Authorship Readings

8/18/2019 Authorship Readings

1/21

Friedland, A. J. and C. L. Folt. 2000. Writing Successful Science Proposals. Yale University Press, New Haven.

Readings selected and compiled by Brad Taylor


2/21


3/21


4/21


5/21


6/21


7/21

MICHIGAN STATE UNIVERSITY GUIDELINES ON AUTHORSHIP

Adopted by the University Research Council

January 15, 1998

(To apply to all academic units which have not adopted their own written

Authorship - A person claiming authorship of a scholarly publication must have met the foll1.

a. Substantial participation in conception and design of the study, or in analysis and interpret

b. Substantial participation in the drafting of the manuscript or in the substantive editing of t

Final approval of the version of the manuscript to be published;c.

Ability to explain and defend the study in public or scholarly settings.d.

(Note: these criteria follow closely those recommended by several professional associations

International Committee of Medical Journal Editors, Annals of Internal Medicine 1988; 108:

e.

Acknowledgment - Contributions that do not justify authorship should be acknowledged se

manuscript. These may include general supervision of a research group, assistance in obtai

support.

2.

“Honorary Authorship” - A claim of authorship by, or assignment of authorship to, personsin some way with a study but do not meet the four criteria in item 1 may constitute an une

3.

Graduate Student Authorship - “Faculty should be especially aware of their responsibility t4.

orship Guidelines


8/21


9/21

PLoS Biology | www.plosbiology.org 0013

Correspondence

January 2007 | Volume 5 | Issue 1 | e18

Author Sequence and Credit for

Contributions in Multiauthored

PublicationsTeja Tscharntke*, Michael E. Hochberg, Tatyana A.Rand, Vincent H. Resh, Jochen Krauss

The increasing tendency across scientific disciplines to write multiauthored papers [1,2] makes the issue of thesequence of contributors’ names a major topic both interms of reflecting actual contributions and in a posterioriassessments by evaluation committees. Traditionally, thefirst author contributes most and also receives most of thecredit, whereas the position of subsequent authors is usuallydecided by contribution, alphabetical order, or reverseseniority. Ranking the first or second author in a two-authorpaper is straightforward, but the meaning of positionbecomes increasingly arbitrary as the number of authorsincreases beyond two. Criteria for authorship have beendiscussed at length, because of the inflationary increase in

the number of authors on papers submitted to biomedical journals and the practice of “gift” authorship [3,4], but asimple way to determine credit associated with the sequenceof authors’ names is still missing [4–7] (http://www.councilscienceeditors.org).

The situation in our area of research—the ecologicaland environmental sciences—has changed in recent years.Following informal practices in the biomedical sciences,the last author often gets as much credit as the first author,because he or she is assumed to be the driving force, bothintellectually and financially, behind the research. Evaluationcommittees and funding bodies often take last authorship as asign of successful group leadership and make this a criterionin hiring, granting, and promotion. This practice is unofficial,and hence not always followed, meaning that sometimeslast authors “mistakenly” benefit when they actually arenot principal investigators. Moreover, there is no accepted

yardstick in assessing the actual contribution of a groupleader to given scientific publications [8,9], so interpretationof author sequence can be like a lottery. Hence, one reallydoes not know if being last author means that the overallcontribution was the most or least important.

Although reducing evaluation of authors’ complexcontributions to simple metrics is regrettable, in reality it isalready in practice in most evaluation committees. Hence, inour opinion, we need a simple and straightforward approachto estimate the credit associated with the sequence of

authors’ names that is free from any arbitrary rank valuation.In multiauthored papers, the first author position shouldclearly be assigned to the individual making the greatestcontribution [4–6], as is common practice. However, authorsoften adopt different methods of crediting contributionsfor the following authors, because of very differenttraditions across countries and research fields, resulting in

very different criteria that committees adopt to quantifyauthor’s contributions [8,9]. For example, some authors usealphabetical sequence, while others think that the last authorposition has great importance or that the second authorposition is the second most important. Still others detail eachauthor’s contribution in a footnote.

We suggest that the approach taken should be stated inthe acknowledgements section, and evaluation committeesare asked to weigh the contribution of each author based onthe criteria given by the authors. This would make reviewersaware that there are different cultures to authorship order.

The usual and informal practice of giving the whole credit(impact factor) to each author of a multiauthored paper isnot adequate and overemphasises the minor contributionsof many authors (Table 1). Similarly, evaluation of authorsaccording to citation frequencies means often overratingresulting from high-impact but multiauthored publications.The following approaches may be identified.

(1) The “sequence-determines-credit” approach (SDC).The sequence of authors should reflect the decliningimportance of their contribution, as suggested by previousauthors [4–6]. Authorship order only reflects relativecontribution, whereas evaluation committees often needquantitative measures. We suggest that the first author shouldget credit for the whole impact (impact factor), the secondauthor half, the third a third, and so forth, up to rank ten.

When papers have more than ten authors, the contribution ofeach author from the tenth position onwards is then valuated

just 5%.(2) The “equal contribution” norm (EC). Authors use

alphabetical sequence to acknowledge similar contributionsor to avoid disharmony in collaborating groups. We suggestthat the contribution of each author is valuated as an equalproportion (impact divided by the number of all authors, buta minimum of 5%).

(3) The “first-last-author-emphasis” norm (FLAE). Inmany labs, the great importance of last authorship is wellestablished. We suggest that the first author should get credit

of the whole impact, the last author half, and the credit ofthe other authors is the impact divided by the number of allauthors [as in (2)].

(4) The “percent-contribution-indicated” approach(PCI). There is a trend to detail each author’s contribution(following requests of several journals) [7]. This should alsobe used to establish the quantified credit.

The SDC approach (as a new suggestion), the EC norm(alphabetical order), the FLAE norm, and the PCI approachmay be combined (e.g., FLAE and SDC), but need to beexplicitly mentioned in the acknowledgements.

Our suggestion of explicit indication of the methodapplied, including the simple method of weighing authors’

Table 1. Comparison of the Credit for Contributions to ThisPaper under the Four Different Models Suggested in the Text

Author SDC EC FLAE PCI Contribution

(%) for PCI

Traditional

Credit

TT 14.7 2.9 14.7 8.8 60 14.7

MEH 7.3 2.9 2.9 2.9 20 14.7

TAR 4.9 2.9 2.9 1.5 10 14.7VHR 3.7 2.9 2.9 0.7 5 14.7

JK 2.9 2.9 7.4 0.7 5 14.7

Sum 33.5 14.5 30.8 14.6 100 73.5

The credit is based on the impact factor, which is 14.7 (2005) forPLoS Biology . Thetraditional but informal practice of giving the whole credit to all authors may be the mostattractive, but often least justified approach.doi:10.1371/journal.pbio.0050018.t001


10/21

PLoS Biology | www.plosbiology.org 0014

rank in publications in a quantitative way, will avoidmisinterpretations and arbitrary a posteriori designationsof author contributions. Multidisciplinary scientificcollaboration indeed must be encouraged, but we need toavoid misinterpretations so that current and future scientificcommunities can evaluate author contributions.

Acknowledgments

We applied the SDC approach for the sequence of authors. We

are grateful for the stimulating discussions and comments by JanBengtsson, Charles Godfray, Bradford A. Hawkins, Christian Körner,

William F. Laurance, Bernhard Schmid, Wim van der Putten, andLouise Vet.

Funding. The authors received no specific funding for this article.Competing interests. The authors have declared that no competing

interests exist.

References1. Regaldo A (1995) Multiauthor papers on the rise. Science 268: 25.2. Johnson S (2006) Are ecologists becoming more gregarious? Bull British

Ecol Soc 37: 23–24.3. Leash E (1997) Is it time for a new approach to authorship? J Dental Res 76:

724–727.4. Hunt R (1991) Trying an authorship index. Nature 352: 187.5. Schmidt RH (1987) A worksheet for authorship of scientific articles. Bull

Ecol Soc America 68: 8–10.6. Verhagens JV, Wallace KJ, Collins SC, Thomas TR (2003) QUAD system

offers fair shares to all authors. Nature 426: 602.7. Anderson C (1992) Writer’s cramp. Nature 355: 101.8. Laurance WF (2006) Second thoughts on who goes where in author lists.

Nature 442: 26.9. Weltzien JF, Belote RT, Williams LT, Keller, JF, Engel EC (2006)

Authorship in ecology: Attribution, accountability, and responsibility. FrontEcol Environm 4: 435–441.

Citation: Tscharntke T, Hochberg ME, Rand TA, Resh VH, Krauss J (2007) Authorsequence and credit for contributions in multiauthored publications. PLoS Biol 5(1):

e18. doi:10.1371/journal.pbio.0050018

Copyright: © 2007 Tscharntke et al. This is an open-access article distributed underthe terms of the Creative Commons Attribution License, which permits unrestricteduse, distribution, and reproduction in any medium, provided the original authorand source are credited.

Teja Tscharntke is Professor and Tatyana A. Rand is Postdoc with the AgroecologyGroup, University of Göttingen, Göttingen, Germany. Michael E. Hochberg isResearch Director at Centre National de la Recherche Scientifique, University ofMontpellier II, Montpellier, France. Vincent H. Resh is Professor at the Departmentof Environmental Science, Policy and Management, University of California,Berkeley, California, United States of America. Jochen Krauss is Postdoc with theInstitute of Environmental Sciences, University of Zurich, Zurich, Switzerland, andthe Department of Animal Ecology, Population Ecology, Bayreuth, Germany.

* To whom correspondence should be addressed. E-mail: [email protected]

January 2007 | Volume 5 | Issue 1 | e18


11/21

43

© The Ecological Society of America www.frontiersinecology.org

Articles published in peer-reviewed journals are themedium by which scientists present their findings tothe scholarly community. The quality and quantity of publi-cations are essential components for building careers, fund-ing projects, and generating a sense of accomplishment andself-worth (Lindsey 1980). The past five decades have seen aproliferation of scientific subdisciplines, an increase in thenumber of researchers and collaborative manuscripts, and acorresponding increase in multi-authored articles (Regalado1995; Cronin 2001). Multiple authorship is an increasingtrend that has now become the norm, but there remains a

paucity of useful and definitive guidelines to aid researchersin addressing authorship issues (Rennie et al. 1997; Kleinand Moser-Veillon 1999). Although several journals (eg TheLancet, Journal of the American Medical Association, andProceedings of the National Academy of Sciences of the United

States of America) have adopted clearly defined guidelinesthat specify criteria for authorship and communicate thatinformation to the readers, most journals have only vague ornon-existent guidelines (Rennie et al. 2000).

In the absence of standardized definitions or guidelines onauthorship (eg criteria for author inclusion or order), scien-tists employ a variety of personal criteria that are unknownto readers and that probably differ from criteria employed byother authors, even for articles in the same journal. Forexample, individual authors, laboratory groups, or even sub-disciplines may determine byline composition and order

based on arbitrary or idiosyncratic traditions, customs, orhabits. As such, the order in which authors are listed com-municates little information about the importance of thecontribution of each individual, since a wide variety of undisclosed methods are used to assign order (Rennie et al.2000). This can create an environment in which credit,accountability, and responsibility for research are neitherpersonally accepted nor publicly acknowledged (Zuckerman1968). A lack of communication about authorship mayengender interpersonal issues and ethical dilemmas if unde-serving individuals are included as authors, or if contributingresearchers are not included (Rennie and Flanagin 1994;Rennie et al. 1997). The purpose of this article is to discuss

potential approaches to deciding who should be included inthe authorship byline, and in what order. We recommendimproved communication among authors during the writingprocess, and outline an approach used by other science disci-plines, wherein authors publish their contributions to amanuscript in a separate byline (eg Panel 1).

Authorship trends in ecology

Ecologists are in a particularly challenging situationwhen dealing with authorship, since our discipline has

CONCEPTS ANDQUESTIONS

Authorship in ecology: attribution,accountability, and responsibility Jake F Weltzin1*, R Travis Belote2, Leigh T Williams1, Jason K Keller3, and E Cayenne Engel1

Quality and quantity of publications are among the most important measures determining the success of

ecologists. The past 50 years have seen a steady rise in the number of researchers and collaborative manu-

scripts, and a corresponding increase in multi-authored articles. Despite these increases, there remains a

shortage of useful and definitive guidelines to aid ecologists in addressing authorship issues, leading to a lack

of consistency in what the term “author” really means. Deciding where to draw the line between those who

have earned authorship and those who are more appropriately credited in the acknowledgments may be one

of the more challenging aspects of authorship. Here, we borrow ideas from other scientific disciplines and

propose a simple solution to help ecologists who are making such decisions. We recommend improving com-

munication between co-authors throughout the research process, and propose that authors publish their con-

tributions to a manuscript in a separate byline.

Front Ecol Environ 2006; 4(8): 435–441

1Department of Ecology and Evolutionary Biology, University of

Tennessee, Knoxville, TN 37996 *([email protected]); 2Department

of Biological Sciences, Virginia Polytechnic Institute and State

University, Blacksburg, VA 20461; 3Smithsonian Environmental

Research Center, Edgewater, MD 21037

In a nutshell:• Ecology is becoming an increasingly collaborative science, with

researchers from various disciplines involved in ecologicalresearch projects; decisions about authorship of a manuscriptare therefore becoming more difficult

• The Ecological Society of America’s Code of Ethics providesonly vague guidelines to determine who should be grantedauthorship; it is therefore time for ecologists to develop a moresubstantial framework for attributing credit to authors

• Here, we propose a byline statement summarizing the contribu-tion of each author to the research, to be published with thearticle (a practice now commonly used in biomedical journals)


12/21

Authorship in ecology JF Weltzin et

436

www.frontiersinecology.org © The Ecological Society of Amer

developed into a collaborative science which frequentlyproduces multiple-author articles without simultaneouslydeveloping useful guidelines for handling this issue.Interdisciplinary teams of scientists are often required toinvestigate questions in ecology, and these teams typi-cally include several layers of participants, such as princi-pal investigators, graduate and undergraduate studentresearchers, technicians, statisticians, and field assistants.Increases in the diversity of funding sources, larger labo-

ratories and centers of research, and advances in technol-ogy all promote the growth of research teams and net-works. This trend is mirrored by patterns of authorshipfor articles published in the journal Ecology; between1925 and 2005, the mean (± 1 SE) number of authorscredited per article tripled from 1.1 ± 0.06 to 3.3 ± 1.1,and the maximum number of authors on a single paperincreased from 2 to 17 (Figure 1).

Ecological research continues to be increasingly collab-orative and interdisciplinary, a pattern that is encouragedby the National Institute for Health and the NationalScience Foundation and facilitated by institutions such asthe National Center for Ecological Analysis and

Synthesis. Despite this, and the fact that ecologicalresearch continues to move in the direction of large-scale,long-term projects, ecologists and the majority of theirjournals have yet to formally address the authorship issue.An ISI Web of Knowledge search on the keyword“authorship” yields only two matches in the top 20 ecol-ogy journals (ranked by impact factor). In contrast, thesame search yields 34 results for the Journal of the AmericanMedical Association, 13 results for Nature, and 14 forScience. There were 1090 results when all journals wereincluded in the search (Thomson ISI 2005). While weacknowledge that not all search results are relevant to thetopic at hand, this pattern suggests that ecology may be

lagging behind other scientific disciplines when address-ing the complexities associated with authorship. Thetrend towards increasing author numbers in the absenceof guidelines for acknowledging involvement in projectswill continue to complicate this already sensitive issue.

The meaning of “author”

The difficulties associated with selecting both who willbecome an author on the final manuscript and in whatorder those authors should appear are neither trivial nor

easily resolved (Panel 2). In the absence of clear guidlines, scientists use individualized criteria, creating a laof consistency in what is really meant by the ter“author” (Rennie et al. 2000). In ecology, it is generaassumed that the person placed first in the list of authocontributed the most time and energy to the project, bhow does one compare their relative contributions to t

second, third, or eighth person named? Are all authoequally responsible for the work presented, and can eabe held accountable for the claims made in the articAre certain authors earning undeserved credit for prjects, while others are unfairly denied credit for perhagreater contributions? There is currently too much dispity between the criteria employed by each set of authowhen submitting a manuscript, allowing researchers lose track of who is truly responsible and accountable, atherefore deserving of credit for the research. This calso lead to situations where potentially unaccountabauthors (ie those not obliged to accept responsibility fcontent) are given credit for the article. This dilutes t

impact of having one’s name listed on a manuscript, amay detract from the professional value of the publisharticle for the secondary authors who appear as “et alrather than having their full name listed in all citations

The contribution of each author is diminished wh“ghost” authors, “guest” contributors, and those wacquired the initial funding for the project are includin the list of authors (Rennie and Flanagin 1994). Ghoauthors, ie those who receive author credit for simpediting completed research, are more often found in tbiomedical sciences, but probably occur in all disciplinGuest contributors are invited to participate in manscript preparation in an undemanding capacity, mainly

justify the addition of their name to the authorship lfor the positive impact gained from their celebrity ingiven field. Finally, there are all-too-common occurences of honorary authors, who obtain authorship providing funding or lab space, or even by “tradinauthorship rights on one paper for inclusion on anoth(Flanagin et al. 1998). Inclusion of such guest contribtors and ghost authors generates ethical questionbecause researchers and co-authors differ in opinion ovthe appropriateness of including “authors” who cotributed neither intellectually nor physically to the prduction of the article (Culliton 1988).

For ecologists, the issue is further complicated wh

assessing the contribution of participants, such as techncians and student researchers, who may have been vito one portion of a project but not another (Panel 2Participants without a PhD may feel as though they hacontributed substantially to the completion of an expement, but they may have no standards on which to statheir claim for inclusion on the authorship list (Heffn1979). Alternatively, investigators may be undecidabout including a technician as an author when that pson worked for only a few years on a longer-term projeand meets only some of their criteria for authorship.

Panel 1. Author contributions for this article

JFW co-conceived and co-developed the idea for the manu-script, co-refined the intellectual content and scope, edited alldrafts, prepared the final version of the manuscript, and facili-tated the gathering of contributors. RTB co-conceived and co-developed the idea, edited all drafts,and assessed historic trendsin authorship in Ecology . LTW initiated the project, co-developedand co-refined the intellectual content, and wrote the first two

drafts. JKK co-developed the idea, edited all drafts, and con-ducted the keyword search. ECE co-developed the idea andcoordinated the authorship survey. JFW is the guarantor for theintegrity of the article as a whole.


13/21

JF Weltzin et al. Authorship in ecology

43


sum, the lack of guidelines leads to anenvironment in which individualsinvolved in a project are often unsureabout their own right to claim or dis-pute authorship and provides no meansto resolve situations that arise over theselection of authors and their order.

Developing authorship

guidelines for ecologists

Scientific journals, professional soci-eties, and individual scientists have pre-viously attempted to create definitionsof authorship and to provide guidelineson how to determine which partici-pants should be credited on the manu-script. A variety of approaches existamong the various scientific disciplines,including listing authors based on

seniority, extent of contribution, impor-tance of contribution, or simply byalphabetical order or the outcome of acoin toss (Rennie et al. 1997). However,these approaches are infrequently com-municated to readers, who must maketheir own assumptions about howauthors were selected and the order inwhich they are listed. Furthermore,these approaches are often ignored bythe authors who submit manuscripts, sothat even if a journal attempts to pro-vide a standardized definition of author-

ship, the scientists may fail to adopt it(Klein and Moser-Veillon 1999).

The Ecological Society of America(ESA) currently suggests that its mem-bers employ a rather vague set of guide-lines presented in the publication sec-tion of its Code of Ethics (ESA 2006). Specifically, theESA guidelines related to the selection of authors state:

1. Researchers will claim authorship of a paper onlyif they have made a substantial contribution.Authorship may legitimately be claimed if researchers

(a) conceived the ideas or experimental design;(b) participated actively in execution of the

study;(c) analyzed and interpreted the data; or(d) wrote the manuscript.

2. Researchers will not add or delete authors from amanuscript submitted for publication withoutconsent of those authors.

3. Researchers will not include as co-author(s) anyindividual who has not agreed to the content of the final version of the manuscript.

Although these guidelines describe who should beincluded as an author, they do not address the question of author order. They also leave interpretation of “substan-tial contribution” to the individual(s) making the deci-sions, leading to confusion and inequities. The ESAguidelines are more lenient than those employed by otherjournals, in that authorship may be granted even if only

one of the four criteria is met. By comparison, theUniform Requirements for Manuscripts Submitted toBiomedical Journals, developed by the InternationalCommittee of Medical Journal Editors, states that con-tributors only qualify as authors if they meet all of thesecriteria (ICMJE 2005).

Seeking a more definitive approach, Galindo-Leal(1996) suggested a two-stage process, using a modifiedscoring system originally proposed by Hunt (1991). Thefirst stage involves improving communication betweenco-authors by drafting a pre-research agreement for all

Figure 1. (a) Mean and (b) maximum number of authors per article published inEcology during 1925, 1955, 1985, and 2005. Notes and comments were excluded fromthe analysis. Mean (± 1 SE) number of authors with the same capital letter did not differ(P > 0.05, Tukey’s HSD; issues as replicates, thus n = 4, 4, 6, and 12; data werenormal).

(a)

(b)

Years


14/21


438


parties to use as a guideline or protocol. Co-authors monitortheir involvement and progress throughout the duration of

the research project by using the scoring system, whichassesses their participation in planning, executing, analyz-ing, interpreting, and writing the manuscript. The secondstage involves reviewing the scoring system and using thescores to determine who has earned credit as an author; theauthorship order is selected by arranging names in thedescending order of their scores. Although this scoring sys-tem may work well for research teams involving few partic-ipants, it is less useful for the multi-year, large-scale complexcollaborative projects that are becoming the norm in ecol-ogy. Complications may arise when participants are highly

involved for only a portion of the prject or when participants are not askto be involved in all aspects of the prject. Weighting the various categorisuch as “planning” or “analysis”, is difcult because it is often a matter of opiion as to how much credit is earned

conceiving the project or analyzing tdata relative to credit earned by phycally collecting the data. Informal intlectual contributions from technicstaff may go unnoticed or be undappreciated by researchers preparingmanuscript. The practice of discussiauthorship before, during, and afterproject is surely one that should adopted by all scientists to avoid confsion and discord over issues of authoship. However, in our opinion, usithe scoring system may allow too na

row a scope for contemporary projein ecology.

Proposals to resolve these challengand establish a realistic and functionset of guidelines for authors shouinclude a way to recognize both credand accountability for the article, whmaintaining flexibility for a diverse sof research participants, projects, asituations. These guidelines must available to the participants for agiven project, as well as to the readeof each manuscript, to ensure that t

meaning of the authorship list is communicated to the scientific communiStandards for determining authorshorder, and for differentiating betweauthors and those whose names moappropriately appear in the acknowedgments, must also be established.

Establishing accountability and

responsibility

Rennie et al. (1997) proposed a system that stresses timportance of accepting responsibility and accountabil

for research in order to earn credit for it (see also Davis anGregerman 1969; Garfield 1983; Moulopoulos et al. 198Huth 1986; Saffran 1989; Mancini 1990; Hunt 1991; anGreen 1994). They propose a system of “contributorshi(as opposed to “authorship”) that recognizes the contribtion of each individual to the manuscript, and establishthe accountability of that person to the content of tmanuscript; in short, the “word and concept contributor”substituted for the “word and concept author” (Rennieal. 1997). Contributors disclose which particular aspectsa manuscript they were responsible for in a byline that

Panel 2. Authorship survey

We invited attendees of the 2004 Annual Meeting of the Ecological Society of America(Portland, OR) to participate in an informal, voluntary survey about authorship.Participants were presented with six hypothetical scenarios involving research collabo-rators, and were asked which characters should be authors, and in what order thoseauthors should appear on manuscripts.Hardcopies of the surveys were posted on a bul-letin board in the main foyer of the conference hall at the meeting;respondents returnedcompleted surveys onsite, or mailed them at a later date. We received 57 completed

surveys. In Part A below, we present three scenarios from the survey. In Part B, weinclude follow-up questions to spur discussion and present highlights of the results fromsurvey respondents.

Part A Part B

Scenario 1:Professor X and new graduate student Y are developing a research project for Y. Y is interested in a project that Z , a gradu-ate student colleague/professor in thedepartment, is conducting. Y discussesproject concepts with X , and decides toconduct a project descended from andclosely related to Z ’s project.The ques-tions, methods, and analysis were devel-oped solely by X and Y , and all physicalwork was conducted by Y. Y and Z met afew times to discuss methods for analy-sis, but Z contributed nothing to manu-script preparation.

Scenario 2:Principal Investigator X developed theintellectual ideas, wrote a proposal, andreceived monies for a new, well-fundedproject. X hires technician T to handleproject logistics, and to ensure that theproject follows X ’s original vision; T col-lects much of the empirical data, andsupervises undergraduate students whoassist during data collection. Researchassistant A is responsible for manipula-tion, analysis, and interpretation of data

collected by T et al .

Scenario 3:Professor X initiates writing of a synthe-sis paper with graduate student Y ontheir favorite topic. After the two meetseveral times to outline a paper, Y takesthe task of writing the first draft. X and Y pass the manuscript back and forth sev-eral times before X does the final revi-sion and submits the manuscript for pub-lication.

Follow-up questions:• Should Z be included as an author?• Who should be first author?

Survey responses:• 25% of respondents thought that Z

deserved authorship.• 84% of respondents indicated that Y

should be first author, whereas 16% of respondents indicated that X deservedto be the first author.

Follow-up question:• Who should be included as an author,

and in what order?

Survey responses:• 78% of respondents thought all three

characters should be included asauthors.

• 78% chose X as first author.• 14% chose A as first author.• 82% included T as an author.• Respondents listed 10 unique combina-

tions for authorship order.

Follow-up question:• Who should be the first author?

Survey responses:• 46% of respondents thought that X

should be the first author.• 46% thought that Y should be the first

author.• 8% could not decide.


15/21


43


published with the article (see, for example, the contribu-torship byline [Panel 1] for this article). This systemrequires each author to publicly accept accountability fortheir particular contribution; moreover, it would enablereaders to more objectively ascribe credit to the namedindividuals, as well as determining the credibility of thearticle as a whole. A “contributorship” policy was recently

adopted by Proceedings of the National Academy of Sciencesof the United States of America (PNAS 2006), which poststhe byline as a footnote to the paper, albeit only online(Panel 3). Similarly, authors submitting a manuscript to Nature are “strongly encouraged to include a statement inthe end notes to specify the actual contribution of each co-author” (Nature 2006; see also Anonymous 1999).

Alternatively, journals could establish standards andconsistency for bylines by providing a list of possible tasksor responsibilities to contributors (Rennie et al. 1997;Klein and Moser-Veillon 1999; see also Panel 3). Thatsaid, it is probably unnecessary and overly complicated toattempt to standardize all job descriptions for all research

projects. It is probably more important to create anopportunity for authors to declare individual contribu-tions, whatever they may be, and to publish them withthe manuscript. Only by disclosing this information canthe contributors guarantee that their relative responsibil-ities, and thus their relative accountability and credit, arepublicly accepted and acknowledged.

Rennie et al. (1997) advise researchers to “meet, dis-cuss, and decide on their respective contributions to theproject, as well as the relative value of the contributionsto the whole, and in what order to list them in publica-tions”. As Galindo-Leal (1996) stressed, communicatingwith collaborators before, during, and after the project is

an important part of ensuring that responsibility isaccepted and acknowledged, credit is assigned fairly, andconflicts are avoided (Figure 2).

By committing to ongoing discourse about authorshipthroughout a particular project, contributors can makeinformed decisions as to individual contributions, whichmay facilitate ordering of authors. Authorship order is gen-erally understood to be designated by placing the name of the persons involved in order of the importance of theirduties, “in descending order, starting with the collaboratorwho made the most substantial contributions” (Rennie etal. 1997). Since each research team may employ uniquecriteria, such as allowing someone to take the last position

on the authorship list for providing funding, it is particu-larly important that the ordering methodology is disclosedto the readers (Klein and Moser-Veillon 1999). In sum,open communication between all participants involved ina research project can yield the best results for understand-ing and determining authorship order.

Guarantors and acknowledgees

Although published papers are typically (and appropri-ately) considered as a whole, complex projects involving

multiple collaborators, each involved in one aspect of theproject, may obscure internal assignment or external per-ception of accountability and responsibility. Thus, it isgood practice for each research team also to designate atleast one contributor as a guarantor for the whole project(eg Panel 1). Guarantors are individuals who have con-tributed substantially to the manuscript and who have

also made an extra effort to ensure the integrity of thepaper as a whole. Guarantors may organize the varioustasks associated with manuscript preparation, ensure theinternal consistency of the final manuscript, and solicitand organize contributorship statements; as such, they areprepared to be accountable for all parts of the completedmanuscript, before and after publication (Rennie et al.1997). Recognizing a guarantor ensures that someone onthe research team accepts and publicly acknowledgesresponsibility and accountability for the entire project,including each component of the manuscript. Guarantorsserve the scientific community by certifying that all workwas done properly and thoroughly, and by guarding

against dishonest scientific practices. Acknowledging aguarantor improves trust and credibility in science andpromotes good research practices.

Deciding where to draw the line between those whohave earned authorship and those who are more appropri-ately credited in the acknowledgments may be one of themore challenging aspects of authorship. One meaningfulway of thinking about the differences between these twosets of participants may be to consider whether or not theparticipant is responsible and accountable for the article.A contributor receiving credit for the article should be

Panel 3. Guidelines for authorship, Proceedings of the National Academy of Sciences of the United States of America (PNAS 2006)

Authorship should be limited to those who have contributedsubstantially to the work.The corresponding author must haveobtained permission from all authors for the submission of eachversion of the paper and for any change in authorship.

All collaborators share some degree of responsibility for any

paper they co-author. Some co-authors have responsibility forthe entire paper as an accurate, verifiable report of the research.These include co-authors who are accountable for the integrityof the data reported in the paper, carry out the analysis, writethe manuscript, present major findings at conferences, or pro-vide scientific leadership to junior colleagues. Co-authors whomake specific, limited contributions to a paper are responsiblefor their contributions, but may have only limited responsibilityfor other results.While not all co-authors may be familiar withall aspects of the research presented in their paper, all collabora-tors should have in place an appropriate process for reviewingthe accuracy of the reported results.

Authors must indicate their specific contributions to the pub-lished work.This information will be posted online as a footnoteto the paper. Examples of designations include:

• Designed research• Performed research

• Contributed new reagents or analytic tools• Analyzed data• Wrote the paper

An author may list more than one contribution, and more thanone author may have contributed to the same aspect of the work.


16/21


440


willing to be held accountable for its contents and not bejust responsible for a portion of the work involved. Incontrast, an acknowledgee may contribute formal orinformal ideas to ongoing projects, collect enormousamounts of data, and develop and/or conduct statisticalanalyses, but may not be accountable for the final con-tents of all or even portions of the final manuscript. Opencommunication about the roles, responsibilities, and

expectations for authors as opposed to acknowledgeesshould be ongoing during the writing process.

Will a system of contributorship work for ecology?

Critics of similar proposals for contributorship advance sev-eral reasons why these systems may not work (Rennie et al.1997; Flanagin et al. 1998; Yank and Rennie 1999; Rennieet al. 2000). Skeptics argue that the system of naming con-tributors and disclosing individual responsibilities is no dif-ferent than current author and acknowledgment lists. Thissystem is different, however, because it eliminates the “arti-ficial distinction, mostly of a social nature, between authors

and non-author contributors – that is, between authors andacknowledgees. The contributions of all (not just those of acknowledgees) are described and disclosed” (Rennie et al.1997). Critics also worry that any systematic change wouldbe resisted by researchers, but this could be overcomethrough the leadership of journals, professional societies,and indexers by requiring that article submissions use thesystem. While no system will put an end to disagreementsover authorship rights, forcing participants to think criti-cally and publish the contribution of each individual mayattenuate problems and abuses of authorship.

Acceptance of a contributorship systewill require behavioral changes on tpart of researchers and technical changby journals and professional societiThose who argue that a system such this would already be in place if it wergood idea may be comfortable with t

status quo. In fact, as described above, tsystem has been used by numerous bmedical journals for some time, andbeing used or considered by top qualjournals that publish ecology papers, suas PNAS and Nature. It will take effortbring about this change, but we argue thsuch a modification is necessary in a pulishing environment where more amore researchers are likely to experienissues related to authorship.

Conclusions

Although no system will completely resolthe challenges associated with authorshsubstituting “contributors” for “authors” aasking that all researchers disclose their r

sons for including authors and their relative order may golong way towards ensuring proper credit and appropriresponsibility for articles. Including this information asbyline, in addition to a statement of acknowledgments, wenable readers (as well as contributors) to better understawhere responsibility, accountability, and credit belong. the number and frequency of multi-author papers continuto rise, ignoring authorship issues may dilute the meaning

“author”. Our ecological journals and professional societshould adopt this system, or its equivalent, as a reasonabresponse that would provide much needed guidance for contemporary researchers and scholars. It is time for ecogists to join the rest of the scientific community in discussiauthorship issues and developing guidelines for our article

Acknowledgments

P Allen contributed to initial discussions of this topic aco-refined the intellectual content of earlier versions the manuscript. C DeVan assisted with data collectiand organization for Figure 1. The survey on authorsh

was developed and implemented with the help of Fitzpatrick, C Iversen, J Nagel, and L Souza. Commenfrom P Cole, S Collins, O Dermody, M Fitzpatrick, Iversen, C Reilly, N Sanders, and L Souza improved elier versions of the manuscript.

ReferencesAnonymous. 1999. Policy on papers’ contributors. Nature 399: 3

Cronin B. 2001. Hyperauthorship: a postmodern perversion or e

dence of a structural shift in scholarly communication pr

tices? J Am Soc Inf Sci Tech 52: 558–69.

Figure 2. Intellectual contribution in ecology can be difficult to quantify because fieldtechnicians or undergraduate students may provide important, informal observationsthat can easily be under-acknowledged by principal investigators. The informalcontributions may drive future research, direct data analyses, and contribute to

manuscripts. Communication among potential contributors before, during, and after a project is critical to ensure assignment and acceptance of responsibility. Each contributoris responsible for drafting his or her own byline; the guarantor is responsible forevaluating each byline relative to the others, and for maintaining internal consistency.


17/21


Culliton BJ. 1988. Authorship, data ownership examined. Science242: 658.

Davis PJ and Gregerman RI. 1969. Parse analysis: a new method forthe evaluation of investigators’ bibliographies. New Engl J Med281: 989–90.

ESA 2006. Ecological Society of America code of ethics. AdoptedAugust 2000. www.esapubs.org/esapubs/ethics.htm. Viewed 29 January 2006.

Flanagin A, Carey LA, Fontanarosa PB, et al. 1998. Prevalence of articles with honorary authors and ghost authors in peer-reviewed medical journals. JAMA 280: 222–24.

Galindo-Leal C. 1996. Explicit authorship. B Ecol Soc Am 77:219–20.

Garfield E. 1983. Essays of an information scientist, Vol 5:1981–1982. Philadelphia, PA: ISI Press.

Green MS. 1994. Authorship! Authorship! JAMA 271: 1904.Heffner AG. 1979. Authorship recognition of subordinates in col-

laborative research. Soc Stud Sci 9: 377–84.Huth EJ. 1986. Irresponsible authorship and wasteful publication.

Ann Internal Med 104: 257–59.Hunt R. 1991. Trying an authorship index. Nature 352: 187.ICMJE (International Committee of Medical Journal Editors).

2005. Uniform requirements for manuscripts submitted to bio-medical journals: writing and editing for biomedical publica-

tion. www.icmje.org/index.html. Viewed 7 December 2005.Klein CJ and Moser-Veillon PB. 1999. Authorship: can you claim abyline? J Am Diet Assoc 99: 77–79.

Lindsey D. 1980. Production and citation measures in the sociologyof science: the problem of multiple authorship. Soc Stud Sci 10:145–62.

Mancini GBJ. 1990. Documenting contributions to authorship. Ann Internal Med 104: 257–59.

Moulopoulos SD, Sideris DA, and Georgilis KA. 1983. Individualcontributions to multiauthor papers. Brit Med J 287 :1608–10.

Nature. 2006. Publication policies. www.nature.com/nature/authors/policy/index.html. Viewed 29 January 2006.

PNAS (Proceedings of the National Academy of Sciences of theUnited States of America). 2006. Information for authors.www.pnas.org/misc/iforc.shtml. Viewed 29 January 2006.

Regalado A. 1995. Multiauthor papers on the rise. Science 268: 25.Rennie D and Flanagin A. 1994. Authorship! Authorship!

Guests, ghosts, grafters, and the two-sided coin. JAMA 271:469–71.

Rennie D, Yank V, and Emanuel L. 1997. When authorship fails: aproposal to make contributors accountable. JAMA 278:579–85.

Rennie D, Flanagin A, and Yank V. 2000. The contributions of authors. JAMA 284: 89–91.

Saffran M. 1989. On multiple authorship: describe the contribu-tion. The Scientist 3: 9.

Thomson ISI. 2005. Web of knowledge: science citation indexexpanded. www.isiwebofknowledge.com. Data retrieved 7December 2005.

Yank V and Rennie D. 1999. Disclosure of researcher contribu-tions: a study of original research articles in The Lancet. AnnInternal Med 130: 661–70.

Zuckerman HA. 1968. Patterns of name ordering among authors of scientific papers: a study of social symbolism and its ambiguity. Am J Sociol 74: 276–91.

44



18/21

EDITORIAL

www.sciencemag.org SCIENCE VOL 301 8 AUGUST 2003

The average number of authors on scientific papers is skyrocketing. That’s partly becauselabs are bigger, problems are more complicated, and more different subspecialties areneeded. But it’s also because U.S. government agencies like the National Institutes of Health (NIH) have started to promote “team science.” As physics developed in the

post–World War II era, federal funds built expensive national facilities, and these served as surfaces on which collaborations could crystallize naturally. That has produced some

splendid results. Multidisciplinary teams have been slower to develop in biology, but now the rush ison. NIH recently sponsored a meeting entitled “Catalyzing Team Science”—something new for anagency traditionally wedded to the investigator-initiated small-project kind of science. Increasinglycomplex problems, NIH seems to be saying, will require larger and more diverselyspecialized groups of investigators. So team science is part of its road map: a “Good Thing.” That may be right.

Multiple authorship though—however good it may be in other ways—presents problems for journals and for the institutions in which these authors work. For the journals, long lists of authors are hard to deal with in themselves. But those longlists give rise to more serious questions when something goes wrong with the pa-

per. If there is research misconduct, should the liability be joint and several, ac-cruing to all authors? If not, then how should it be allocated among them? If thereis an honest mistake in one part of the work but not in others, how should an eval-uator aim his or her critique? Such questions plagued the committee that examined the recent high-profile case of fraud in the physics community, the Schön affair,and surely will trouble others.

When penalties for research misconduct are considered, it is often argued thatan identification of each author’s role in the research should be required, in order to help us fix blame. Critics of the notion that authors should share the blame ask,for example, “how can the molecular biologist be expected to certify the honestyand quality of the crystallographer’s work?” Some would answer “by knowing that

person well enough to rely on him or her.” I rather like that response, so with re-

spect to assigning blame for research misconduct, I take the “joint and several” position, knowingthat it puts me in a quirky minority.

Various practical or impractical suggestions have emerged during the longstanding debate on thisissue. One is that each author should provide, and the journal should then publish, an account of thatauthor’s particular contribution to the work. Although Science will make it possible for authors todo that, we cannot monitor the authors’ designations or negotiate possible disputes over which au-thor actually did what (there’s enough of that already, thank you). And listing the individual contri-

butions of each of a couple of dozen authors will, even if it appears only electronically, add somelength and complexity to the communication.

But a different view of the problem, and perhaps of the solution, comes as we get to universitycommittees on appointments and promotions, which is where the authorship rubber really meets theroad. Half a lifetime of involvement with this process has taught me how much authorship matters.I have watched committees attempting to decode sequences of names (“is it good or bad that her ma-

jor professor’s name wasn’t at the end of the author roster?”), agonize over whether a much-cited paper was really the candidate’s work or a coauthor’s, and send back recommendations asking for more

specificity about the division of responsibility.Problems of this kind change the argument, supporting the case for asking authors to define their

own roles. After all, if quality judgments about individuals are to be made on the basis of their per-sonal contributions, then the judges better know what they did. But if questions arise about the va-lidity of the work as a whole, whether as challenges to its conduct or as evaluations of its influencein the field, a team is a team, and the members should share the credit or the blame. Thus, Sciencewould be glad to see authors define their roles—briefly, please!—but has no plans to pass out the

Newcomb Cleveland prize, our annual award for the best Science paper, in little bits and pieces.

Donald Kennedy

Editor-in-Chief

Multiple Authors, Multiple Problems

A team isa team, and

the members

should share

the credit or

the blame.


19/21

L E T T E R S

www.sciencemag.org SCIENCE VOL 302 3 OCTOBER 2003

Ph.D. in a relevant field, but who worksunder the auspices of a “lobbying” organiza-tion, be considered worthy of appearing onCSN? What reasoned standard would exclude this “expert,” but not a scientist withsimilar academic credentials who works at auniversity? And what about scientists whowork at “unbiased” institutions, but whoreceive funding from “biased” sources?

I think that a CSN could be a valuable addi-tion to the sources of scientific informationcurrently available to the public, to legislators,and to administrative officials, but it is goingto be very difficult—if not impossible—todistinguish “biased” from “unbiased” sourcesof information on the brief time scale required for 24-7 television programming.

MICHAEL N. NITABACH

Department of Biology, New York University, 100

Washington Square East, New York, NY 10003, USA.

E-mail: [email protected]

ResponseNITABACH RAISES THE IMPORTANT ISSUE OFhow to ensure that CSN presents a fair and unbiased view of scientific issues. Onecornerstone of avoiding bias on CSN would

be a scientific advisory board with stellar credentials (already being formed) thatcould provide advice about programming.

Scientific societies would also be a source of advice and support. Many scientific societieshave media outreach programs that includetaped lectures, interviews, and press confer-ences, and CSN could serve as a central nexusfor all of these efforts, disseminating them toa wider audience. It will be essential to

present a balanced view of both sides of controversial issues. We see this more as an

opportunity to engage the audience with alively debate.

Nitabach is not quite correct when he saysthat “C-SPAN’s goal—to broadcast officialgovernmental proceedings and statements— is an easy one to achieve from a programming

perspective.” C-SPAN’s mission statement (1)talks of providing access to governmental

proceedings “with a balanced presentation of points of view”; of providing a forum for people who influence public policy “withoutfiltering or otherwise distorting their points of view”; of providing access, through call-in

programs, to decision-makers; and of employing “production values that accuratelyconvey the business of government rather than distract from it.” Substitute “science” for “government,” and you have CSN.

Broadcasting governmental proceedings isonly a part of C-SPAN’s schedule. The network also carries a broad spectrum of nongovern-

mental meetings, book readings, history, and advocacy lectures (clearly identified). C-SPANstrives to be an honest broker in allocating air-time. CSN would do the same.

TERRENCE J. SEJNOWSKI1 AND ROGER BINGHAM2

1Howard Hughes Medical Institute, Salk Institute

for Biological Studies, 10010 N. Torrey Pines Road,

La Jolla, CA 92037, USA. E-mail: [email protected] for Brain and Cognition, Department of

Psychology, University of California, San Diego,9500 Gilman Drive, La Jolla,CA 92093–0109, USA.

E-mail: [email protected]

Reference1. S.Frantzich,J. J.Sullivan, The C-SPAN Revolution (Univ.

of Oklahoma Press, Tulsa, OK, 1996),p. 2.

A Suggestion for theMultiple Author Issue

DONALD KENNEDY’S SUPERB AND LONG-overdue Editorial “Multiple authors,multiple problems” (8 Aug., p. 733) raisesthe increasing problem of too many authorson scientific papers. This is clearly a case of dilution of importance—how is one toappreciate the importance of an author of a

paper with more than 50 coauthors?I would like to propose a possible solution

that should clarify this issue while allowingrecognition of important technical contribu-


20/21

L E T T E R S

www.sciencemag.org SCIENCE VOL 301 3 OCTOBER 2003

tions. Only those individuals who were inti-mately involved in (i) experimental design,(ii) data acquisition, (iii) data analysis and interpretation, and (iv) writing and editingshould be listed as authors. Technical contri-

butions (e.g., a specific measurement) could be acknowledged as a separate list identifyingthe specifics. This could be done in a smallfont so that space requirements are mini-

mized. Those listed for technical contribu-tions could cite this in their CV under a sepa-rate category, thus getting “credit” for promo-tion, tenure, and grant applications.

It is important to note that in addition togiving credit where credit is due, this would

protect coauthors from any guilt by associa-tion if scientific misconduct was discovered in parts of the publication not related to aspecific contribution.

HERSHEL RAFF

St. Luke’s Medical Center, Medical College of

Wisconsin, 2801 West KK River Parkway,Suite 245,

Milwaukee,WI 53215,USA. E-mail: [email protected]

A HeterozygoteAdvantage

THE EVIDENCE FOR BALANCING SELECTION AT

the prion protein gene ( PRNP ) due to kuru in

the Fore group of the Papua New GuineaHighlands is compelling (“Balancing selec-tion at the prion protein gene consistent with

prehistoric kurulike epidemics,” S. Mead et al., Reports, 25 April, p. 640). That is, their analysis of worldwide haplotype diversityand sequence analysis demonstrates that themajor alleles at the PRNP locus are main-tained by selective factors favoring the main-

tenance of heterozygotes. In addition, theextent of the “heterozygote advantage” in theFore in terms of their viability in the presentgeneration can be calculated from Mead et al.’s genotypic data (provided by S. Mead).In 30 women over the age of 50 that had ahistory of multiple exposures to mortuaryfeasts, 4 were homozygous MM , 23 wereheterozygous MV , and 3 were homozygousVV ( M and V indicate methonine and valineat position 129), a large deviation fromHardy-Weinberg proportions. In another sample of unexposed Fore individuals, thegenotypes were in Hardy-Weinberg propor-tions (31 MM , 72 MV , and 37 VV ). Usingthese two groups as the frequencies of thegenotypes after (indicated by primes below)and before selection, the viability of genotype MM relative to genotype MV can be estimated (1) as V MM = ( P ′ MM P MV )/( P ′ MV P MM ) =(0.133)(0.514)/(0.767)(0.221) = 0.403, and

the viability of genotype VV relative to geno-type MV can be estimated as V VV =( P ′VV P MV )/( P ′ MV P VV ) = (0.100)(0.514)/(0.767)(0.264) = 0.254. In other words, therelative viabilities of the genotypes MM , MV ,and VV are 0.403, 1.0, and 0.254, respectively,a very strong heterozygote advantage in theface of kuru.

Because adult males participated little

at feasts, this heterozygote advantage acts primarily in females. Therefore, theaverage selection coefficient ( s = 1 – V )against MM homozygotes is approximately- s MM = (1 − V MM )/2 = 0.299, and against VV homozygotes, it is - sVV = (1 – V VV ) /2 =0.373. The expected equilibrium frequencyof the V allele is therefore qV =

- s MM /(- s MM +

- sVV ) = 0.45, not very different from theobserved frequency of 0.55. Although it isnot known whether selection has been thisstrong in previous generations, the strengthof balancing selection in this one genera-tion appears to be the strongest yet docu-mented in any human population.

PHILIP W. HEDRICK

School of Life Sciences, Arizona State University,

Tempe,AZ 85287,USA. E-mail:philip.hedrick @asu.edu

Reference

1. P. W. Hedrick, Genetics of Populations (Jones &Bartlett, Boston, ed.2, 2000).


21/21

E D I T : J O E S U T L I F F

Biofuels: Clarifying

Assumptions

THE REPORT BY T. SEARCHINGER ET AL.(“Use of U.S. croplands for biofuels in-

creases greenhouse gases through emissionsfrom land-use change,” 29 February, p. 1238)

provides one scenario for the conversion

from a fossil-based energy economy to a bio-

based, renewable-energy economy. However,

Searchinger et al . failed to include several

important considerations.

It is inaccurate and misleading to allocate

the cutting down of Brazilian rainforest,

which is done often for timber production, to

biofuels use. The economic signals driving

different from the timber-driven economic

signals driving land-use change patterns. The

deforestation estimates of Searchinger et al .

are appropriate for biodiesel production in the

Far East. A cheaper and more likely use of

land for increased biofuels production is the 6

billion acres of underutilized or unused rain-fed agriculture land available, according to a

Food and Agriculture Organization report (1).

Searchinger et al . analyze switchgrass as

an energy crop when miscanthus and sorghum

have much higher yields [a recent study esti-

mated that miscanthus yields are 250% that

of switchgrass (2)] and would dramatically

reduce the demand for land. Furthermore, be-

cause these crops have not been optimized for

biomass, they are likely to produce substantial

oretical maximum yield of 40 to 50 tons p

acre in a region with an average of 40 inches

rain, practical yields of 50 to 60% of this ma

imum are likely. It has even been suggeste

that maximum theoretical yield values will b

reached and possibly surpassed (3).

Searchinger et al . assume that crops growin developing countries will have lower yield

The yields are lower because of low prices an

lack of farmer income. In these condition

farmers cannot afford the best seed crops an

other inputs such as fertilizer (1). It is like

that if farmer incomes improve, yields w

also increase.

Searchinger et al . state that “[h]igh

prices triggered by biofuels will accelera

forest and grassland conversion there even

LETTERS I BOOKS I POLICY FORUM I EDUCATION FORUM I PERSPECTIVE

378

Toward perfect ceramics Transforming light control

383 384

A place of life

LETTERSedited by Jennifer Sills

Quantifying Coauthor Contributions

FIFTY YEARS AGO IN SCIENCE , D. MCCONNELL ARGUED THAT “FORanything short of a monographic treatment, the indication of more than

three authors is not justifiable” (1). He was never cited. Coauthor num-

bers kept rising, and it has been recently suggested that in some fields

“multiple authorship endangers the author credit system” (2). In 2006,

more than 100 papers had over 500coauthors, and one physics paper had

a record 2512 coauthors (3). With

research groups growing larger (4),

this trend will continue. Given the

increasing interest in the quantifica-

tion and standardization of scientific

impact with various metrics like the

h index (5, 6 ) and the growing debate

on potential biases (7 , 8) and unethical

behavior (4, 9, 10), a standardized

method to quantify coauthor contribu-

tions is needed (10–13).

Rarely do all coauthors contributeto a paper equally. However, aca-

demic search engines (such as Google

Scholar, Scopus, and Web of Science)

calculate citations, h indices, and rank-

ings without regard to author rank.

Quantification of coauthor contributions will motivate coauthors t

clarify each person’s percent of contribution.

I propose that the k th ranked coauthor be considered to contribu

1/k as much as the first author. This way, coauthors’contributions can b

standardized to sum to one, regardless of the author number or ho

authors are ranked. Author rank can be different from author order, pro

vided that this is declared in the paper. Multiple authors can have th

same rank, as long as this is stated and is reflected in the calculations.

Quantifying coauthors’contributions will encourage a healthy dilogue about the meaning of coauthorship and author rank (2, 4, 1

13), will promote better consideration of author rank in assessing sc

entific impact, and will lead to improved ways to measure and repo

coauthor contributions.

CAGAN H. SEKERCIOGL

Department of Biology, Center for Conservation Biology, Stanford University, Stanford, 94305, USA. E-mail: [email protected]

References

1. D. McConnell, Science 128, 1157 (1958).2. M. Greene, Nature 450, 1165 (2007).3. C. King, Sci. Watch 18, 1 (2007).4. P. A. Lawrence, Curr. Biol. 17, R583 (2007).5. J. E. Hirsch, Proc. Natl. Acad. Sci. U.S.A. 104, 19193 (2007).6. P. Ball, Nature 448, 737 (2007).

7. C. D. Kelly, M. D. Jennions, Nature 449, 403 (2007).8. D. C. Mishra, Nature 451, 244 (2008).9. P. A. Todd, R. J. Ladle, Nature 451, 244 (2008).

10. D. Kennedy, Science 301, 733 (2003).11. R. Hunt, Nature 352, 187 (1991).12. J. V. Verhagen, K. J. Wallace, S. C. Collins, T. R. Scott, Nature 426, 602 (2003).13. W. F. Laurance, Nature 442, 26 (2006).

COMMENTARY

Documents

Authorship Readings