A colleague of mine claims that in addition to journals that accept only well-written manuscripts about quality research, there are journals that accept badly written articles about good research, journals that accept well-written articles about bad research, and others still that accept badly written articles about bad research. In other words, getting a manuscript published does not necessarily signify a major scholarly accomplishment. Yet an individual's scholarly productivity is often summed up as a simple publication count that may influence hiring, tenure, and promotion decisions; funding opportunities; and other professional rewards. Journal quality and author order are also taken into consideration in more nuanced evaluations of scholarly productivity (Geelhoed, Phillips, Fischer, Shpungin, & Gong, 2007; Seipel, 2003). Journal quality may be assessed by impact factor, rejection rate, and circulation size by the more assiduous evaluators of curricula vitae; however, there are limitations to these metrics. For example, although social work scholars are expected to show a degree of loyalty to the social work literature (Seipel, 2003), social work journals in general have notoriously low impact factors, and circulation size is a function of topical and methodological specificity, not just quality. Citation counts, which are now easily available online, provide additional useful information for evaluating scholarly contributions and influence, but they, too, suffer from various limitations that reduce their comparability across individuals, such as size of research community in a topical area and publication data (personal communication with M. O. Howard, Frank A. Daniels Distinguished Professor, School of Social Work, University of North Carolina at Chapel Hill, August 26, 2012). In sum, it is difficult to gauge the quality and impact of social work scholars' contributions, even if information on publication outlet and impact is sought out. Most often it is not sought out; simplistic counts are relied on to sum up scholarly activity.
Although the only way to get published in better journals, presumably, is to do superior research and present it well, less honorable options may be used to influence publication counts and author order on manuscripts, such as using status and power to obtain authorship credit or deny it to others (Arthur et al., 2004; Geelhoed et al., 2007; Jones, 1999; Marusi, Bosnjak, & Jeroni, 2011; Street, Rogers, Israel, & Braunack-Mayer, 2010). The quest for long publication lists and sole or first authorship inject zero-sum dynamics into authorship decisions in academic settings, which are characterized by power differentials, competing professional needs, and differences of opinion about appropriate assignment of authorship credit. The socially constructed stakes of publication credits combined with the conditions in which authorship decisions are negotiated make it advisable for social work researchers to establish professionwide ethical standards for authorship assignment. Two goals of such standards are to increase the validity of publication records as indices of scholarly productivity, and thereby enhance their utility for numerous purposes, and to reduce the occurrence of unfairness and coercion in authorship decisions. Developing comprehensive standards applicable to all publishing situations and preventing all abuses of authorship policies are not realistic goals. But social work scholars may be able to reach consensus on a number of acceptable and unacceptable authorship practices and on mechanisms to promote ethical assignment of authorship. This editorial proposes authorship guidelines for social work researchers and challenges those in positions of authority to promote their use.
Encouragement of collaboration by funders and the numerous specialized skills required in large projects have increased the likelihood that multiple individuals make substantial contributions to the same project and manuscripts. The mentor and supervisor roles of faculty with graduate students also lead to multiple-authorship situations (Arthur et al., 2004; Louis, Holdsworth, Anderson, & Campbell, 2008). At the same time, being sole author and being first author are the most highly valued authorship statuses (Geelhoed et al., 2007; Seipel, 2003). Trends toward longer author lists (Howard & Walker, 1996) support the need for guidelines on authorship decisions that increase the validity of publication records as a measure of scholarly contributions.
When individuals with different opinions, competing career objectives, and statuses that range from first-year graduate students, lab technicians, and statisticians to full professors work together on projects and manuscripts, it may be inevitable that disagreements about authorship credit will emerge. Even within individuals there can be opposing pressures, for example, the potential conflict between a faculty member's perceived "need" for publications (that is, to gain tenure) and his or her role as a mentor of graduate students (Netting & Nichols-Casebolt, 1997), or what Louis et al. (2008) referred to as the "sponsorship" of young scientists. The forces affecting authorship assignment practices can lead to situations in which decisions reflect factors other than actual contributions to a manuscript.
A recent publication in Social Science and Medicine (Street et al., 2010) speaks harshly of all kinds of misrepresentation of authorship, labeling them variously "misconduct," "dubious practices," and "outright falsehoods." The authors contend that misrepresentation of authorship undermines the accountability and integrity of academic research, for example, because one or more putative authors whose names lend credibility to a study may not even be aware of how the study was conducted. And publication records that are based on questionable methods of assigning authorship threaten "the meaning and value of all track records in the academy" (Street et al., 2010, p. 1459). Clearly, the application of different rules for authorship results in publication records that are difficult to evaluate and compare. Authorship histories that look the same on paper may represent widely divergent contributions, skills, and future publishing potential.
According to one estimate, one-third of social work scholarship is published in non-social work journals (Seipel, 2003). Therefore, authorship research from a range of disciplines is potentially relevant to a discussion of social work authorship. Cross-disciplinary research suggests that experiences with authorship disagreements and perceptions of misrepresentation of authorship are common in many research settings. A conservative figure that may be most relevant to social work conies from a recent cross-disciplinary systematic review of 123 research articles on 118 studies about authorship (Marusi et al., 2011). Based on a subset of 12 survey studies published in American, British, and international journals that asked respondents about their authorship experiences, the authors found that an average of 23% of respondents reported prior authorship problems. Authorship problems were defined as personal experiences with or observation of authorship disagreements or "misuse of authorship" (Marusi et al., 2011). Other data from studies in the same systematic review suggested that anywhere from 10% to 89% of researchers have observed the practice of giving authorship to undeserving individuals or denying it to deserving individuals. Studies have suggested that giving unearned authorship credit is more prevalent than denying credit (Geelhoed et al., 2007; Louis et al., 2008).
Central to many discussions in the literature on authorship attribution is the relationship between faculty and graduate students (Apgar & Congress, 2005a; Sullivan & Ogloff, 1998). A common concern on the part of scholars who write on authorship issues is the potential for unfair treatment of graduate students in research and writing endeavors, and the need for protective guidelines for students (American Psychological Association [APA] Science Student Council, 2006; Apgar & Congress, 2005b; Arthur et al., 2004; Sullivan & Ogloff, 1998). Power differentials likely also occur among faculty, however, on the basis of tenure status, gender, age, seniority, and other factors. The role of power differentials in assigning authorship credit is illustrated in findings from a study of psychologists (Geelhoed et al., 2007). Students and untenured faculty were more likely than tenured faculty to report that power differentials adversely affected authorship credit in a study they had recently coauthored. Untenured faculty were also more likely to report that "unwarranted authorship had been granted" or that authorship credit had been influenced by % sense of loyalty or obligation" (Geelhoed et al., 2007, p. 110). Tenured faculty in the study perceived themselves as the most powerful in authorship negotiations and, unsurprisingly, were most likely to feel satisfied with the process of determining authorship credit. In contrast to the sense of power and satisfaction enjoyed by tenured faculty, a "fear of negative consequences" (Marusi et al., 2011) was one of the reasons doctoral psychology students did not seek help when they disagreed with authorship decisions on papers to which they contributed. Similarly, in an earlier focus group study, social work doctoral students reported feeling little control of authorship decisions (Netting & Nichols-Casebolt, 1997).
SOCIAL WORKERS AND AUTHORSHIP PROBLEMS
In the extensive review of research on authorship by Marusi et al. (2011), only three of 123 articles on 118 studies were completed by social workers (that is, Apgar & Congress, 2005a, 2005b, Seipel, 2003). Findings of the three studies and language in the NASW (2008) Code of Ethics and the Council on Social Work Education's (CSWE) National...