Sci Eng Ethics DOI 10.1007/s11948-015-9722-5 ORIGINAL PAPER

Trouble in Paradise: Problems in Academic Research Co-authoring Barry Bozeman1 • Jan Youtie2

Received: 19 February 2015 / Accepted: 3 November 2015  Springer Science+Business Media Dordrecht 2015

Abstract Scholars and policy-makers have expressed concerns about the crediting of coauthors in research publications. Most such problems fall into one of two categories, excluding deserving contributors or including undeserving ones. But our research shows that there is no consensus on ‘‘deserving’’ or on what type of contribution suffices for co-authorship award. Our study uses qualitative data, including interviews with 60 US academic science or engineering researchers in 14 disciplines in a set of geographically distributed research-intensive universities. We also employ data from 161 website posts provided by 93 study participants, again US academic scientists. We examine a variety of factors related to perceived unwarranted exclusion from co-author credit and unwarranted inclusion, providing an empirically-informed conceptual model to explain co-author crediting outcomes. Determinants of outcomes include characteristics of disciplines and fields, institutional work culture, power dynamics and team-specific norms and decision processes. Keywords Research collaboration
Co-authorship
Ghost authors
Guest author
Contributorship

& Barry Bozeman [email protected]; [email protected] Jan Youtie [email protected] 1

Center for Organization Research and Design, School of Public Affairs, Arizona State University, Phoenix, AZ 85004, USA

2

Enterprise Innovation Institute, Georgia Institute of Technology, Atlanta, GA 30308, USA

123

B. Bozeman, J. Youtie

Introduction Today’s science, technology, engineering and mathematics (STEM) research is almost always a team enterprise. Bozeman and Boardman (2014) use the term ‘‘collaborative imperative,’’ suggesting that in many cases research collaboration is essentially a prerequisite of contemporary research. Despite some variation by field, discipline and geography (De Stefano et al. 2013), increasingly STEM research occurs within networks and is dominated by teams, often quite large ones, entailing co-authorships that are forged for many different reasons and employing a wide variety of rules and norms. One of the best reasons to collaborate is simply that collaboration enhances effectiveness. According to Wuchty et al. (2007) collaborative research, compared to individual investigator research, more often leads to high impact research and to patents. Lee and Bozeman (2005) provide evidence that co-authored research garners more citations and research by Gaughan and Ponomariov (2008) echoes those findings. But rational expectations of higher impacts provide only one reason for the collaborative imperative. Recent trends toward increased research collaboration and co-authorship have also been attributed to needs to share resources (Nedeva et al. 1999; Ynalvez and Shrum 2011), including large-scale technology and equipment but also samples and specimens (Austin et al. 2012). Other researchers (Porter and Rafols 2009; Haustein et al. 2011) point to the growing specialization of science and the interdependence of fields, indicating that in some cases it is impossible for one person to have the specialized knowledge requisite for research, especially interdisciplinary research. Research collaboration is abetted by new collaborative technologies (Zhou et al. 2012; Muscio and Pozzali 2013; Tacke 2011), not only the most ubiquitous and obvious ones such as email, electronic file transfer and video conferencing, but also more exotic ones (see Olson et al. 2013; Sonnewald 2014). Interacting with each of these drivers of increased collaboration, public policies increasingly encourage or even mandate collaboration (Ponomariov and Boardman 2010; Wallerstein and Duran 2010; Van Rijnsoever and Hessels 2011). In this study, our focus is on one major aspect of collaboration, co-authorship of publications, especially journal articles. We focus particularly on problems in coauthorship crediting, seeking to identify types of problems encountered and to develop a preliminary model of determinants of co-author crediting disputes. Most, not all, such crediting problems fall into one of two categories, excluding deserving contributors or including undeserving ones. But our study shows, as do previous studies (e.g. Laudel 2002; Birnholtz 2006), that there is no consensus on ‘‘deserving’’ or on what type of contribution suffices for co-authorship award. We explore these issues using qualitative data, including interviews with 60 US academic researchers in 14 sciences and engineering disciplines in a set of geographically distributed research-intensive universities. We also employ data from 161 website posts provided by 93 study participants, again US academic scientists.

123

Trouble in Paradise: Problems in Academic Research…

Collaboration and Co-authorship Not everyone agrees as to the nature of research collaboration or what qualifies as collaboration and it is thus understandable that there is not universal agreement on what collaborative contributions warrant co-author status. Bozeman and Boardman (2014, p. 12) define research collaboration as ‘‘social processes whereby human beings pool their experience, knowledge and social skills with the objective of producing new knowledge, including knowledge as embedded in technology.’’ These experiences may or may not lead to co-authorship and, just as important, some awarded co-author status are not involved in the pooling of effort but, instead, may be included to provide resources, authority, or need, among other factors. Thus, collaboration and co-authorship are related but certainly not identical. While at first blush it may seem as though co-authoring is simply a subset of collaboration that is not actually the case. Whereas there are many different definitions and measures of collaboration (see Katz and Martin 1997; Melin 2000; Bozeman et al. 2013; Bozeman and Boardman 2014), co-authorship can be determined by simple and convenient measures, typically just author crediting in published works. However, the matter of author crediting is not at all straightforward. Clearly, it is possible for individuals to be involved in research collaborations but not be named as authors. For example, some of those persons who ‘‘pool their experience, knowledge and social skills,’’ to use the Bozeman and colleagues’ (in press) definition, may include research technicians, computer programmers, research managers and others who make a significant contribution but are not generally included as co-authors. Perhaps more worrisome, some listed as coauthors may have made minimal or no contributions to the research but are included as co-authors because of author trading (you make me an author in yours and I will make you an author in mine), because of hierarchical authority, or because they have pressured colleagues or even threatened those over whom they have power. There is remarkably little evidence about co-author crediting decisions and behaviors. Laudel (2002) provides some very useful qualitative data based on a set of interviews with research group members in science centers. She identifies what is essentially a gradient of behaviors more or less likely to lead to the awarding of coauthorship. When there is a clear division of labor and individuals have contributed importantly to an aspect that is clearly a requisite to the work they are likely to awarded co-authorship. In the case of ‘‘service collaborations’’ one of the collaborators sets the goal and design of the work and others perform routine tasks; the collaboration input may be ‘‘weaker’’ and the assurance of co-authorship less. Still weaker is the ‘‘provision of access to research equipment’’ collaboration or, in Birnholtz’ (2006) terms, ‘‘infrastructure’’ contributions. In such cases a specific collaborator may perform none of the research tasks but rather share equipment. Similar relationships can be found in the providing of samples. Such provision collaborations often do not lead to co-authorships. A fourth and especially murky category is ‘‘transmission of knowledge,’’ in which case an individual may describe a procedure, give advice about relevant theory or provide knowledge of a research operation but not otherwise engage in the actual research. The awarding of co-authorship seems uncommon in such cases but is certainly not unprecedented.

123

B. Bozeman, J. Youtie

In sum, co-authoring is not a sub-set of research collaboration because some who collaborate are not given co-author status and some who are given or claim such status have not been truly involved in collaboration. The rules for awarding coauthor credit are certainly not hard and fast and vary according to discipline, personal preference, need for credit and, quite often, negotiation. Problems occur when individuals have divergent ideas about what factors should and should not be rewarded with co-author credit. Moreover these informal decision rules about crediting are not usually explicit, even among the researchers directly involved (Youtie and Bozeman 2014; Bozeman et al. 2015) and, thus, there is in many cases a strong likelihood of divergent views about crediting. Problems in co-author crediting can be sorted into two overlapping categories. In one category there are what we refer to as ‘‘decision structure problems’’ and in the other there are ‘‘ethical problems.’’ Research ethics problems have received a good deal of attention (Plemmons 2012; Saha and Hurlbut 2011; Va¨ha¨kangas 2013),1 but co-authoring ethics have received less attention and the effect of decision structure problems in co-authorship almost none. We can think of co-author crediting problems as (with some exceptions) falling into one of two categories: either failing to credit persons who are deserving of co-author status are awarding co-author credit to those who have done little or nothing to deserve it. Either type of collaboration error can be the result of ethical problems, just as either type can result from flawed (but not unethical) decision processes. While ethical problems take many forms, some examples include bullying colleagues, exploiting students, or generally making claims one knows to be untrue or unwarranted. Ethical Issues and Co-authorship Much of the attention on co-authoring ethics has come from the medical and biotechnology fields and, possibly, problems are especially pronounced in those fields. Many have focused on so-called ‘‘guest’’ authors and ‘‘honorary’’ authors (for an overview Levsky et al. 2007), persons listed as authors who make no significant contribution. Some attribute problems with sorting out co-authorship and crediting to the explosion in research and the funding imperatives driving collaboration among investigators from multiple sites and numerous disciplines (Devine et al. 2005; Drenth 1998). Ultimately the system of scientific authorship is built on trust that the published work reflects the data and analysis of the authors (Lagnado 2003). We contend that the co-authorship choices relate closely to the integrity of the research. Specifically, the integrity of the research can be undermined in (at least) the following ways: 1.

In cases where authors, especially lead or corresponding authors, make no substantial contribution to the research, authorship claims are essentially scientific fraud.

1

Much of the attention to unethical behavior has to do with research fraud and falsification of result (e.g. Steen 2011; Kornfeld 2012). That is only one type of unethical behavior in research and not our focus here. We are interested only in unethical behavior in co-author crediting.

123

Trouble in Paradise: Problems in Academic Research…

2.

3.

4.

5.

Control of authorship award and co-authorship credit can in some instances be a ‘‘weapon’’ that the powerful use to obtain resources, knowledge or obedience from the less powerful. Since grants and contracts and other important research-related resources are provided in part on the basis of scientific reputation and apparent productivity, measured in terms of the number and quality of publication, misattribution of co-author credit undermines the effective allocation of science resources. In cases where ‘‘legitimate’’ collaborators are not included as co-authors and not provided their due in the collaboration, there is a possibility that data privileges and other resource controls can act as, essentially, a restraint of trade. Conflicts of interest in STEM research, often revealed in the authorship of scholarly literature, are problematic for obvious reasons (McCrary et al. 2000).

The last integrity threat enumerated above, conflict of interest, has received a good deal of attention by researchers (see Mowery et al. 2001 for an overview) examining the implications of ever-increasing commercialization of research, especially in the wake of the massive changes in intellectual property ushered in with the Bayh-Dole Act of 1980 (which enable research institutions, including universities, to license the intellectual property of their employees, even in cases where the research is based on federal funds). However, most of the attention is on conflicts between the researchers and the university, not conflicts among researchers seeking to parse their relative contribution. Co-authorship stakes loom large here. We add to this the possibility of exploitation, real or perceived, usually of the more junior and inexperienced researcher. We do not presume that most unequal status collaborations result in unethical or otherwise bad outcomes. On the contrary, we assume that most collaborations are instrumental for all parties rather than either helping or exploitative and that, for the remainder, helping behavior is more common than exploitative (see Green and Bauer 1995; Tenenbaum et al. 2001; Bozeman and Corley 2004). One key issue is to determine if exploitative outcomes can be explained by situational or structural variables. For example, Green and Bauer (1995) suggest that outcomes may relate to senior researchers’ assessments of ability and quality of students’ and postdocs contributions.

If Co-author Crediting is a Problem, is Contributorship Policy the Solution? Partly because of a high level of perceived threat to research ethics, biomedical researchers and editors have taken the lead in identifying and moving to resolve ethical problems in collaboration. The International Committee of Medical Journal Editors (ICMJE, known as the Vancouver Group) created a set of ‘‘Uniform Requirements’’ for authorship in 1985. But by the mid-1990’s, these protocols were still employed by only a handful of journals. Drummond Rennie, a deputy editor of the Journal of the American Medical Association, and a strong proponent of contributorship policies, acknowledged this deficit in an editorial colorfully subtitled, ‘‘Guests, ghosts, grafters, and the two-sided coin’’ (Rennie and Flanagin

123

B. Bozeman, J. Youtie

1994). Contributorship is defined as authors declaring in detail, in advance of publication, their individual contributions to scholarly papers (Rennie 2001, p. 1274). In the medical and biotechnology research realm there have been strong advocates of ‘‘contributorship’’ (Rennie et al. 1997; Council of Science Editors 1999; Devine et al. 2005) as a means of combating ethical problems. Following a series of articles that describe a growing problem with irresponsible authorship of medical research articles, Rennie proposed a major change in instructions to authors to JAMA (2000, p. 89). These changes in contributorship requirements have provided clear signals where none were given before and, presumably, have enhanced collaborators’ ability to communicate effectively with one another about contribution and credit. Contributorship policies have been developed and adopted in many biomedical and health sciences fields but are less common in other STEM fields (Council of Science Editors 1999; Washburn 2008). Other policies related to contributorship have been proposed and some implemented. For example, Marusic et al. (2004) and Pichini et al. (2005) describe the many international Uniform Requirements for coauthorship information, including policies building on the Vancouver protocols International Committee of Medical Journal Editors (1997). Contributorship policies and norms are viewed as increasing transparency and fairness, not only according to Rennie et al. (1997, 2000), but other researchers and journal editors (Smith 2012; Hayter et al. 2013). Despite some progress in journals and editorial boards attending to ethical issues, one certainly cannot assume that ethical hazards are vitiated by current editorial and journal policies. In the case of contributorship approaches, there is not much evidence as to whether contributorship policies have diminished unethical behavior. One problem is the veracity of contributorship statements and the possible social and potential power dynamics entailed in developing them. Moreover, contributorship policies are not yet widespread. Resnick et al. (2009) gathered data from a random sample of 399 journals listed in the ISI Web of Knowledge database. Not only were crediting and co-authorship policies very uncommon, the majority did not have misconduct policies of any sort. Responses from 197 editors and publishing officials (49.4 % response rate) indicated that: 47.7 % of the journals had a formal (written) misconduct policy; 28.9 % had a policy that only outlined procedures for handling misconduct, 15.7 % had a policy that only defined misconduct, 10.2 % had a policy that included both a definition and procedures. The sources of policies varied with 26.9 % of journal policies developed by the publisher, 13.2 % by the journal, and 14.7 % had a policy that was generated by another source, such as a professional association. Contributorship is perhaps the best known approach to co-author crediting issues but other approaches have been suggested including systematic evaluation of human resources management in research projects and programs (Thune 2009) and more extreme approaches (for an overview see Birukou et al. 2011) that would alter the meaning of peer reviewed research and crediting such as Research Gate’s Open Review (ResearchGate Press Office 2014), and post-publication reader review (Bozeman and Gaughan 2011).

123

Trouble in Paradise: Problems in Academic Research…

If Decision Structure is the Problem, then What is the Solution? Almost as important as co-author credit is the decision process by which co-authorship is decided. Decisions about assigning authorship, credit, and co-author order are highly varied and often provide only an oblique signal as to who has done what. One problem is that scientific fields and even work groups within fields vary substantially in their practices for assigning co-authoring credit. For example, in some cases first authorship means that the individual made the most significant scientific and intellectual contributions to the research, but in other cases it means that the individual was the lab director or the principle investigator and may or may not have had direct involvement in the research. An alternative practice—alphabetizing authorship order—would seem to reflect more ‘‘fairness’’ but it also lacks explicit information as to which author is primarily responsible for the work. Some fields, particularly those for which numbers of authors of 100 or more, also known as ‘‘hyperauthorship’’, or are not uncommon (i.e., in certain subfields of physics such as astrophysics) even use randomization techniques for the same fairness rationale; this method even further muddies the ability to trace authorship contribution (Davenport and Cronin 2001; Birnholtz 2006). As decision analysts have known for years, often process is the primary determinant of outcome (Brockner and Wiesenfeld 1996). While there is remarkably little evidence about collaboration and co-authorship decision processes and norms, most agree that these vital processes affect not only scientific career trajectories and advancement but also the very course of science (Katz and Martin 1997; Melin 2000). The choice of scientific topics and the configuration of research teams depend in part on collaborative and co-authorship norms. In the vast majority of instances, researchers have considerable autonomy in their collaboration choices and collaboration strategies are based in part on judgments about the conferring of co-authorship and status (Heffner 1981; Bozeman et al. 2015). The issue is who decides. From companion research (Youtie and Bozeman 2014) we have learned that decision structure problems in co-authoring most often occur simply because one or more of the authors, usually the most senior one, assumes tacit agreement and consensus when there is none and then decides on co-authoring with little or no discussion. While that is by far the most common decision process problem, others pertain to problems well known to group dynamics and group decision-making, such as suppression of dissent (Nemeth 1986; Tollefsen 2006) and ‘‘groupthink’’ (Janis 1982; MaCauley 1989; Esser 1998; Baron 2005), ‘‘squeaky wheel’’ phenomenon (Plous 1993; Labianca et al. 1998) whereby the most vocal opinion wins the day, and various forms of ‘‘halo effects,’’ such as assuming that one’s scientific leadership or excellence implies superior decision acumen (Do¨rner and Schaub 1994; McShane et al. 2013).

Data As part of our broader study of collaboration, one focusing on collaboration processes, norms, decision-making, and effectiveness, we included efforts to identify problems in contributorship. Our initial methodology was to develop

123

B. Bozeman, J. Youtie

questionnaires. For the questionnaire data (see Youtie and Bozeman 2014 for full account of procedures) we asked respondents to think about the most recent published article that involves co-authors but did not ask them to specify either the article or the co-authors but rather to frame their responses in terms of that article that they would keep in mind. This methodology proved effective for most of the broader study’s objectives but less so for the current paper’s purpose of examining co-authoring and contributorship. Two other approaches, semi-structured interviews and gathering of web post data, proved more helpful and the data from this paper are from those efforts, as discussed below. The chief relevance of the survey approach is that it provided the sampling population for the interviews and the respondents for the web post request. The sampling population was drawn from the NSF’s categories in its Survey of Earned Doctorates and we focused on science and technology categories. We received 641 completed or mostly completed questionnaires, for a 36 % response rate. Respondents were very similar to the population in terms of gender, rank and departmental discipline. The Interview Data We and other research team members conducted in-person and telephone interviews with 60 academic scientists in 14 STEM disciplines in 13 US states. The STEM disciplines were: biology, computer science, mathematics, physics, earth and atmospheric science, chemistry, chemical engineering, economics, civil engineering, electrical engineering, mechanical engineering, materials engineering/materials science, industrial/manufacturing engineering, and biomedical engineering. In-person interviews were conducted with faculty at geographically distributed universities in Arizona, California, Georgia, South Carolina, and Massachusetts. Additional telephone interviews were conducted with faculty at universities in Colorado, Delaware, Georgia, Indiana, Iowa, Kansas, Massachusetts, Mississippi, New Mexico, New York, North Carolina, Pennsylvania, South Carolina, Texas, Utah, Virginia, and Washington. Interview subjects were selected at random based on their affiliation with schools and departments in the 14 STEM disciplines. Senior assistant, associate, and full professors were selected to ensure a sufficient range of research collaboration experience for discussion (although a few non-tenure track researchers wound up being included as well). The interviews were held in the spring, summer, and autumn of 2012 and the first two months of 2013, typically in the offices of the faculty being interviewed. Each interview lasted from 30 min to an hour or more. Notes were either handwritten or typed into a computer by a second interview team member. Interview questions covered topics about the faculty member’s research characteristics, nature of research collaborations, decision-making processes for determining authorship and author order, the role of policies such as promotion and tenure, and recounting of best and worst research collaborations. The interview protocol allowed for flexibility to adapt to each interview subject’s situation. For example, some interview subjects did not initially report ever having any collaboration problems. The interviewer was able to use prompts such as asking about a colleague’s problematic experiences or ‘‘urban legends’’ in the field to obtain such information;

123

Trouble in Paradise: Problems in Academic Research…

sometimes these accounts of other colleagues would prompt interview subjects to recollect problematic collaborations of their own. The Website Posts Instructions for the open-ended qualitative survey responses to be routed to the secondary website were as follows: Please tell us about an especially positive or an especially negative collaboration experience (or one of each). As indicated, our study seeks to understand research collaboration dynamics, including perceived quality of collaborations. We are especially interested in extreme cases, ones where collaborations are especially beneficial and effective, or ones that are highly disruptive or ineffective. If you have an anecdote to share we would be very grateful. To gather such anecdotes confidentially we have set up a separate website. If you click below you will be taken to this separate website and any information you care to give us will remain confidential. As with the rest of the survey, your responses will not be individually identifiable. We hope you will take just a few minutes and share one or two interesting anecdotes, either positive or negative. Thank you for considering our request. Whether or not you wish to share an anecdote, please accept our thanks for participating in this study. Please visit the following website to share your experiences confidentially: [website information redacted]. The website posts provided 161 comments, many of them quite extensive, from 93 different participants. More than three quarters of these responses were about negative experiences, perhaps not surprising since research (Watson and Clark 1984; Nabi 1999) shows that people are likely to report especially salient experiences and ones that are easily recalled and, moreover, that highly negative experiences are particularly salient and easier to recall. Analytical Approach Invariably, an issue with qualitative data is organizing the analysis in a way that exploits key findings but at that same time gives some ability to generalize and draw conclusions. Our approach is similar in some respects to grounded theory. While we had some expectations that colored the interview data (for the analytical model used see Bozeman et al. 2015), the web-posted data are entirely open-ended, framed only by our request for extremely good or extremely bad research collaboration experiences (we focus here on bad outcomes as most of the reports of good outcomes were quite homogeneous). We refrained from more detailed explication because we had no desire to color or influence the reports we received. But if this open-ended approach is useful for obtaining respondent-driven data, it does not suggest an analytical approach. We felt that a good place to start would be a quite simple analytical distinction in crediting errors, excluding the deserving vs. including the undeserving.

123

B. Bozeman, J. Youtie

Though most co-author problems can be sorted into one of these two simple categories, there are many different routes to these extreme outcomes. Some are commonplace and others quite unusual. We see that just a few issues run through a great many of the co-author problems. One pervasive issue has to do with the fact that various STEM fields and disciplines have different norms and one of the things we see from the reports below is that quite a few co-authoring disputes occur when different disciplines come together, bringing along disparate norms and expectations. Power dynamics, likewise, provide threads that run through much of the data. This most often occurs in different but sometimes interrelated sub-categories, ones that often interact. First, gender conflict is a frequent theme. Second, power dynamics are almost always important in supervisor-subordinate and senior-junior relations. In the context of universities we can be more specific and distinguish senior versus junior relations and student versus faculty relations. As one might expect, gender conflict also is frequently intertwined with power dynamics. Each of these predominant themes is reflected below, but in a subsequent section we seek to refine these major determinants of co-authoring problems, including presenting a conceptual model that reflects many of the general themes identified here.

Excluding the Deserving: Failing to Assign Credit when Credit is Due Our concern here is why deserving people do not receive co-authoring credit. We begin with a caveat. It is not always easy to determine who is and is not deserving (Laudel 2002). In the first place we are relying on first hand accounts by interested parties. We can expect something less than an impartial judgment. Perhaps even more important, some co-authoring credit claims are at the margins very difficult to adjudicate. It is not possible to say in any absolute sense how much of a contribution should suffice to warrant co-authorship. Reasonable people disagree on this. Overall, we have somewhat more cases of the undeserving being included than the deserving excluded. Further, there seems to be a bit more predictability as to deserving people getting squeezed out as compared to the many different ways in which undeserving people manage to insinuate themselves. The quotes provided immediately below give an indication of the most common ways in which people who seem deserving fail to receive co-author credit. After discussion of author exclusion we look at the related problem of a lack of consultation on author order and similar issues. Author Exclusion If our data provide a full account, just a few interacting variables explain author exclusion: (1) geographic separation and, especially, relocation, (2) differential in power and experience, (3) disagreements about the role of technical contributions, and (4) gender dynamics (which generally interact with power differentials). The first of these issues in co-author exclusion brings to mind the aphorism: ‘‘out of sight, out of mind.’’ If persons, especially less experienced individuals, leave a lab or a company or a research team and move to another site, there seems to be a

123

Trouble in Paradise: Problems in Academic Research…

stronger likelihood that unwarranted exclusion will occur. It is perhaps unsurprising that when people relocate they are more likely to fall prey to unethical behavior or to a lack of care. Research in other realms shows a close connection between unethical behavior or corruption and both geographic and social distance (ZeyFerrell and Ferrell 1982; Brass et al. 1998; Gino et al. 2009). There was a case where I was co-author on a conference paper that resulted from a funded grant I was co-PI on but I was not included on the final journal publication because I had left that university. I had to request that I be a coauthor on a journal paper in which I was co-PI for the grant that funded the research, and the paper was related to a specific research question that I raised during the proposal process. I was lucky to catch that publication right before final printing. Here are two other location-related snubs: I did great work for a company for two years, then left to go to graduate school. They later published a journal article in a professional society transaction but forgot to include me as coauthor. I have never forgotten, and have made sure that it never happens to someone who works with me. I spent years setting up an experimental system and shared the methods and resources with a colleague who did not have the expertise. We worked collaboratively until I was away from work for a period of time due to serious illness. During that time, the colleague published a paper without offering authorship or even mentioning it to me. The paper beat one of my own publications to press, and that colleague is now credited with the work that I developed. Crediting is not always a matter of the deserving person leaving. Sometimes the person who fails to award credit is the one who has relocated and perhaps at a distance feels safer wresting credit away from another colleague: A professor from a Japanese university sent his student to work with me; we worked together and produced some results. Upon returning to her home institution this student defended her MS thesis, of which more than 80 % was based on work done with me, without even mentioning our work, even though that work had been published and presented in a [co-authored] conference paper. It was as if we had never even met. That student’s advisor got all the credit although he knew very well that he [the advisor] did not contribute anything on that work. In a different situation I had a visitor from Korea with whom I spent considerable amount of time to teach him a certain technique, and we wrote a paper together, which we did not publish before he left. He just disappeared, only to find out two years later that he had translated that paper into Korean and published it under his own name. Another common reason for undeserved exclusion is that individuals provide a service or a technical contribution, expecting credit but not receiving it. Here are some examples, the first one, like many of our cases, interacting with gender:

123

B. Bozeman, J. Youtie

I have wasted a great deal of time and energy teaching techniques that I have developed to people who visit my lab, only to have them fail to give me coauthorship. Worse, they often give credit to a male colleague who doesn’t actually do any of the work. In fact, he has no clue how to do it. I now have explicit conversations about collaborations with all lab visitors before they arrive. An odd aspect of co-authorship in economics is that it is considered standard to get significant help from those in fields that can be of use without offering co-authorship (odd because in other scientific fields this is rare). As an econometrician I am often solving econometric problems for faculty doing applied work, including determining the main approach of the data analysis, without being offered co-authorship. Recently, we did a lot of work with another university, brought in a person from their university as a postdoc, backed up the data with in situ hybridization, but we got no credit for it. The reason in part is because this is practical work. People know we do this work and they contact us and say ‘hey, can you just set up a gene expression for this [research project]. It will be easy for you to do. We just want you to confirm our results by using another expression approach.’’ They don’t know much about this approach so they just assume it is easy to do even though it isn’t. Then they don’t give us any credit. In general, power dynamics seem to have a more prominent role in undeserved inclusion, but in some instances unequal relationships are at the heart of persons being excluded from co-author credit. Here is an example, one that interacts power with gender and has a touch of relocation thrown into the mix. I wrote a proposal during my Ph.D., which was funded, and I supported myself and [another] Ph.D. student (let’s call the student A). I designed the research question, performed experiments, and trained A for more than a year on a topic since no one in their group was familiar with that topic. I then graduated and moved to another university but continued communicating and answering his questions about that work over the phone for another year. They finished the paper after that and didn’t include me as a coauthor despite my numerous contributions. When I asked the faculty member why this is the case, he said, ‘‘because your main expertise is on numerics and the paper was experimental.’’ I’m female in a male dominated field and all these collaborators were male. Here is yet another contributor exclusion problem that seems largely owing to differentials in experience and power, again interacting with relocation: I started a new project in my postdoctoral lab and worked on the project for two years by myself. I left the lab prior to the project’s completion to start my own lab. Before leaving, I trained a graduate student and I left all of the data that I had collected as well as a document outlining the different sections of the manuscript (some sections were fully written) and partial figures with my data included and formatted with my postdoc advisor. I was told by my postdoc advisor that I would be contacted when the final experiments were completed to assist with

123

Trouble in Paradise: Problems in Academic Research…

completing the manuscript. The work was published 1.5 years later and I was not included as a co-author. My data (multiple figures worth) and writing (exactly word for word) were included in the manuscript. Upon seeing the manuscript in press, I contacted my postdoc advisor and inquired about gaining co-authorship and she said that she felt that my contributions were appropriately highlighted in the acknowledgement section. Author Order Disputes In many instances co-author problems relate to issues other than who will and will not be included as a co-author. A particularly common issue is a lack of consultation and a subsequent perceived bad outcome in author order or credit as a corresponding author. Sorting out the problems is complicated by the fact that fields differ in their preferred author positions. In some fields, being the last author is the most prestigious, in many others first author is preferred and in still others the order makes less difference than designation as corresponding author. To a large extent the roots of problems seem to be identical to those affecting author exclusion and, thus, we provide just a few examples. Before tenure I worked with a more senior person. We had their [sic.] student in our lab. The senior person had expertise in a completely different field, so the measurements and analysis were completely guided by myself. The paper was written, but my name was buried in the middle of the list, rather than near the end, which signifies supervisor in my field. When this was brought up with the senior colleague they [sic.] blew up. Initially I backed down, but as promotion loomed, I made an additional request. The senior colleague blew up again, but did make the change. Moral: junior faculty, if you foresee that the majority of the work is occurring in your lab, you need to be proactive and upfront about author order BEFORE paper is written. I have never again collaborated with this faculty member. Here is another author order example in which power and relocation play a role: One of the papers I wrote as part of my thesis research was co-authored by a former student of my advisor who had moved on to another institution. She did not do any of the research or writing herself, but was very familiar with the problem and made many good suggestions during the discussion of this new type of analysis we were attempting. She was very meticulous with the content of the paper and inspired me to keep very high standards. I was clearly the first author, but my advisor asked me if I thought he should be second and the other woman third, because that accurately represented the amount of effort contributed by each of us to the paper, although he clearly felt guilty about it. (He is a well-established professor in the later part of his career and the other co-author was a new, untenured faculty member.) I have always been under the impression that the last author indicates the senior person running the group in which the research was done (and funding the operation), but clearly this is not a universal understanding.

123

B. Bozeman, J. Youtie

Including the Undeserving: Guest Authors, Honorary Authors, and the Idea of Significant Contribution A common concern, as we suggested above, is the ‘‘guest author,’’ the person who makes no significant contribution whatsoever to research but by one means or another manages to be included as a co-author. This is different from we call an ‘‘honorary author,’’ in which case a person does not seek to be included as a co-author but is included either out of respect, out of a desire to associate the work with a prestigious person or for some other reason such as misunderstandings about the relation of authority to authorship. But in the case of the honorary author, unlike the guest author, the co-authorship status not initiated by the recipient. It is also different from the ‘‘ghost author’’ who makes a sizeable contribution to and in many cases exerts influence on the results of the work but is not included as an author. The ghost author could be considered the polar opposite of the guest author because the ghost author purposely does not wish to be listed as an author, even though the ghost author has considerable sway over the paper, while the guest author does wish to be—indeed sometimes insists on being—registered as a coauthor despite making no contribution to the paper. The problem in the case of the guest author is the idea of a ‘‘significant contribution.’’ In some cases our data show that while the respondent obviously felt an individual did not do enough to merit co-authorship, there seems to us be some room for interpretation. In other cases, it seems that promises were made and not kept, but there is no implication of unethical behavior; busy people sometimes do not come through with their expected work in some cases receive credit nonetheless. In still other cases, it seems very clear that individuals use positions of influence or simply cajole others into including them as co-authors even though all parties are aware that the individual has done nothing to warrant inclusion. Sometimes coauthorships are awarded as, essentially, charity. If people are in tenure jeopardy or will soon be on the job market, then colleagues will sometimes stretch the idea of a significant contribution to the breaking point in order to enhance the career prospects of someone they care about. In short, knowing which case falls into which category turns out not to be so easy, especially when each of our cases relies only on one reporter. Thus, our approach below is to start out with what seems to us to be particularly egregious cases and then move to ones that appear to be less clear-cut. The most problematic and usually unethical cases are the ones that we refer to as guest authors: people who insinuate themselves as co-authors against others’ objections, either vocal or unspoken objections. Guest authorship seems clearly to have strong implications for unethical collaborative behavior and for the overall veracity and integrity of the scientific enterprise. Guest Authors: How People Acquire Baseless Co-authorships The following case is a fairly typical instance of a junior person feeling as though a senior faculty member has exploited them: Recently I collaborated with a very senior person on a high-impact project. The senior person suggested that they had ‘‘interest in mentoring a younger,

123

Trouble in Paradise: Problems in Academic Research…

developing faculty’’ and ‘‘help them in their career.’’ I considered this to be a positive sign and genuinely took them at their word. I conducted all of the research, including analysis, mentoring the students and writing the paper. When the student sent the first version of the manuscript to everyone, surprisingly the only edit offered by the senior person was the change in authorship. There were no suggestions on analysis, data presentation, discussion etc. And guess where the senior person had themselves listed? In the pre-edit version authors were: Student A, Student B, Senior Author, Me. Now [after then senior person’s changes] it was Student A, B, Me, Senior Author. [Authors’ note: in this field the most prestigious position is the last one]. And when I objected, I was told that ‘‘this is how it works.’’ I did not settle and unfortunately this very high-impact research will not be published because of the authorship issue. From now on, I am discussing authorship before furthering any potential collaboration – unless I personally know the authors. We note in passing that one common and positive outcome of bad co-authoring experiences is that collaborators learn quickly and often take more care in the future in their choices of co-authors and in making early joint decisions about co-author crediting order. While this is a useful strategy for the future, it can, in the extreme, limit scientific progress and dissemination of results. We come back to this point in a later section focusing on compensatory and preventive strategies in research collaboration. The idea of experienced and very senior people taking advantage is one that is, unfortunately, ubiquitous in our data. This is certainly not due entirely to young, inexperienced people not knowing the ropes. Many of the reports are from very senior people talking about bad formative experiences early in their career. Here is another example, this one from a senior professor recounting a problem from doctoral student days: While working on my doctorate, I was pressured to include a PM&R physician on two papers. He contributed nothing to the work or to the two manuscripts! On the first, he received second author status but I wrote the paper (with much assistance from my mentor, who was deservedly first author) and the physician was third author. On the second paper, I was first author (I had graduated and led the manuscript effort), but I still had to ‘‘carry’’ the doc as a third author. Both cases were unjustified and borderline unethical (he did contribute to the concept). I prefer not to collaborate with physicians unless absolutely necessary. They do not pull their weight; yet want to claim most of the accomplishments! This was one of several cases where individuals in our data set discuss problems working with physicians and medical researchers. Unfortunately, since no physicians or medical researchers were included in our study we see only one side of what is likely a disciplinary dispute or a reflection of pressures associated with the broader medical system. Perhaps a more interesting issue illustrated by the above quote is that the credited physician did actually ‘‘contribute to the concept.’’ As persons who are outsiders looking in it is very difficult to make any determination of what exactly this might mean. Sometimes the concept is

123

B. Bozeman, J. Youtie

everything that is original about research and sometimes contributing to the concept can be a negligible contribution. Even as we acknowledge the difficulties with seeing only one side of disciplinary disputes, it is also worth noting than quite a few complaints were STEM respondents working with medical researchers and physicians. Here is another: I find working with individuals on the medical side a bit frustrating. An NIH proposal can include almost 15 co-authors (some are co-PIs, some are not). When writing a paper that is based only on one task (contributed by perhaps 4-5 individuals at the most), they all insist on being a co-author even if not everyone contributed. It is truly unethical to include someone as a co-author if they have not contributed to the ideas, writing, or analysis, regardless of whether their name was on the proposal or not. This does not happen often in my career (I am in the engineering field), but the 2-3 times that it did happen, it was ALWAYS with someone in the medical side. In the following case we see an instance that is perhaps a true guest author experience but, at the same time, the idea that ‘‘he only commented on the final draft of the paper’’ is open to multiple interpretations. It is not clear from the case below that ‘‘his research scientists’’ made any contribution at all. Finally, as is so often the case, the specter of gender dynamics appears. Here, as in other cases, we have no idea from the evidence provided whether gender really plays a major role or whether there is some degree of attribution error. During my doctoral education, I started to talk to an experimental faculty so that I could perform my experimental ideas (I was working in a purely numerical lab). The faculty member allowed me to use his laboratory to perform experiments, but he didn’t have any other contributions in designing the research question, performing the research, or writing the paper. He only commented on the final draft of the paper. He emphasized so much on the authorship that I ended up adding him and his research scientist as co-author of the paper. In my opinion these undesired experiences [note: respondent provides three similar cases of unwarranted co-author exclusion] were affected by gender difference. I’m female in a male dominated field and all these collaborators were male. Sometimes the claim to a ‘‘significant contribution’’ hangs by the most slender thread: In one case, a senior professor wanted to be listed as a co-author just because he attended a workshop where our research was discussed. Beyond his attendance, there was no tangible contribution to the paper. In some cases the insinuating co-author, the would-be guest author, is a person associated with the funding of research: Several times, I’ve run into situations where a project officer from the agency/company that funded the research has hinted that he/she would like to be a co-author – but had virtually no role in forging the direction of the

123

Trouble in Paradise: Problems in Academic Research…

research, gathering and analyzing the data, or in writing the manuscript. In each case I did not list them as co-author, explaining that I’d be happy to have a collaborative relationship with them that would naturally lead to coauthorship, but that our current relationship warranted acknowledgement, not co-authorship. Hard feelings resulted, but I would not budge on my position. In some cases, undeserving co-author cases relate to the role of resources, such as equipment. Consider the following case. One male colleague in my department with an expensive piece of equipment buys co-authorship on papers by completing analyses for people at no cost. He sets the cost of analysis as being named a co-author. His publication productivity is judged in the same way mine is – by number of publications. Although he does not participate in any aspect of the intellectual enterprise of developing the research questions, designing a research approach, interpreting the data, or writing about the significance of the findings, he gets more publications than anyone else in the department. It is completely unfair, but the Dean doesn’t seem to know how to do anything other than count numbers of papers. As expressed by the respondent this seems a clear-cut case of guest authorship and unethical behavior. However, one could easily envision cases where someone who is an expert only on equipment would make a ‘‘significant contribution’’ especially if the expert ran or manipulated or applied the equipment and if the equipment application was central to the ability to construct original research. Powerful Guests By far the most common factor in guest authorships is power. This is the theme that occurs again and again. Some of the examples above involve the exercise of power but most of them are multidimensional. The cases we present here are ‘‘pure’’ power, those who control resources, status, degrees or careers seemingly taking advantage of others. Here are some examples. We begin with a bald, direct threat: I had collaborated with a colleague on a previous study which had been published about a year ago. We did a follow up study, using the knowledge that was published in that previous study as starting point to design new experiments, but with no further involvement of this colleague. When I prepared the manuscript for submission, that colleague insisted that she and her student be added to the paper, while they had not contributed to any of the work. When I replied that I did not think it was fair, that colleague (who was a senior member of my department) replied to me that ‘‘this isn’t the smartest thing to do for your tenure promotion.’’ I ended up adding both people to the paper. Perhaps more typical, but with similar and in a sense worse outcomes, is the case of a Ph.D. advisor pressuring a student to share authorship: My PhD advisor had no contributions to my publications (8 journal papers and a number of conference papers) other than providing funding and checking for grammatical errors. I think PhD advisors should be more involved.

123

B. Bozeman, J. Youtie

A somewhat unusual take on the exploitation of power for gain focuses on garnering fame from media exposure: I had issues on some occasions with collaborators that because they were senior and tenured, they took all the credit for the research done in the collaboration because it yielded good media coverage and extremely important results. In this occasion, the initial idea of the project, even the conception and most of the experiments of the project came from the junior investigator, but the senior investigator was the one that got the majority of the credit about the research and findings and the junior investigator seemed like he/she was just a helper on the research, when in reality it was completely the opposite. Here is another description of a power dynamic in which student supervision is the medium by which guest authorship is enforced: My negative experiences involve working with colleagues who have no contribution to the paper, but claim authorship solely based on the fact that they are senior and the students working on the projects are jointly supervised. In the interview phase of our study (Bozeman et al., in press) we learned that in some fields supervising a student, discussing and trading ideas is simply viewed as the job of an advisor or mentor and anyone claiming co-author credit for such activity would be viewed as unethical. However, in other fields this is literally the most common form of collaboration and entire careers are built on supervising students and, as a result, claiming co-author status. Is this routine and accepted practice or ethical violation? Does it depend on the discipline or the field?2 Here is another case that involves people with different work norms, a case perhaps related to the fact that the collaborators are ‘‘from another country.’’ Thus, contributorship may be complicated because the behavior and its perceived ethical dimensions vary by field, nation and culture. Still, the result of the case, whatever the difference in norms, is that people who did not work directly related to the research nonetheless receive credit: I have collaborators from another country and I have had difficulty with assigning co-authorship appropriately as my collaborators are happy to put their entire research group as co-authors though not everyone has contributed to the manuscript. My collaborators also submit manuscripts, for which I am not corresponding author, without allowing all authors to view the version submitted and suggest changes. In this case, when I am not corresponding author, I have had difficulty with manuscripts sent in that do not acknowledge the funding sources properly, and thus, I am not able to claim these papers as resulting from NSF funding. I am not eager to work with this group in the future. 2

All our data are from the U.S. However, in career-long discussion with colleagues in other countries, colleagues who are mostly but not exclusively, social scientists, we know there is considerable variance in accepted practice for claiming co-author credit for work with doctoral students. In some cases it is widely accepted that supervision implies co-authorship even if the supervisor contributes not at all to the writing or analysis or even to the major aspects of the core idea. In other cases that is frowned on and may be considered unethical.

123

Trouble in Paradise: Problems in Academic Research…

Power dynamics and co-authorship sometimes work out in ways that pertain less to particular research papers and projects than to the heuristics and processes people dictate to those in subordinate positions. Here is such an example: I had negative experiences in the lab where I was a post-doc. We were set up to compete with one another. Then when one person ‘‘won’’ [had a paper accepted for publication], the others were put on the paper as co-author as a booby prize. I think my mentor thought this was a fair way of dealing with the situation but I found it a very negative training environment. Finally, not all guest authors are aggressive or push, as the data below show: [In one instance] I asked to remove my name from a paper because I didn’t feel I didn’t do enough work. My coauthor took this the wrong way and thought that I didn’t think well enough of the work. I reinstated my name as author.

Two Residual Co-author Crediting Problems: Unwanted Inclusion and Research Suppression While the vast majority of co-authoring problems relate to either awarding underserved credit or neglecting the deserving, these are not the only problematic issues in co-authoring. In this section we explore two less common co-authoring credit issues, (1) including people who may not wish to be included and (2) suppressing publication or presentation of research. Inclusion Problems Sometimes a lack of consultation leads to a problem very different from either exclusion or order disputes. It is not common to be included without one’s consent but it does happen: I had a manuscript with primarily Chinese co-authors that was submitted without my being afforded the opportunity to read the final version and correct the English. I made sure that did NOT happen again. I have a Ph.D. student who I inherited from a professor who was denied tenure. He is strong technically but has some really poor training related to research and publishing ethics. He recently submitted a paper we worked on together without telling me about it until after the fact. The paper had at least 70 % overlap with another paper we recently finished and also had some random co-authors – one of these was on the committee evaluating the papers for inclusion in a very competitive IEEE conference. Oh and he failed to reference the overlapping paper! He also tends to insert random references, which he thinks will help us get published. We pulled the paper and thankfully he will graduate this quarter.

123

B. Bozeman, J. Youtie

Suppression Problems While there is a research literature that evinces a good deal of concern about research, especially research with industry, being suppressed owing to intellectual property and profit motives, we found very little evidence of suppression problems, but we did find some. Here is a brief but quite straightforward account: One negative collaboration occurred because one of the co-PIs (from a different university) did not want to publish the results, which may have been influenced by IP reasons and his startup company. Here is a much more complex instance of suppression: The single worst result came in a collaboration between myself (a tenured professor of chemistry) and a tenured professor of chemical engineering at my institution. The latter individual was seeking new organic compounds for use as organic electronic materials. Preliminary testing of samples that we had in hand indicated that one of our compounds was of interest as an electron transport material in the solid state. However, a relatively large amount of that compound would be required for proper testing (chiefly because of the inefficiency of vacuum deposition techniques). One of my students devoted an entire summer to the synthesis of the required material, and, when it was tested, it was found to be superior to the standard electron transporter (ALQ) for such applications. However, the engineering professor declared that there was no point in pursuing this work any further, because, although our material was significantly better, it would require that they re-optimize their device fabrication methods. Thus, we had gone to great effort to prepare a compound that was in fact a superior electronic material, but the collaboration was terminated (and never published) because it would have been inconvenient for the engineer to change the way the he was doing things! Why had he sought new materials in the first place? The reference to samples brings up an issue about the role of samples in determining contributorship. While the provision of samples in many cases constitutes a definitive contribution, some samples are standard and provided as a convenience, yet the supplier of the sample still pushes for authorship despite having made no other contribution to the research. This situation again has the potential for bordering on guest authorship.

Conclusions: Understanding Co-author Problems A close reading of the cases and anecdotes provided above shows that co-authoring problems have many idiosyncratic elements. Despite their variety, we feel that it is possible to draw some general lessons about the nature and sources of co-authoring problems. Using the data from these anonymous reports, but also our experience in developing critical reviews of the empirical literature on research collaboration (Bozeman and Boardman 2014; Bozeman et al. 2013) we have developed a

123

Trouble in Paradise: Problems in Academic Research…

Fig. 1 Model of credit sharing determinants

preliminary model of credit sharing determinants, presented in Fig. 1. From the discussion below we see below that some elements of the model are, from the standpoint of improvements or reforms, essentially fixed, whereas others are highly amenable to change. Note that the primary dependent variable in the model is credit-sharing outcomes not decisions. This is in recognition of the fact that in some cases there are no explicit decisions about credit sharing, which, not coincidentally, is often a major factor in co-authoring disputes. As we mentioned above there is a great divergence in the processes, norms, and rules that collaborative teams use in coming to conclusions about crediting. Thus, one of the most important elements of our model, and the one most important from the standpoint of quickly improving credit-sharing outcomes, is a category pertaining to research ‘‘team-specific processes.’’ The least profound suggestion we infer from our studies is perhaps the one most efficacious. All of our evidence shows that if collaborative teams have open and participative discussions about crediting, ones where even Junior and inexperienced team members feel safe to express their views, then the likelihood of bad co-authoring outcomes plummets. Our companion research (Youtie and Bozeman 2014) shows that only 40 % of collaborative teams has such explicit discussions about co-authoring credit, and only 31 % of these discussions are before or at the start of commencement of the research, suggesting that most explicit discussions, when they do occur, are in response to problems after these problems have occurred, thereby minimizing the usefulness of such explicit discussion. These statistics hold aside whether they are open, ‘‘safe’’ and participative (Youtie and Bozeman 2014). As our model suggests, team norms do not emerge out of the ether; rather they are strongly influenced by attributes of fields and disciplines and the work culture of particular labs, research centers, academic departments, and ongoing teams bring along their history. What this suggests to some degree is that problems in teams emerge from norms and rules, and just as often lack of norms and rules, but that the likelihood of bad effects is often a joint function of team rules and multidisciplinary structure of a team. It is crystal clear from our survey results and interviews that collaboration practices that are viewed as routine in some fields are viewed as unethical in others. This clash of norms virtually ensures that multidisciplinary

123

B. Bozeman, J. Youtie

teams will face challenges and, moreover, that the problems will be greatly exacerbated if these differences are never discussed explicitly. One of our respondents, after reporting a horrendous multi-university, multidisciplinary collaboration experience, also offered an interpretation of it, one we feel can be generalized to many other such disputes: Often people have strong opinions on how things ‘‘should’’ be done, based on how they were trained, but the rules change depending on field, university, etc. There are often times truly no hard and fast ‘‘rules.’’ …Working across interdisciplinary lines can often lead to these types of misunderstandings. There is no sign that the trend toward interdisciplinary and multidisciplinary research is abating. Thus, we feel that it is safe to predict that in the absence of better team norms, rules and procedures co-authored crediting disputes are likely to only get worse. Another combustible factor increasingly common in research collaborations is concern in relation to commercial objectives and intellectual property. When researchers who have been brought up with open science norms began to team with other researchers focused on developing technology and patents the possibilities for conflict are clear. Our only surprise in this study is that these factors were not more often mentioned. While we do not depict in the model as a separate category, there are important determinants that are related to factors intersecting institutional work culture and disciplinary field effects, issues related to evaluation and reward. One of the most obvious factors affecting collaboration behavior in general, and crediting in particular, is the structure of tenure and promotion systems. This feature is vitally important but also highly variable. In the first place, some nations such as the US, the UK, Canada, Australia, and the Netherlands have ‘‘up or out’’ procedures (also known as ‘‘publish or perish’’ systems) whereby authoring credit determines not just career success but, more dire, whether one will remain unemployed. This presents a very different set of incentives than one finds in, say, Korea, Taiwan, France and Denmark, to name just a few of the countries where one does not necessarily perish if one does not publish. An interesting cross-national study could focus on the impacts of various tenure and promotion requirements on collaboration behavior. Indeed, one interest in comparative work need not focus only on the cross-national. Many US universities vary extensively in their publishing requirements. In the US, those working in the 300 or so ‘‘research-intensive’’ universities have very different work lives than those working in the thousands of other colleges and universities. Those in research-intensive universities work longer hours (Link et al. 2008; Anderson and Slade 2015) and they are expected to provide more extensive knowledge and technology outputs (Crane 1965; Fox 1983; Abramo et al. 2014). Another ‘‘system level’’ factor should be acknowledged, though it is not prevalent in our data. A great many nations, including the UK, Italy, and Spain, among others, have well developed and precisely articulated government-based evaluating schemes for rating universities, academic units and sometimes individuals (Whitley and Glaser 2008; Barker 2007). Even in those cases where future employment is not at stake, there may be highly significant funding allocation decisions contingent on the particulars of national evaluation schemes (Guena and

123

Trouble in Paradise: Problems in Academic Research…

Martin 2003). The reason our data shows no evidence of this is because the US is not yet one of the nations with such central government mandated or imposed evaluation. In some respects, the most disappointing feature of our findings is the enormous role played by power dynamics (and thus its centrality in our conceptual model). It is easy enough to understand the miscommunications and differences in expectations that accrue from clashes in disciplinary backgrounds and differences in work culture and context (see Becher and Trowler 2001). However, many of the power dynamics at play seem to us to give strong evidence of, in some cases, unethical behavior and, in other instances, what we have referred to elsewhere (Bozeman et al., in press), for lack of a better and more decorous name, as the ‘‘jackass factor.’’ What we mean by the jackass factors is that individuals engaged in research will necessarily encounter ruthless and unsavory human beings. Indeed, this occurs in virtually any area of work life and we have no reason to believe that such individuals are any more or less prevalent in academic research. But the solutions, ones that seem to be quite evident to more benign individuals, is to simply not work with a person who is a revealed jackass. Our interview-based research shows that avoidance is the most common approach and seems to be a generally ineffective one. Naturally, the avoidance strategy is not such a good solution when the jackass in question is one’s doctoral or postdoctoral supervisor, or the senior member of a project, or the one who controls most of the resources. Unethical behavior, while it almost assuredly overlaps with the personality flaws that permeate the jackass factor, is not perfectly coincident with it. In some cases there are respondents who report perceived bad behavior by people whom they trusted and considered friends. Why do individuals exploit their friends and also those who are in less powerful positions than themselves? Let us say that we feel on less solid ground when speculating about motive than when focusing on structural causes. Nevertheless, we hazard a supposition that in many instances bad behavior is rooted in desperation. Many of the reports we received show that individuals may be willing to cut corners if they feel their ability to earn tenure is threatened or if they are students underemployed or unemployed. But what of senior investigator desperation, why would this occur? We are not sure that it is particularly common but one of our respondents notes that: [S]ome ‘smart’ senior investigators can run out of ideas and stick with young investigators claiming that they are there to help them through tenure, but in reality it is their way of getting fresh and new ideas to them to keep their brilliant career going. Related, one individual who had spent much of his career in industry said that one of this biggest disappointments coming to academia: In academia, the worst offense that I see is advisors not letting their students off after graduation, i.e. continuing to expect that their [the advisors’] name to show up on their former students’ papers. I loathe this, actually I think it’s an ethical issue.

123

B. Bozeman, J. Youtie

Given the nature of the problems, as reflected in our data and generalized in the model, what is the likelihood of contributorship policies improving co-authoring outcomes? Certainly they have the potential to help, if by other means simply sensitizing people to problems. However, we feel that institutional actors such as journal editors and professional society committees and funding agents have little chance, at least in the absence of better training and socialization, of ushering in major improvements in co-author crediting. For most researchers their first ‘‘training’’ in research collaboration occurs, if it occurs at all, as a by-product of early research experiences. If there are no discussions of standards, choices or norms and such, we should not be surprised to see bad practices perpetuated.

References Abramo, G., D’Angelo, C. A., & Rosati, F. (2014). Career advancement and scientific performance in universities. Scientometrics, 98(2), 891–907. Anderson, D. M., & Slade, C. P. (2015). Managing institutional research advancement: Implications from a university faculty time allocation study. Research in Higher Education, 1–23. doi:10.1007/s11162015-9376-9. Austin, M. A., Hair, M. S., & Fullerton, S. M. (2012). Research guidelines in the era of large-scale collaborations: An analysis of Genome-wide Association Study Consortia. American Journal of Epidemiology, 175(9), 962–969. Barker, K. (2007). The UK Research Assessment Exercise: The evolution of a national research evaluation system. Research Evaluation, 16(1), 3–12. Baron, R. S. (2005). So right it’s wrong: Groupthink and the ubiquitous nature of polarized group decision making. Advances in Experimental Social Psychology, 37, 219–253. Becher, T., & Trowler, P. (2001). Academic tribes and territories: Intellectual enquiry and the culture of disciplines. New York: McGraw-Hill Education. Birnholtz, J. P. (2006). What does it mean to be an author? The intersection of credit, contribution, and collaboration in science. Journal of the American Society for Information Science and Technology, 57(13), 1758–1770. Birukou, A., Wakeling, J. R., Bartolini, C., Casati, F., Marchese, M., Mirylenka, K. et al. (2011). Alternatives to peer review: Novel approaches for research evaluation. Frontiers in computational neuroscience, 5. http://dx.doi.org/10.3389/fncom.2011.00056. Bozeman, B., & Boardman, C. (2014). Research collaboration and team science: A state-of-the-art review and agenda. New York: Springer. Bozeman, B., & Corley, E. (2004). Scientists’ collaboration strategies: implications for scientific and technical human capital. Research Policy, 33(4), 599–616. Bozeman, B., Fay, D., & Slade, C. P. (2013). Research collaboration in universities and academic entrepreneurship: The-state-of-the-art. The Journal of Technology Transfer, 38(1), 1–67. Bozeman, B., & Gaughan, M. (2011). How do men and women differ in research collaborations? An analysis of the collaborative motives and strategies of academic researchers. Research Policy, 40(10), 1393–1402. Bozeman, B., Gaughan, M., Youtie, J., Slade, C. P., & Rimes, H. (2015). Research collaboration experiences, good and bad: Dispatches from the front lines. Science and Public Policy. doi:10.1093/ scipol/scv035. Brass, D. J., Butterfield, K. D., & Skaggs, B. C. (1998). Relationships and unethical behavior: A social network perspective. Academy of Management Review, 23(1), 14–31. Brockner, J., & Wiesenfeld, B. M. (1996). An integrative framework for explaining reactions to decisions: Interactive effects of outcomes and procedures. Psychological Bulletin, 120(2), 189–208. Council of Science Editors. (1999). White paper on promoting integrity in scientific journal publications. http://www.councilscienceeditors.org/editorial_policies/white_paper.cfm Crane, D. (1965). Scientists at major and minor universities: A study of productivity and recognition. American Sociological Review, 30, 699–714.

123

Trouble in Paradise: Problems in Academic Research… Davenport, E., & Cronin, B. (2001). Who dunnit? Metatags and hyperauthorship. Journal of the American Society for Information Science and Technology, 52(9), 770–773. De Stefano, D., Fuccella, V., Vitale, M. P., & Zaccarin, S. (2013). The use of different data sources in the analysis of co-authorship networks and scientific performance. Social Networks, 35(3), 370–381. Devine, E. B., Beney, J., & Bero, L. A. (2005). Equity, accountability, transparency: Implementation of the contributorship concept in a multi-site study. American Journal of Pharmaceutical Education, 69, 455–459. Do¨rner, D., & Schaub, H. (1994). Errors in planning and decision-making and the nature of human information processing. Applied Psychology, 43(4), 433–453. Drenth, J. P. H. (1998). Multiple authorship: The contribution of senior authors. Journal of the American Medical Association, 280(3), 219–221. Esser, J. K. (1998). Alive and well after 25 years: A review of groupthink research. Organizational Behavior and Human Decision Processes, 73(2), 116–141. Fox, M. F. (1983). Publication productivity among scientists: A critical review. Social Studies of Science, 13(2), 285–305. Gaughan, M., & Ponomariov, B. (2008). Faculty publication productivity, collaboration, and grants velocity: Using curricula vitae to compare center-affiliated and unaffiliated scientists. Research Evaluation, 17(2), 103–110. Gino, F., Ayal, S., & Ariely, D. (2009). Contagion and differentiation in unethical behavior the effect of one bad apple on the barrel. Psychological Science, 20(3), 393–398. Green, S., & Bauer, T. (1995). Supervisory mentoring by advisers: Relationships with doctoral student potential, productivity, and commitment. Personnel Psychology, 48(3), 537–542. Haustein, S., & Siebenlist, T. (2011). Applying social bookmarking data to evaluate journal usage. Journal of Informetrics, 5(3), 446–457. Hayter, M., Noyes, J., Perry, L., Pickler, R., Roe, B., & Watson, R. (2013). Who writes, whose rights, and who’s right? Issues in authorship. Journal of Advanced Nursing, 69(12), 2599–2601. Heffner, A. G. (1981). Funded research, multiple authorship, and subauthorship collaboration in four disciplines. Scientometrics, 3, 5–12. International Committee of Medical Journal Editors. (1997). Uniform requirements for manuscripts submitted to biomedical journals. Journal of the American Medical Association, 277(11), 927–934. Janis, I. L. (1982). Groupthink: Psychological studies of policy decisions and fiasces. New York: Houghton Mifflin. Katz, J. S., & Martin, B. R. (1997). What is research collaboration? Research Policy, 26, 1–18. Kornfeld, D. S. (2012). Perspective: Research misconduct: The search for a remedy. Academic Medicine, 87(7), 877–882. Labianca, G., Brass, D. J., & Gray, B. (1998). Social networks and perceptions of intergroup conflict: The role of negative relationships and third parties. Academy of Management Journal, 41(1), 55–67. Lagnado, M. (2003). Increasing the trust in scientific authorship. British Journal of Psychiatry, 183(1), 3–4. Laudel, G. (2002). What do we measure by co-authorships? Research Evaluation, 11(1), 3–15. Lee, S., & Bozeman, B. (2005). The impact of research collaboration on scientific productivity. Social Studies of Science, 35(5), 673–702. Levsky, M. E., Rosin, A., Coon, T. P., Enslow, W. L., & Miller, M. A. (2007). A descriptive analysis of authorship within medical journals, 1995–2005. Southern Medical Journal, 100, 371–375. Link, A. N., Swann, C. A., & Bozeman, B. (2008). A time allocation study of university faculty. Economics of Education Review, 27(4), 363–374. Marusic´, M., Bozikov, J., Katavic´, V., Hren, D., Kljakovic´-Gaspic´, M., & Marusic´, A. (2004). Authorship in a small medical journal: A study of contributorship statements by corresponding authors. Science And Engineering Ethics, 10(3), 493–502. McCauley, C. (1989). The nature of social influence in groupthink: Compliance and internalization. Journal of Personality and Social Psychology, 57(2), 250. McCrary, S. V., Anderson, C. B., Jakovljevic, J., Khan, T., McCullough, L. B., Wray, N. P., & Brody, B. A. (2000). A national survey of policies on disclosure of conflicts of interest in biomedical research. New England Journal of Medicine, 343(22), 1621–1626. McShane, M., Nirenburg, S., & Jarrell, B. (2013). Modeling decision-making biases. Biologically Inspired Cognitive Architectures, 3, 39–50. Melin, G. (2000). Pragmatism and self-organization: Research collaboration on the individual level’’. Research Policy, 29(3), 1140–1670.

123

B. Bozeman, J. Youtie Mowery, D. C., & Sampat, B. N. (2001). Patenting and licensing university inventions: Lessons from the history of the research corporation. Industrial and Corporate Change, 10(2), 317–355. Muscio, A., & Pozzali, A. (2013). The effects of cognitive distance in university-industry collaborations: Some evidence from Italian universities. The Journal of Technology Transfer, 38(4), 486–508. Nabi, R. L. (1999). A cognitive-functional model for the effects of discrete negative emotions on information processing, attitude change, and recall. Communication Theory, 9(3), 292–320. Nedeva, M., Georghiou, L., Halfpenny, P. (1999) Benefactors or beneficiary—The role of industry in the support of university research equipment. Journal of Technology Transfer, 24(2-3), 139–147. Nemeth, C. J. (1986). Differential contributions of majority and minority influence. Psychological Review, 93(1), 23. Olson, G. M., Malone, T. W., & Smith, J. B. (Eds.). (2013). Coordination theory and collaboration technology. New York: Psychology Press. Pichini, S., Pulido, M., & Garcı´a-Algar, O. (2005). Authorship in manuscripts submitted to biomedical journals: An author’s position and its value. Science and Engineering Ethics, 11(2), 173-175. Plemmons, D. (2012). Challenges for research ethics education in the social sciences. Teaching Ethics, 12(2), 145–147. Plous, S. (1993). The psychology of judgment and decision making. New York: Mcgraw-Hill Book Company. Ponomariov, B. L., & Boardman, P. C. (2010). Influencing scientists’ collaboration and productivity patterns through new institutions: University research centers and scientific and technical human capital. Research Policy, 39(5), 613–624. Porter, A. L., & Rafols, I. (2009). Is science becoming more interdisciplinary? Measuring and mapping six research fields over time. Scientometrics, 81(3), 719–745. Resnik, D. B., Peddada, S., & Brunson Jr, W. (2009). Research misconduct policies of scientific journals. Accountability in research, 16(5), 254–267. Rennie, D. (2001). Who did what? Authorship and contribution in 2001. Muscle and Nerve, 24(10), 1274–1277. Rennie, D., & Flanagin, A. (1994). Authorship! authorship!: Guests, ghosts, grafters, and the two-sided coin. JAMA, 271(6), 469–471. Rennie, D., Flanagin, A., & Yank, V. (2000). The contributions of authors. JAMA, 284(1), 89–91. Rennie, D., Yank, V., & Emanuel, L. (1997). When authorship fails. A proposal to make contributors accountable. The Journal of the American Medical Association, 278(7), 579–585. ResearchGate Press Office. (2014). Peer review isn’t working—Introducing open review. Posted March 13, 2014 at https://news.researchgate.net/index.php?/authors/8-ResearchGate-Press-Office Saha, K., & Hurlbut, J. B. (2011). Research ethics: Treat donors as partners in biobank research. Nature, 478(7369), 312–313. Smith, R. (2012). Let’s simply scrap authorship and move to contributorship. BMJ, 344. Sonnewald, DH. (2014). Some points about ASIS&T global initiatives. Bulletin of the American Society for Information Science, 40(5), 32–34. Steen, R. G. (2011). Retractions in the scientific literature: Do authors deliberately commit research fraud? Journal of Medical Ethics, 37(2), 113–117. Tacke, O. (2011). Open science 2.0: How research and education can benefit from open innovation and Web 2.0. In On Collective Intelligence (pp. 37-48). Springer: Heidelberg. Tenenbaum, H., Crosby, J., & Gliner, M. (2001). Mentoring relationships in graduate school. Journal of Vocational Behavior, 59(3), 326–341. Thune, T. (2009). Doctoral students on the university–industry interface: a review of the literature. Higher Education, 58(5), 637–651. Tollefsen, D. P. (2006). Group deliberation, social cohesion, and scientific teamwork: Is there room for dissent? Episteme, 3(1–2), 37–51. Va¨ha¨kangas, K. (2013). Research ethics in the post-genomic era. Environmental and Molecular Mutagenesis, 54(7), 599–610. van Rijnsoever, F. J., & Hessels, L. K. (2011). Factors associated with disciplinary and interdisciplinary research collaboration. Research Policy, 40(3), 463–472. Wallerstein, N., & Duran, B. (2010). Community-based participatory research contributions to intervention research: The intersection of science and practice to improve health equity. American Journal of Public Health, 100(S1), S40–S46. Washburn, J. (2008). Encouraging research collaboration through ethical and fair authorship: A model policy. Ethics and Behavior, 18(1), 44–58.

123

Trouble in Paradise: Problems in Academic Research… Wuchty, S., Jones, B. F., & Uzzi, B. (2007). The increasing dominance of teams in production of knowledge. Science, 316(5827), 1036–1039. Ynalvez, M. A., & Shrum, W. M. (2011). Professional networks, scientific collaboration, and publication productivity in resource-constrained research institutions in a developing country. Research Policy, 40(2), 204–216. Youtie, J., & Bozeman, B. (2014). Social dynamics of research collaboration: Norms, practices, and ethical issues in determining co-authorship rights. Scientometrics, 101(2), 953–962. Zey-Ferrell, M., & Ferrell, O. C. (1982). Role-set configuration and opportunity as predictors of unethical behavior in organizations. Human Relations, 35(7), 587–604. Zhou, W., Zou, Y., Zhu, Y., Fei, S., & Lu, X. (2012, September). Wiki lab: A collaboration-oriented scitentific research platform. In Electronic System-Integration Technology Conference (ESTC), 2012 4th (pp. 411-414). IEEE.

123