscholars:Aliterature-basedtheoreticalframeworkofscholarlyactivitiesandastate-of-the-artappraisalof

59

thesocialnetworkingservicesusedbyscholars,tobuild,maintainandshowcasetheirreputation.EuropeanCommission,JointResearchCentre,InstituteforProspectiveTechnologicalStudies.Availableat:http://publications.jrc.ec.europa.eu/repository/bitstream/JRC94955/jrc94955.pdf

Nicholas,D.,Herman,E.andJamali,H.R.(2015d).Analysisofemergingreputationmechanismsforscholars.In:Vuorikari,R.andPunie,Y.(Eds).AnalysisofemergingreputationandfundingmechanismsinthecontextofOpenScience2.0.Part1,3-72.EuropeanCommission,JointResearchCentre,InstituteforProspectiveTechnologicalStudies.Availableat:http://publications.jrc.ec.europa.eu/repository/bitstream/JRC94952/jrc94952.pdf.Poli,S.(2016).TheannualconferenceoftheGermanAssociationforHEResearch.Blogpost,May10,2016.SRHE-TheSocietyforResearchintoHigherEducationBlog.Availableat:https://srheblog.com/2016/05/10/the-annual-conference-of-the-german-association-for-he-research/

Poli,S.(2016).TheannualconferenceoftheGermanAssociationforHEResearch.Blogpost,May10,2016.SRHE-TheSocietyforResearchintoHigherEducationBlog.Availableat:https://srheblog.com/2016/05/10/the-annual-conference-of-the-german-association-for-he-research/

Procter,R.,Williams,R.,Stewart,J.,Poschen,M.,Snee,H.,Voss,A.andAsgari-Targhi,M.(2010)."AdoptionandUseofWeb2.0inscholarlycommunications".PhilosophicalTransactionsoftheRoyalSocietyA:Mathematical,PhysicalandEngineeringSciences,368(1926),4039-4056.

PublishingResearchConsortium(2016).Peerreviewsurvey2015.MarkWareConsulting.Availableat:http://publishingresearchconsortium.com/index.php/prc-projects/peer-review-survey-2015

RIN(ResearchInformationNetwork)(2010).IfYouBuildIt,WillTheyCome?HowResearchersPerceiveandUseWeb2.0.AResearchInformationNetworkReport.Availableat:http://www.rin.ac.uk/system/files/attachments/web_2.0_screen.pdf

Rowlands,I.,Nicholas,D.,Russell,B.,Canty,N.andWatkinson,A.(2011).Socialmediauseintheresearchworkflow.LearnedPublishing,24(3),183-195.

Shin,J.C.andCummings,W.K.(2013).Teachingandresearchacrosshighereducationsystems:Typologyandimplications.In:Shin,J.C.,Arimoto,A.,Cummings,W.K.,&Teichler,U.(Eds).TeachingandResearchinContemporaryHigherEducation:Systems,ActivitiesandRewardsVol.9.SpringerScience&BusinessMedia,pp.381-394.

Shneiderman,B.(2008).Science2.0.Science,319(5868),1349-1350.Sinclair,J.,Barnacle,R.,&Cuthbert,D.(2014).Howthedoctoratecontributestotheformationofactiveresearchers:Whattheresearchtellsus.StudiesinHigherEducation,39(10),1972-1986.

Sugimoto,C.R.,Lariviere,V.,Ni,C.Q.,Gingras,Y.,&Cronin,B.(2013).Globalgenderdisparitiesinscience.Nature,504(7479),211-213.SpringerNaturePublishingGroup(2015).NaturePublishingGroupreleaseslandmarkWhitePaper-TurningPoint:ChineseScienceinTransition.Availableat:http://www.springernature.com/gp/group/media/press-releases/nature-publishing-group-releases-landmark-white-paper/6630926

Teichler,U.andCummings,W.K.(2015).Forming,recruitingandmanagingtheacademicprofession:Avariedscene.In:Teichler,U.andCummings,W.K.(Eds).Forming,recruitingandmanagingtheacademicprofession,1-10.Springer.Tennant,J.(2016).Breakingthetraditionalmouldofpeerreview:Whyweneedamoretransparentprocessofresearchevaluation.http://blogs.lse.ac.uk/impactofsocialsciences/2016/02/17/breaking-

60

the-traditional-mould-of-peer-review/

Tenopir,C.,Allard,S.,Bates,B.,Levine,K.,King,D.W.,Birch,B.,Mays,R.andCaldwell,C.(2010).Researchpublicationcharacteristicsandtheirrelativevalues:AreportforthePublishingResearchConsortium.CenterforInformationandCommunicationStudies,UniversityofTennessee.Availableat:http://www.publishingresearch.net/

Tenopir,C.,Allard,S.,Bates,B.,Levine,K.,King,D.W.,Birch,B.,Mays,R.andCaldwell,C.(2011).Perceivedvalueofscholarlyarticles.LearnedPublishing,24(2),123-132.

Tenopir,C.,Volentine,R.andKing,D.(2013).Socialmediaandscholarlyreading.OnlineInformationReview,37(2),193-216.

Tenopir,C.,Levine,K.,Allard,S.,Christian,L.,Volentine,R.,Boehm,R.,Nichols,F.,Nicholas,D.,Jamali,H.R.,Herman,E.andWatkinson,A.(2015),Trustworthinessandauthorityofscholarlyinformationinadigitalage:Resultsofaninternationalquestionnaire.JournaloftheAssociationforInformationScienceandTechnology.DOI:10.1002/asi.23598

VanDalen,H.P.andHenkens,K.(2012).Intendedandunintendedconsequencesofapublish-or-perishculture:Aworldwidesurvey.JournaloftheAmericanSocietyforInformationScienceandTechnology,63(7),1282-1293.

Veletsianos,G.andKimmons,R.(2012).AssumptionsandchallengesofOpenScholarship.TheInternationalReviewofResearchinOpenandDistanceLearning,13(4),166-189.

Waaijer,C.J.,Macaluso,B.,Sugimoto,C.R.andLarivière,V.(2016).StabilityandlongevityinthepublicationcareersofUSdoctoraterecipients.PloSone,11(4),e0154741.

Watkinson,A.,Nicholas,D.,Thornley,C.,Herman,E.,Jamali,H.R.,Volentine,R.andTenopir,C.(2016).Changesinthedigitalscholarlyenvironmentandissuesoftrust:Anexploratory,qualitativeanalysis.InformationProcessing&Management,52(3),446-458.

Weir,K.(2011).U.S.DepartmentofEducationstatisticscitedinThenewacademicjobmarket.gradPSYCHMagazine,September.Availableat:http://www.apa.org/gradpsych/2011/09/job-market.aspx

WilliamsJune,A.(2016).Someday,Altmetricswillnolongerneed'Alt'.TheChronicleofHigherEducation,July18,2016.SpecialReport.Availableat:http://chronicle.com/article/Someday-Altmetrics-Will-No/237118?cid=at&utm_source=at&utm_medium=en&elqTrackId=b54ac59fb55040ee87830472d3b673ac&elq=50e40719099e4514a7c8f09393c12857&elqaid=9906&elqat=1&elqCampaignId=3641

60

Appendix1:QuestionsforECRinterviews

62

1. Backgroundinformation

Hypothesestotest:Theydomanyjobsforshortperiodsoftime;theydomanythingsonaproject(multi-taskers).Theenvironmentinwhichtheyareisprecarious.Thereisabigdrop-outrate.

Researchjob/projectcurrentlyworkingon:

Q1. What research project(s) areyouinvolvedinatthemoment?Aretheyinter/multi-disciplinary?

Q2.Whatisyourrole/statusintheproject?Q3.Areyoupartofaresearchgroup/centre?Ifso,isyourgroupworkingwithothergroupsonthisproject,ifso,whatistheroleofyourowngroupandtherolesoftheothergroupsintheproject?Q4.If,sincereceivingyourdoctorateorbeforeyoustartedyourdoctorate,youhaveworkedinothergroupswhatwasyourroleinthosegroups?Q5.Ifyourcurrentresearchisnotpartofagroup,doyoustillworkwithotherresearchers?Inwhatways?Q6.Howwouldyoudescribethecurrentandpreviousgroupsyouhavework(ed)in?Couldyoucharacterisethemastopinternationalgroups,groupsofinternationalstanding,wellthoughtofgroupsworkingmainlyatanationallevel?Inwhatwayshasyouraffiliationwiththesegroupsinfluencedyourcareer?

[Mentoring/training]Q7.Howwouldyoudescribethequalityofyourmentoring?[Whodoyouturntoforadviceandhowgoodaretheyatansweringyourquestions?]Q8.Dotheorganisationsyou

haveworkedformakespecialprovisionsforECRs,forexample,inrespecttotraining?

2. Careeraims

Hypothesistotest:Gettingagoodjobisthemajormotivation,notchangingtheworld/science.

Q1.Areyouaimingtohaveacareerasauniversityresearcher?Q2.Whereareyouinyourcareerdevelopment/progression?Q3.Mightyouconsiderresearchingoutsidetheacademy,inindustry,forexample?Q4.Doyouseeyourresearchactivityendingwithyourcurrentpostorwillyoucontinueresearchinginanotherpost?Q5.Whatisyourmainmotivationfordoingyourcurrentresearch?Forinstance,addingtoknowledge,havinganinterestingcareerorawell-paidcareer?

3. General(scholarly)

communication

behaviour

Hypothesestotest:Earlycareerresearchersadoptthepracticesoftheirmentorsandheadsofthegroupstowhichtheybelong.Newbehavioursarenotreallytakinghold,whileacademicsaretypicallyrecruited,promotedandobtainfundingonthebasisoftheirpublicationrecordandcitationscoresbasedonaccumulatedreputation.

63

Q1.Whatareyourscholarlycommunicationspracticesinrespecttodissemination,citingandreadingresearch?[Gointodetailaboutthethreeindividualactivities.]Dotheydifferfromthoseofyourresearchmentorsincurrentandpreviousjobs?

[Changes]Q2.Asyouhaveprogressedthroughyouracademiccareerhaveyourattitudestowardsestablishedscholarlycommunicationbehaviourchanged?Ifso,whatarethemainfactorsthatinfluencedthechange(technologies,policies,peerinfluence/pressure,etc.)?

[Discovery/usage]Q3.Howdoyoufindthescholarlyinformationyouneed?Google,librarycatalogues,onlinenetworks,etc.?Q4.Doyousearchforandreadscholarlypapersonyoursmartphone?

4. Influenceofsocialmedia

andonlinecommunities

Hypotheses to test: Early careerresearchers would like to use socialmedia more, but traditional normsthat dominate scholarly behaviourpreventthemfromdoingso.ECRsdonot see social media as beingscholarly ‘noise’ but useful forresearch purposes. Social scientistsaremore favourable to thescholarlyuse of social media. Early careerresearchers are detached frominstitutions and more closelynetworked/connected with theirpeers.

[Socialmedia]Q1.Doyouusesocialmediainyourscholarlyactivities?a)Tofindout

informationand(ifso)fromwhatmedia?b)Doyoucitesocial(new)mediainyourdissertations,orarticlesorinblogs?c)Todisseminateyourresearchfindings/ideas/data?d)Toconnect/network/collaboratewithyourpeers?Q2. Are you encouraged to usethesocialmediainyourwork?Ifso, by whom (seniors,administrators)andforwhatpurpose?Ifnot,doyoustillusethem?Ifyoudo,why?Ifnot,whynot?Q3.Doyoufindthat,thankstothesocialmediaandonlinecommunityplatforms(e.g.,ResearchGate),youare:a)detachingfromyourinstitutions;b)gettingclosertoyourpeerselsewhere?c)both?Ifso,whatarethepracticalconsequencesofthis?

[Onlinecommunities]Q4.Isityourexperiencethatonlinecommunities,suchasResearchGate,giverisetoresearchcollaboration?Q5.Isityourexperiencethatonlinecommunitieshelpinbuilding/enhancingyourreputation?Q6.Arethenewvirtualgroupings,courtesyofonlinesocialnetworks,adifferentphenomenonfromthestructuredresearchgroups?Hasengagementwiththemenabledyoutodomoreoriginalresearch?

5. Authorship,

publishingandopen

access

Hypothesestotest:ECRstoe-theline.ECRsnotveryproductive.Notveryhappywiththeirlotasresearch‘apprentices’.UseOA

64

becausetheyareeasiertogetinto(alsoseeReputation).

[Authorship]Q1.Whatcontributionshaveyoumadetothepaperswhichyouhaveco-authored?Q2.Doesyourresearchteam/department/universityhaveanauthorshippolicy?Q3.Wouldyoudothingsdifferentlyifyouhadasayinthis?[Awardofcorrespondingauthorisanimportantissue;alsocronyism-partialitytolong-standingfriends.]Q4.Whatinfluence(ifany)haveyouhadonthechoiceofjournal?[OpenAccess]Q5.Doesyourresearchteam/department/universityhaveapolicyinregardtoOApublishing?[Probablyimposedbytheirfunders.]Q6.WhatdoyouthinkaretheadvantagesanddisadvantagesofOApublishingfromthepointofviewoftheauthor?Q7.DoyouthinkOApublishingadvancesscienceandresearch,orareyouworriedthatitwilldilutethequalityofpublications,ordoyouagree/disagreewithbothpropositions?Ifso,how?

[Innovating]Q8. Do you have a preferencefor journals with innovativefeatures, such as video articles(e.g., Jove), when placing yourresearch?

[Publishingstrategy]Q9.Istherepressureonyoutopublishinparticulartop-rankedjournalsand,ifso,howdoyouthinkthisaffectsscholarlycommunications,ingeneral,andyourcareer?Q10.Doyouhaveaconsciouspublicationstrategyrelatingtoyourresearchandisthattodowithobtainingatenured/establishedpositionand,ifso,pleasedescribe?Q11.Wouldyouprefertomake

publicyourresearchfindingsinlessformalways,suchasblogs,whichcouldmakethemmorevisible?

[Dataetc.]Q12.Ifyouhaveproduceddataorsoftwareinthecourseofyourworkandthishasbeenyourmaincontributionwouldyoulikethisaspectofwhatyouhavedonetoberecognisedandcreditedandwouldyoulikethedataitselftobemademorevisible?

6. Peerreview

Hypothesestotest:ECRsfeelalienated/lockedoutbytheexistingpeerreviewsystem,whichtheythinkofasaclosedgentleman’sclub.Preferdoubleblindpeerreviewbecauseitprovidesfairerappraisal.Earlycareerresearchersareworriedbytoomuchtransparencyinpeerreviewbecauseitwillmakeitdifficultforthemtocriticisethesubmissionsoftheirseniors.

Q1.Doyouhaveexperienceinrespondingtocommentsfrompeerreviewersonpapersyouhavewrittenand,ifso,howdidyoufindtheexperience?Q2.Haveyouyourselfbeenareviewerand,ifso,whatdidyoulearnfromtheexperience?Q3.Doyoufeelthepeerreviewsysteminitscurrentformisfairordoesitfailyouinanyway?Q4.Doyoufeelthatpeerreviewformostjournalsisinthehandsofestablishedresearcherswhoarenotalwayssympathetictonewideas?[Thisisaquestionaboutwhetherinnovationisbeingsuppressedbythepeerreviewprocess.]Q5.Doyoufeelpeerreview

65

couldbeimproved,andifso,how?Forexample,doyouthinkthatdoubleblindpeerreviewispreferableorwouldyoulikeallpeerreviewtobeopen?Q6.Mostpeerreviewisorganisedbypublishers.Doyouthinkthisisagoodideaordoyouthinkitshouldbedonebyotherentities–forexample,learnedsocieties?

7. Employment,

reputationandcareer

progression

Hypotheses to test: ECRs havelittlepersonalfreedomandsecurity.They are ‘slaves’ to a metric-based/journal focussed system,which they have to adhere to inordertoclimbtheacademicladder.ECRsmakeuseofsocialnetworkingsites inorder tobuildup theirownnetworks, separate from thenetworksalreadyestablishedbytheresearchgroupstheyworkinortheconnections of their mentors (seesocialmediaquestions).

[Employment]Q1.Inyourexperience,howareyoungresearchersemployed/treated?Isthepositionofyoungresearchersinyourpresentandpreviouspoststhesubjectofapolicymadeclearbytheinstitutionordepartment?Q2.Howareyoungresearchersevaluated?Whatarethecriteriaandaretheyobjective?Whatreforms,ifany,wouldyousuggest?

[Careerprogression]Q3.Haveyousufficientfreedomtodevelopyourcareeralongthepathyouwouldlike?Forinstance,towhatextentareyoufreetochoosethearea/topicsyouwouldliketostudy?

Q4. To what degree do you agree

withtheviewthatECRsare‘slaves’toametric-based/journalfocussedsystem to which they have toadhere to in order to climb theacademic ladder? If you agree, atleasttosomeextent,haveyouanyideas on how things can beimproved?

[openscience]Q5.DothetechnologicalinnovationsScience2.0/OpenSciencemeananythingtoyou,andifso,dotheyhaveanysignificanceforyou?Q6. Would you say that openaccess publishing or depositingyour material in institutionalrepositories can fast track yourcareer/build your reputation? Doyou habitually utilise then theseoptions?

8. Sharingand

collaborating

Hypothesestotest:Earlycareerresearchersshareandcollaborateextensivelyevenattheriskoflosingtheircompetitiveedge.

Q1.Inwhatwaysdoyoushareyour:a) ideas and interim researchresults;b) research findings, dataandpublications?Q2.Whatkindsofcollaborationareyouinvolvedwith?Q3.Doyouusethesocialmediaandonlinesocialnetworkstolookfor,buildandmaintaincollaboration?Doyougodifferentlyaboutlookingfor,buildingandmaintainingnational,asopposedtointernationalcollaboration?Q4.Isthereariskoflosingyourcompetitiveedgethroughsharingandcollaboratingextensively?

66

Q5.Isyoursharing/collaboratingbehaviourdifferentfromthatofyourresearchmentorsincurrentandpreviousjobs?Q6.Hasyourbehaviourinrespecttosharing/collaboratingchangedfromearlierinyourcareerandinwhatways?

9. Metrics

Hypothesestotest:ECRsareinterestedmoreinsocialmediaandusagemetricsbecausecitationstakesolongtocount.

Q1.How,ifatall,doyouemploycitationdata,usagedata,socialmediaindicatorsinyourdailyresearchwork(searchingforarticles,etc.),inpresentingyourresearch,identifyingleadingresearchers,etc.[Youmightneedtobreakthisintothreesections,citations,usage,socialmediaindicators]Q2.Howimportantdoyouthinkmetricscoresareforyourreputation?Andforyourcareerprogress?

10. Unethicalbehaviours

Hypothesestotest:ECRsarewillingto‘bend’thesystemtoprogressandgetpublished.

Q1.Doyouhaveaclearunderstandingofwhatisgenerallyregardedasethicalandunethicaliresearchand/orpublishingpracticesorareyouuncertainaboutwhatismeantbytheseterms?Q2.Areyouawareofanyunethicalpublishing/citingbehaviouramongyourpeersoramongthosehigherinacademicstructure?Q3.Doyoubelievethereismorescrutinytodaywhichwillkeepthelidonanyproblembehaviours?

11. Impact

Hypothesestotest:Theyseeconnectingtoawideraudienceasbeinganimportantimpact.

Q1.Howimportantisittoyouthattheresearchyouareinvolvedinshouldhaveanimpactonyourpeers,onpolicyformers,onindustryor/andonthegeneralpublic?Whichgroupsaremostimportanttoyou?Q2. What is the best way toinfluence those groups you thinkyoushouldbereachingoutto?Q3. If you had thetime/opportunity to do more toincrease the impact of yourresearch,whatwouldyoudo?

12.Transformations

Hypothesestotest:Thesystemisunchangingandunbending,butthereislittleevidenceofthedesireforchangeamongECRs.

Q1.Doyouagreethatabigopportunityforthecurrentgenerationofresearchersistofundamentallychangethewaythatthescholarlycommunicationsystemworks?Q2.Ifso,doyouhaveanyoverallpictureofwhatformachangedsystemofscholarlycommunicationmighttake?Q3.Doyouthinkthatfiveyearsfromnowacademicswillstillbetypicallyrecruited,promotedandobtainfundingsolelyonthebasisoftheirpublicationrecordandcitationscoresbasedonaccumulatedreputation?

Q4.Doyouthinkjournalsandlibrarieswillstillhaveacentralrolefiveyearsdowntheline?

67

Appendix2:CodingtemplateforNationalReports

Interview 1 Interview 2 0. Bio & CV Gender (M, F)

Age: under 30 (Y); 30-34 (M); 35 and over (O) Doctoral student (D) or PostDoc (PG) University rank top (T), medium (M), low (L) Subject specialisation (in bullet list) Countries in which worked (in bullet list) 1. Background Research projects (current number) of which Inter- or multi-disciplinary (number) Role and status (number of projects as PI/leader) Research group (RG), centre (RC) or none (N) Past research groups worked with (number) of which the number as PI/leader Status of research groups largely international (I), national (N), university (U)

Those not in groups collaborating (C) or not Mentoring – existence & quality. None (N), poor (P), Good (G) Mentor/advisor (typically): PI, colleague (C), other (O), various (V)

ECR special provision. Yes (Y), no (N) 2. Career Want a career as university researcher Y, N, not sure (NS) Career progression: progression OK (P), with difficulty (D), still uncertain (U)

Consider working elsewhere (Y), (N), unsure (UN) Main motivation: promotion (P), curiosity or interest (I), other (O)

3. Scholarly Communication Behaviour Dissemination (chief characteristics, provide bullets – max. 5 strongest first)

Reading (as above) Citation (as above) Differ from mentors Y, N, no answer (O) List chief sources of information (databases, search engines, bibliographic services etc. – max 5 most important first).

Changes in scholarly behaviour N, Y (In case of latter what were they? List up to 5 bullets)

Main causes of change. List as bullets. Smartphones & Mobile use. Not used (NU), used (U). In used for what?

4. Social Media Finding scholarly info. Y, N. Citing social media. Y, N. Dissemination. Y, N. Connecting with other scholars (e.g. collaborating)

68

Encouragement. Y, N. If yes, who encouraged. List. Does use of social media and online networks result in: a) detachment from your institutions (1); b) getting closer to your peers elsewhere (2); c) both (3)

Do online scholarly networks lead to greater collaboration/connectivity. Y, N.

Do online scholarly networks help towards building reputation? Y, N.

Are new virtual groups different? Y, N. 5. Authorship Number of publications Author contribution/role. Position (first etc.); role (wrote articles, did corrections, literature review etc. – list up to three bullets)

Authorship Policy Y, N, don’t know (DK). List policies as short bullets

Would you do things differently Y, N. If so list as bullets the different things.

Influence on choice of journals. Y, N. If Y what was the influence? List as bullets, up to 3.

Policy towards OA. Y, N. don’t know (DK). If yes, list policies as up to 3 bullets.

OA advantages/disadvantages. List up to 3 bullets each OA publishing advances science and research (1) or will it dilute quality (2). 1, 2 or disagree with both (3). List any reasons as bullets (max. 3)

Like innovative features/Videos journals. Yes (Y), no (N), don’t know (DK).

Pressure to publish in top-ranked journals? Y, N. If yes, how does this affect scholarly communications and your career? (list up to 3 affects as bullets)

Conscious publication strategy. Y. N. What is it? List strategies as up to 3 bullets

Like to make public my research in less formal ways? Y, N.

Produced data/software. Y, N. if Y Data like credit for it? Y, N.

If Y, like it to be more visible. Y, N. 6. Peer Review Responding to comments. Y, N. If Y, how did you find the experience? Good (G), bad (B), mixed (M).

Experience as a reviewer. Y, N. Peer review fair? Y, N. if not why not – list up to 3 reasons as bullets

Peer Review clique? Y, N. Improve it? Y, N. If Y, give up to 3 reasons as bullets Should publishers do it? Y, N. If N, who else should do it? List up to 3.

7. Employment, reputation & career progression ECR treatment. Fair (F), unfair (U), mixed (M) Have ECR policy (Y, N); Clarity. Clear (C), not clear (NC). List as bullets up to 3 policies mentioned

69

Have ECR Evaluation (Y, N). List as bullets up to 3 methods mentioned

Suggested reforms (Y, N) If Y, give max. 3 Freedomtodevelopcareer.Y,N,partly(P) Slaves (Y, N). Any reforms mentioned, list 3 max. Does open science mean anything (Y, N). If Y what is significance for them. Give up to 3 examples.

Can OA fast track career. Y, N Do you publish in OA jnls (Y. N) or deposit in IRs (Y, N)? 8. Sharing & collaborating Ways of sharing ideas (Give up to three ways) Ways of sharing results, data, publication (Give up to three ways)

Collaborations & social media (give up to three examples) Use of social networks for building reputation (Y, N) Sharing behaviour different from mentors (Y, N) Has collaboration changed with experience (Y, N) Risk of losing competitive edge through collaboration (Y, N) 9. Metrics Use of metrics (Y, N). If Y, give up to 3 ways) Metrics and reputation (Y, N) Metrics and career progression; Important (I), not important (NI)

10. Unethical behaviours Ethical understanding. (Y, N). If Y and provide examples give up to 5

Are they personally aware (Y, N) If Y and provide examples give up to 5

Scrutiny. Y, N, not sure (NS) 11. Impact Importance for research to have impact (Y, N). List groups. Best ways to influence groups. List Strategy for impact. List ideas. 12. Transformations ECRs as « change players Y, N. Nature of change. Describe change briefly Will reputation system be the same in 5 years’ time? Y, N, not sure (NS)

Central role of journal in 5 years Y, N, not sure (NS) Central role of libraries in 5 years Y, N, not sure (NS)

70

Appendix3:DetailedsubjectrepresentationofECRs

Subject China France Malaysia Poland Spain UK US Total

Agricultureandforestry 3 2 5Anthropology 1 1Biology 4 1 3 1 2 11Botany 1 1Chemistry 2 1 3 1 7Computerscience 4 2 1 2 1 10Earth/environmentalscience

1 1 1 3

Economicsandbusiness 1 1 1 1 4Education 1 2 3Engineering(andTechnology)

2 2 2 5 11

Law 1 1Libraryandinformationsciences

2 2 4

Mathematics 2 2Medicineandhealth 4 2 2 7 15Microbiology 1 1Nutrition/foodsciences 1 1Physics 2 3 1 1 1 8Physicalchemistry 2 1 3Physiology 3 2 5Politics 1 1 2Psychology 1 1 4 6Socialsciences 2 2Sociology 1 1 1 3Statistics 1 1Technology,withsocialsciences

1 1

Travelandtourism 1 1Veterinaryscience 1 1 2Zoology 2 2Total 13 14 12 10 18 20 29 116