The Fox in the Hen House: A Critical Examination of Plagiarism Among Members of the Academy of Management

Page created by Edith Hardy
 
CONTINUE READING
姝 Academy of Management Learning & Education, 2012, Vol. 11, No. 1, 101–123. http://dx.doi.org/10.5465/amle.2010.0084

........................................................................................................................................................................

                  The Fox in the Hen House:
                  A Critical Examination of
               Plagiarism Among Members of
                the Academy of Management
                                                                       BENSON HONIG
                                                                      McMaster University

                                                                        AKANKSHA BEDI
                                                                       Bishop’s University

            Research on academic plagiarism has typically focused on students as the perpetrators of
            unethical behaviors, and less attention has been paid to academic researchers as likely
            candidates for such behaviors. We examined 279 papers presented at the International
            Management division of the 2009 Academy of Management conference for the purpose of
            studying plagiarism among academics. Results showed that 25% of our sample had some
            amount of plagiarism, and over 13% exhibited significant plagiarism. This exploratory
            study raises an alarm regarding the inadequate monitoring of norms and professional
            activities associated with Academy of Management members.
........................................................................................................................................................................

ETHICS AND THE ACADEMY OF MANAGEMENT                                                       deter such behaviors (e.g., Kock, 1999; Martin, 1994;
                                                                                           Schminke, 2009; Shahabuddin, 2009; Von Glinow &
Student plagiarism, facilitated by the Internet, is a
                                                                                           Novelli, 1982). To our knowledge, there has been no
pervasive and frustrating problem that appears to
be increasing in recent years (Flynn, 2001; Roberts,                                       empirical research that has either investigated the
2008; Trinchera, 2001). It is not surprising, therefore,                                   issue of plagiarism in social science research or
that a considerable amount of research has been                                            examined some of its predictors. However, there is
conducted to investigate the factors that lead to                                          evidence to suggest that scholars may also plagia-
such behaviors (e.g., Bolin, 2004; Davis, 1992; Gran-                                      rize and claim portions of someone else’s work as
itz & Loewy, 2007; Kisamore, Stone, & Jawahar,                                             their own (e.g., Bedeian, Taylor, & Miller, 2010; End-
2007; McCabe, Butterfield, & Treviño, 2006). Al-                                           ers & Hoover, 2006).
though research on the complex interplay of situ-                                             Our purpose in this exploratory study is to em-
ational and individual variables related to student                                        pirically examine the issue of plagiarism by aca-
plagiarism is important (Kisamore and colleagues,                                          demic researchers and to investigate the institu-
2007), surprisingly little attention has been paid to                                      tional and demographic predictors of such
plagiarism in academic research and publication.                                           behaviors. To date, the empirical literature on pla-
Moreover, much of what has been written about                                              giarism has focused only on the predictors and
the topic is based on anecdotal and speculative                                            effects of plagiarism conducted by students. Un-
evidence and is limited to the discussion of gen-                                          derstanding plagiarism among academics is im-
eral principles of ethical research and strategies to                                      portant because these are the very individuals
                                                                                           who are responsible, through mentoring and
                                                                                           teaching, for developing a new generation of
The authors would like to acknowledge the following individ-                               scholars. Furthermore, they are responsible for dis-
uals, whose engaging conversations and active encouragement                                seminating novel intellectual contributions and for
at the Academy of Management (AOM) meeting in Chicago,
2009, inspired this research project: Israel Drori; R. Edward
                                                                                           upholding the highest ethical standards in society.
Freeman; Yuval Kalish; Joseph Lampel; Kathleen Montgomery;                                 From a theoretical standpoint, our study contrib-
Amalia Oliver.                                                                             utes to existing literature on plagiarism that sel-
                                                                                     101
Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s
express written permission. Users may print, download, or email articles for individual use only.
102                                Academy of Management Learning & Education                               March

dom addresses the potential for academics engag-          student plagiarism and plagiarism in academic
ing in unethical behaviors, and that offers little        research. We then describe the research and the-
guidance regarding the nature and causes of the           ory used to support the study hypotheses. We
diffusion of plagiarism. From a practical stand-          conclude with a discussion of the study’s find-
point, an empirical examination of the prevalence         ings, their implications for the Academy (includ-
of plagiarism in academic research may have im-           ing recommended policies for the Academy of
plications that go beyond the hypotheses tested in        Management), as well as suggestions for future
this study and influence the way management re-           research.
search is conducted and reviewed.
  Using evidence from past theory and research,
we argue that plagiarism by academic scholars             BACKGROUND
occurs due to the growing pressure to publish re-
                                                          Student Plagiarism—A Review
search, as well as increasing pressure to publish
in high-impact top-tier journals (DiMaggio & Pow-         The issue of student plagiarism has generated a
ell, 1983; Kock, 1999; Martin, 1994; Von Glinow &         great deal of media and research attention and is
Novelli, 1982). Other factors, such as top social sci-    an increasingly studied phenomenon in higher ed-
ence journals’ demand for complex research, in-           ucation research. With respect to the nature of pla-
creasing competition, and availability of impact          giarism, a variety of definitions have been offered
factor and citation count softwares such as Publish       (e.g., Cottrell, 2003; Fialkoff, 1993; Hannabuss, 2001),
or Perish, have further exacerbated the pressure to       which, although distinct in certain ways, generally
publish (Bedeian et al., 2010; Harzing, 2010; Lampel      converge on the notion that plagiarism involves
& Shapira, 1995). However, despite these pressures,       intentionally and without authorization presenting
existing mechanisms (i.e., the use of plagiarism          someone else’s ideas or words, as one’s own work.
detection software such as Turnitin or Ithenticate)       This ranges from minor instances, such as sloppy
to monitor academic research are rarely used in           paraphrasing, to major incidents, such as inten-
the field of social science, suggesting greater in-       tional word-for-word copying of someone else’s
centives for, and perhaps a higher likelihood of          work without proper acknowledgment (Hawley,
“getting away” with, unethical behavior. We aim           1984). Irrespective of the type of plagiarism, that
to explore the issue of academic plagiarism, why it       plagiarism and cheating among students is wide-
occurs, and how to prevent it. From a practical           spread and on the rise is noteworthy (e.g., Bennett,
standpoint, discussing these questions may help           2005; Park, 2003; Whitley, 1998). For example, a lon-
raise awareness about the issue, aid in the train-        gitudinal study of 474 undergraduate students by
ing of ethically responsible researchers, and influ-      Diekhoff (1996) reported a significant increase in
ence scholars toward more ethical behavior.               overall cheating levels from 54.1% in 1984 to 61.2%
                                                          in 1994.
                                                             Student plagiarism is by no means a new phe-
[P]lagiarism by academic scholars occurs                  nomenon, nor is it relegated to marginalized or
due to the growing pressure to publish                    periphery scholarship. For instance, examinations
research, as well as increasing pressure                  of Dr. Martin Luther King’s graduate work unveiled
                                                          portions of his dissertation that were directly pla-
to publish in high-impact top-tier                        giarized without citation (King, Jr., Papers Project,
journals.                                                 1991). Had this come to light when he was a re-
                                                          cently minted PhD, there would have been strong
                                                          grounds to revoke his degree—in fact, a committee
  To summarize, we aim to contribute to the liter-        met to consider this very issue after his death,
ature on plagiarism by (1) exploring the issue of         deciding to attach a letter to his dissertation indi-
academic plagiarism and outlining the reasons             cating serious improprieties (Radin, 1991).
that warrant its study; (2) exploring the role of            The issue of student plagiarism also received an
demographic and institutional factors, namely             unprecedented surge of media attention when Pro-
gender, academic status, education, and country,          fessor Bloomfield, a physics professor at the Uni-
on the percentage of plagiarism; and (3) discussing       versity of Virginia, designed a computer program
practical implications for the Academy and edito-         to analyze past submitted papers for repetition and
rial boards in general, by examining papers pre-          plagiarized content (Schemo, 2001). The examina-
sented in one division of the 2009 annual meeting         tion revealed 158 students who had plagiarized
of the Academy of Management.                             their work. As a result, 45 students were expelled
  We begin by briefly reviewing the literature on         from the university, three graduates had their de-
2012                                            Honig and Bedi                                             103

grees revoked (Trex, 2009), and a new industry was       Lack of English proficiency in an Anglo environ-
launched specifically to examine the authenticity        ment, as well as a different cultural understanding
and originality of student papers (Braumoeller &         regarding what constitutes plagiarism and the
Gaines, 2001). These new systems, along with the         sharing of knowledge, have been associated with
increasing attention directed toward intellectual        higher levels of academic dishonesty (Carroll,
property and copyright protection, as well as the        2002; Cohen, 2004; Larkham & Manns, 2002;
expansion of electronic media, have resulted in a        Park, 2003).
greater awareness of the implications for and eval-         Research examining the impact of students’ per-
uations of plagiarism (Drinan & Gallant, 2008). Not      sonal characteristics on plagiarism suggests that
surprising, therefore, is that a large volume of re-     some individuals have a greater propensity to in-
search has been conducted to investigate why stu-        dulge in plagiarist activities. One such individual
dents plagiarize and what can be done to discour-        difference factor is the propensity to rationalize
age such behavior (e.g., Coleman & Mahaffey,             dishonest behaviors. Research on plagiarism and
2000; Crown & Spiller, 1998; Kisamore et al., 2007;      deviant behaviors in general has argued the rele-
McCabe & Treviño, 1997; Howard & Davies, 2009).          vance of rationalizing or neutralizing attitudes in
   With respect to the causes of plagiarism, the         shaping reasoning where individuals believe they
relative importance of demographic, individual,          are justified in behaving immorally. Research on
and situational predictors of student plagiarism         student plagiarism has shown a strong positive
has been examined and demonstrated in a number           association between cheating and neutralizing at-
of studies (e.g., Bennett, 2005; Bonjean & McGee,        titudes (Daniel, 1994; Haines, 1986; Jordan, 2001). A
1965; Howard, 2002; McCabe & Treviño, 1993; Mc-          recent study by Rettinger and Kramer (2009), for
Cabe & Treviño, 1997; McCabe, Treviño, & Butter-         instance, surveyed 154 undergraduate students
field, 1999; Scanlon & Neumann, 2002). Although          and found that students with neutralizing attitudes
there is inconsistency in the literature regarding       were more likely to cheat and plagiarize. More-
the association of demographic factors with pla-         over, direct knowledge of others’ cheating or see-
giarism, males and younger students are gener-           ing others cheat had a stronger effect on those high
ally found to have engaged in higher levels of           in neutralization compared to those low in neutral-
plagiarism than females or older students (Gra-          izing attitudes. Another individual difference vari-
ham, Monday, O’Brien, & Steffen, 1994; Lyer & East-      able found to predict plagiarism is attitude toward
man, 2006; McCabe & Treviño, 1997; Newstead,             plagiarism. In his meta-analytic review of factors
Franklyn-Stokes, & Armstead, 1996; Straw, 2002).         predicting cheating, Whitley (1998) found that stu-
One theoretical rationale for the gender difference      dents with favorable attitudes toward cheating
in plagiarism has been provided by sex-role so-          were more likely to cheat than students with unfa-
cialization theory, which argues that women are          vorable attitudes. Some other personality factors
more socialized to obey rules and regulations and        found to be positively associated with plagiarism
are, therefore, less likely to engage in dishonest       are aggressive (Type A) personality (Buckley, Wi-
behaviors (e.g., Ward & Beck, 1990; Whitley, Nel-        ese, & Harvey, 1998); external locus of control
son, & Jones, 1999). Another demographic variable        (Crown & Spiller, 1998); low self-efficacy (Murdock,
that has been associated with plagiarism is gen-         Hale, & Weber, 2001); low self-esteem (Lyer & East-
eral cognitive ability. Results indicate that stu-       man, 2006); and lower levels of school identifica-
dents with lower GPA scores are more likely to           tion (Finn & Frone, 2004).
engage in plagiarism than those with higher GPAs            Plagiarism behaviors are also shaped by the
(e.g., Diekhoff, 1996; McCabe & Treviño, 1997;           context or the situation faced by the student (Mc-
Straw, 2002), although various factors beyond an         Cabe, 1993). One factor that has been positively
individual’s ability may also be relevant. One rea-      associated with plagiarism is students’ percep-
son why students with low GPAs may plagiarize            tions of peer behavior. Using the theoretical per-
more is because they have a higher incentive to          spective offered by social learning theory (Ban-
cheat in order to raise their grades than students       dura, 1986), McCabe and colleagues (McCabe, 1993;
with higher GPAs (Leming, 1980). Indeed, the de-         McCabe & colleagues, 2006; McCabe, Treviño, &
sire to get good grades has been reported as one of      Butterfield, 2002) found that students who wit-
the primary motives to cheat (Bjorklund & Wenes-         nessed successful cheating by their peers were
tam, 1999; McCabe, 2001; Rettinger & Jordan, 2005).      more likely to engage in cheating. A second factor
Finally, another important factor that has been          is the easy access to other’s work that the Internet
associated with plagiarism is an individual’s cul-       offers (Park, 2003). A recent study by Selwyn (2008)
tural and linguistic background (see Hollinger,          found that 69.1% of 1,222 undergraduate students
1965 for a cross-cultural view of student cheating).     had engaged in some form of on-line plagiarism
104                                Academy of Management Learning & Education                             March

during the past 12 months. In another study, Mc-          ies have described specific instances of research
Cabe (2005) surveyed over 80,000 undergraduate            misconduct and have offered suggestions to detect
and graduate students in the United States and            and prevent such behaviors (e.g., Clarke, 2006;
Canada and found that roughly 74% of undergrad-           Enders & Hoover, 2006; Errami & Garner, 2008; Kock,
uate students and 49% of graduate students para-          1999; Shahabuddin, 2009; Yank & Barnes, 2003). Our
phrased or copied a few sentences from a written          purpose here is to address this gap in the literature
or an electronic source without proper acknowl-           and explore the prevalence as well as the predic-
edgment. The problem is further compounded by             tors of plagiarism in academic research. As we
evidence that suggests some students lack a clear         explain below, plagiarism by academic scholars is
understanding of what constitutes on-line plagia-         likely and may be affected by a variety of demo-
rism. For instance, some consider cutting and past-       graphic and institutional factors.
ing from the Internet a good research practice
rather than an act of plagiarism (Poole, 2004;
                                                          Plagiarism in Academic Research
Straw, 2002). Research also indicates a discrep-
ancy between faculty and student perceptions of           Only a few leading social science academic jour-
Internet plagiarism (McCabe, 2005). McCabe (2005),        nals have acted decisively to curb plagiarism. For
for instance, found that only 57% of undergraduate        example, Enders and Hoover (2004), in their survey
students and 68% of graduate students considered          of 127 editors of leading economics journals, found
paraphrasing or copying a few sentences from the          that only 19% had a formal plagiarism policy in
Internet without proper acknowledgment a serious          place. This is surprising given the evidence that
offense. In contrast, when the same behavior was          plagiarism in academic research is widespread.
presented to the faculty, 82% reported it as serious.     For instance, Bedeian et al. (2010) surveyed 438
In another study, Scanlon and Neumann (2002) sur-         faculty members from 104 business schools and
veyed 698 students and reported that 3% of stu-           found that over 70% of the faculty members were
dents felt that faculty members did not view pur-         aware of colleagues who engaged in plagiarism.
chasing papers from on-line paper mills as wrong.         In another study, Enders and Hoover (2006) sur-
Interestingly, in one instance where a student was        veyed 1,208 economists and found that 24.4% of
caught plagiarizing, the student went on to be-           respondents identified themselves as victims of
come an on-line paper mill entrepreneur (Mannix,          plagiarism—although the percentage might have
2010). Other factors that have been positively            been inflated due to self-reporting bias. Evidence
linked to plagiarism are faculty tolerance of pla-        from other fields suggests that plagiarism is not
giarism (McCabe, 1993); fraternity or sorority mem-       only widespread, but often goes undetected. For
bership (McCabe & Treviño, 1997); difficulty of the       example, a year after implementing the plagiariz-
test and decreased surveillance (Whitley, 1998). In       ing screening process for its new submissions, the
contrast, a factor increasingly linked with lower         editorial board of the British Journal of Anesthesia,
levels of student cheating is the presence of insti-      a high-impact medical journal, reported rejecting
tutional honor codes or institutional policies that       4% of submissions on that basis (Yentis, 2010).
require students to take a pledge and maintain an            In a recent editorial in the Academy of Manage-
environment of academic integrity (Bowers, 1964;          ment Review, Schminke (2009) discussed both the
McCabe, 1993, 1997). McCabe (1993, 1997), for in-         considerable temptation to engage in ethical vio-
stance, found that students attending academic            lations at the Academy, and the rarity with which
institutions with an honor code system were not           the audits, either formal or informal, are pursued.
only less likely to cheat, but also less likely to        Reporting on his informal survey, approximately
rationalize or justify cheating behaviors, and more       half the editors he queried had no difficulties re-
likely to discuss the importance of morality and          counting ethical violations that contravened the
compliance with standards of academic integrity.          clearly stated policy formulated in the Academy’s
   The extensive literature on student plagiarism         code of ethics. Most of the cases recounted in his
indicates not only the widespread nature of the           essay reflected violations either of submission
problem, but also a growing awareness of its grav-        (e.g., authors submitted to more than one journal at
ity among faculty and academics alike. And while          a time) or violations of originality, referring to pa-
student plagiarism is well-discussed and policies         pers “conspicuously similar to previously rejected
are in place to limit and control unethical behav-        manuscripts or to papers already published in
ior, what appears to be missing is an empirical           other journals” (Schminke, 2009: 587). Notably ab-
examination of the extent to which plagiarism pre-        sent were instances of data fabrication or exam-
vails in academic research. This paucity of re-           ples of plagiarism by one author of another. Unfor-
search is rather surprising given that several stud-      tunately, given the current norms of the profession,
2012                                              Honig and Bedi                                            105

data fabrication is particularly difficult to ascer-          An example of a nascent violator, on the other
tain, as we do not require the distribution of raw         hand, is the case of a PhD student from Greece who
data, encourage the retesting of similar studies           submitted a plagiarized paper to an academic con-
(how many top journals would consider publishing           ference. As a result of his attempt to distribute
a replication study?), or make any significant at-         plagiarized papers at conferences (in this case,
tempts to independently verify the integrity of the        Euro-Par, a Computer Science Conference), a letter
author(s)’ source or quality of data. In short, despite    was distributed warning other potential conference
the existence of unethical conduct in academic             organizers of his proclivities (Anonymous, 1995).
research (e.g., Yentis, 2010), our monitoring sys-            Although many in our profession appear to be
tems to control such behaviors are either nonexis-         suspicious of students cutting corners in an effort
tent, insufficient, or infrequently implemented.           to marginally improve their grades, we seem to
   Schminke’s (2009) commentary illustrates both           have full confidence in our colleagues, whose in-
the range and the variability of plagiaristic activ-       centives to skirt rules and policies are “limited to
ities. For example, he distinguishes between expe-         less significant issues” such as tenure, reputation,
rienced scholars, who knowingly violate conven-            and six-figure salaries. As editors, we place con-
tion, and new scholars, who either lack the                siderable trust in our submitting authors, believ-
knowledge regarding appropriate processes or               ing that the data they report have been repre-
take shortcuts to secure tenure. These arguably            sented fairly, acquired honestly, and analyzed
reflect different incentives, pressures, and norma-        precisely as depicted. However, there is evidence
tive practices. In the present study, we distinguish       to suggest that this wholesale trust may be mis-
these two groups as either habitual plagiarizers or        placed (Yentis, 2010). Schminke (2009) provided ex-
nascent plagiarizers. Habitual plagiarizers have a         amples of contraventions encountered in the sub-
history of plagiarizing, while nascent plagiarizers        mission process, questioning whether our
are more likely to be either doctoral students or          normative scholarly expectations may be some-
junior colleagues who are new to the profession.           what naive and misplaced. He cited examples of
An example of a habitual plagiarizer is the case of        authors resubmitting rejected manuscripts to the
Dr. Madonna Consantine, a tenured professor at             same journal, and others submitting papers under
Columbia’s Teacher’s College, who was found                second review simultaneously to other journals.
guilty of plagiarizing 36 passages from a junior           These individuals may be described as procedural
colleague and two students over a period of 5 years        deviants or those who, rather than plagiarizing the
(Arenson & Gootman, 2008). Another example of a            work of others, engage in unethical behaviors dur-
habitual plagiarizer is found in the field of man-         ing the publication process. Examples include tak-
agement sciences, where author Dǎnuţ Marcu pub-          ing research shortcuts, falsifying data, or ghost
lished three plagiarized papers in Studia Univer-          writing. For instance, a recent court case uncov-
sitatis Babes-Bolyai Series Informatica during             ered an apparently well-entrenched process of
2002–2003 and subsequently tried to publish an-            ghost writing conducted by pharmaceutical com-
other “lifted” piece in the Quarterly Journal of the       panies for established academics (Wilson &
Operations Research (4OR; Bouyssou, Martello, &            Singer, 2009). In many cases, renowned academics
Plastria, 2009). One of Marcu’s victims wrote:             were taking credit and even payment for allowing
                                                           their names to be used in a peer-reviewed publi-
   A very odd thing has happened. A fellow by              cation, reflecting a study they had not participated
   the name of Dănuţ Marcu has plagiarized my            in, and a paper they had not authored. The practice
   paper in its entirety! The first two pages of his       remained a largely undisclosed secret until litiga-
   paper. . . are basically just a rewording of my         tion brought it to light, effectively opening a Pan-
   paper, down to the details of the proofs. Ap-           dora’s box regarding this conduct. Further evi-
   parently this is not the first time this has            dence of ethical violation in medical research
   happened— he has been plagiarizing papers               emerged in the study by Long et al. (2009) that
   for years and passing them off as his own.              found 9,120 highly similar citations between pub-
   Unfortunately, many of his papers fool both             lished works in MEDLINE. A full text analysis of
   the referees and the journals (Bouyssou et al.,         these citations revealed 212 articles with signs of
   2009: 12).                                              duplication with an average similarity rate of
                                                           86.2% between the duplicated and the original pa-
   Marcu was subsequently “outed” and banned               per. Moreover, only 22.2% of these 212 articles ref-
from publishing his work in the above journals.            erenced the original article, and approximately
The plagiarized rejected piece from 4OR was later          42% contained evidence of data fabrication, incor-
published in another journal.                              rect calculation, and manipulated diagrams. Ob-
106                                Academy of Management Learning & Education                           March

viously, these practices, no matter how rampant or        reported student cheating to the appropriate
normative, contradict editorial guidelines for peer-      authorities.
reviewed journals. Thus, despite the ostensible
rigor of blind peer review, opportunities exist in
                                                          Isomorphism of the Peer Review Process
the peer review system for considerable manipu-
lation and ethical violation.                             Although the concept of “publish or perish” has
                                                          been synonymous with academic life in contempo-
                                                          rary times, this has certainly not always been the
                                                          case. The history of the peer review process is
[O]pportunities exist in the peer review                  surprisingly absent from academic discourse, de-
system for considerable manipulation                      spite its obvious preeminence and its implications
and ethical violation.                                    for prestige, notoriety, and success. Academic pub-
                                                          lication evolved from journalism, when early
                                                          newspapers relied upon a single editor as sole
                                                          adjudicator. Medical journals, for example, main-
   Estimating the severity of plagiarism in the           tained a single editor as a gatekeeper well into the
publication process warrants an ethical yard-             19th century, as did other academic publications in
stick, if only to assess and implement appropri-          the United States and United Kingdom. In France,
ate measures and censures when questionable               senior editors of academic journals thought of
behaviors are identified. Bartlett and Smallwood          themselves as journalists well into the 20th century
(2004) provide a 10-step hierarchical list of pla-        (Burnham, 1990). Many academic journals emerged
giarist offenses, with the top five consisting of (1)     primarily to broadcast the success of a particular
copying the entire work, a substantial part, para-        research institute and were typically published
graphs, sentences, or clauses; (2) copying highly         and edited by a single editor who, as director,
original ideas; (3) paraphrasing segments of sub-         considered himself an expert in all areas related to
stantial size without new contributions; (4) para-        the journal’s topic. As knowledge specialization
phrasing segments of moderate size without new            increased, various editors relinquished some of
contributions; and (5) verbatim or nonverbatim            their editorial control and sought external review
copying of unremarkable segments of small size            advice (Burnham, 1990). Thus, it was only in the
(e.g., clauses, phrases, expressions, and neolo-          1940s, with the combination of increasing submis-
gisms). As with all rules, the above offenses and         sions and the specialization of knowledge, that the
the probability of violation are dependent on             systematic blind review process we are familiar
both contextual conditions, such as the degree of         with today first emerged (Burnham, 1990).
transparency involved, and the characteristics of            As the model for tenure diffused throughout
the rule, including enforceability and procedural         North America, so did the accompanying pace of
factors (Lehman & Ramanujam, 2009). Also of in-           peer-reviewed academic journals, to accommodate
terest is that there seems to be disagreement on          the growing needs of junior faculty to demonstrate
the conceptualization of plagiarism and the ap-           productivity to tenure-and-promotion committees.
propriate penalties attached to such behaviors.           Academic tenure, originating in 12th century Eu-
For instance, Enders and Hoover (2006) found that         rope, disseminated through North America by 1915,
approximately 80% of the 1,208 respondents held           both as a consequence of the influence of German
the view that unattributed sentences were either          institutions (White, 2000), and in response to sev-
likely or definite examples of plagiarism. How-           eral faculty terminations at Stanford University
ever, 2.8% believed it was “not at all” plagiarism,       (Ludlum, 1950). Peer-reviewed scholarship subse-
and 16.6% thought it was “not likely.” Recommen-          quently became the primary duty of faculty (Ad-
dations regarding penalties were surprisingly             ams, 2006). While peer-reviewed publication con-
restrained: 74% would likely or definitely notify         tinues to be central for promotion and tenure, some
the plagiarist’s department chair, dean, or pro-          universities limit the absolute number of articles
vost, 72% would place a ban on future journal             submitted for promotional review (Bickel, 1991),
submissions, but slightly less than half would            thus increasing the pressure to publish in high-
make the plagiarism a matter of public notice             impact journals. Peer-reviewed scientific articles
(2006: 96). This reluctance to publicly report acts       continue to be the means by which academic hon-
of plagiarism is not unique to plagiarism in ac-          ors and promotion are distributed (Hargens, 1988).
ademic research, and is also discussed in the             For example, citations in peer-reviewed journals
literature on student plagiarism where McCabe             have been explicitly linked to increased income,
(1993) found that only 40% of 789 faculty members         which may be considered a proxy for reputation
2012                                             Honig and Bedi                                            107

and prestige (Diamond, 1986). Thus, our increased         igins of institutional theory are embedded in at-
reliance on academic journal rankings to assess           tempts by sociologists to understand the diffusion
individual potential in creating and publishing           of mass education, compulsory state education,
new knowledge is quite evident. A survey of 252           and tertiary education (Meyer & Rowan, 1977;
management department chairs across the United            Meyer, Hannan, Rubinson, & Thomas, 1979). In-
States indicated that approximately 14% of institu-       stitutional theorists were struck by the apparent
tions used a formal list of journals to make person-      isomorphism of educational processes, including
nel decisions such as promotion and tenure (Van           subjects, curriculum, instructional guidelines,
Fleet, McWilliams, & Siegel, 2000). Moreover, the         classrooms, enrollments, school design, peda-
faculty’s “intellectual capital” or the number of the     gogy, and a host of associated educational de-
faculty’s publications in top-tier journals also has      signs. Such expansion occurred irrespective of
implications for other indicators, such as business       controls for variation of the nation state (e.g.,
school rankings, prestige, and access to grants and       urbanization, energy consumption, political re-
other resources (Beamish, 2000; Miller, Glick, &          gime). As stated by Ramirez and Boli (1987: 172),
Cardinal, 2005).                                          “What all these disparate bodies of evidence un-
                                                          derscore is the universality and uniformity of edu-
                                                          cational development in recent decades . . . Edu-
                                                          cation is institutionalized at the world level and
The Institutionalization of Peer Review and
                                                          acts as a social imperative for nation states inte-
Its Consequences
                                                          grated within this institutional environment.” Us-
Higher education in management continues to fol-          ing this perspective, we may observe that business
low a pattern largely set in the United States over       education, including the MBA, taught with similar
a century ago. American journals, such as the pub-        subdisciplines (e.g., finance, accounting, OB, oper-
lications of the Academy of Management, continue          ations, strategy, etc.) has diffused through North
to play a central role in tenure, advancement, and        America, across Europe, and extending to every
university rankings. The United States has contin-        continent. As management education expanded,
ued to lead and even dominate in the field of man-        joint ventures were established between the
agement for more than a century, establishing the         United States and international universities. As a
AACSB accreditation process in 1916 and promot-           result, United States professors developed interna-
ing the growth of named schools and research              tional affiliations, which served to tacitly carry
chairs. Although the process and achievement of           forward existing models of research and publica-
academic rank and tenure according to publica-            tion standards (Bandelj, 1989). To quote one re-
tion are primarily North American inventions,             search study in Eastern Europe, “Aid from interna-
there have been considerable isomorphic trends            tional organizations, the activities of professional
worldwide. For example, the research assessment           associations, and mimicking peer behavior have
exercise (RAE) in the United Kingdom that first           all helped establish a management school as a
began in 1986 has become a systematic 5-year re-          legitimate organizational form in post-socialism”
view of every public higher education facility in         (Bandelj, 1989: 13).
the United Kingdom (it will be called the research           Institutional theory is helpful in understanding
excellence framework in 2014). Universities are           the process by which faculty publication stan-
rated according to the prestige and frequency of          dards, tenure, promotion, and peer review pro-
their faculty publications and are directly re-           cesses, including those related to management ed-
warded through the provision of financial re-             ucation, diffused throughout the world. It helps
sources. Australian universities are presently un-        predict and explain the isomorphic pressures that
dergoing a similar comparative assessment                 result in the diffusion of habitual and nascent pla-
exercise (Gallagher, 2010).                               giarizers. For example, virtually overnight, univer-
   Examining how higher education models have             sities that were unfamiliar with management stud-
globally disseminated is one important element in         ies, such as those in Eastern Europe, suddenly
understanding where, when, and why individuals            found themselves attempting to conform to stan-
choose to be habitual or nascent plagiarizers. Ex-        dards and methods with which they had little or no
isting theories may assist in predicting where and        experience. Institutional theorists have very well-
when the pressures to plagiarize may be greatest.         conceptualized and empirically tested models by
Neo-institutional theory examines the social pro-         which institutional norms are disseminated. Indi-
cesses by which structures, policies, and programs        viduals engage in normative behavior either
are developed and subsequently acquire a “taken-          through coercion (they are forced or cajoled into
for-granted” status (Meyer & Rowan, 1977). The or-        conformity), through mimetic means (they are at-
108                               Academy of Management Learning & Education                            March

tracted by what appears to be a successful model),       larly important for individuals on the periphery
or due to normative expectations (they comply to         who may not otherwise have good access to
appear rational, sensible, modern, or legitimate;        world-class scholarship or opportunities to col-
DiMaggio & Powell, 1983). Regarding the evolution        laborate and network with other members of a
of business schools, coercion took place through         particular scholarly field. We maintain that an
international donors, such as the European Union,        emerging untenured scholar or doctoral student
and the United States Agency for International De-       has a higher incentive to plagiarize than a senior
velopment (USAID); normative isomorphism came            professor or someone not at all in the professor-
through the extension of national and regional pro-      ate (Kock, 1999). This is not to imply that only
fessional organizations, including accreditation         doctoral students and junior faculty are likely to
bodies; and mimetic isomorphism took place               engage in these activities; it is for this reason
through the alliances and cooperative agreements         that we distinguish between nascent plagiariz-
developed with U.S. business schools (Ban-               ers and habitual plagiarizers. Further, ceteris pa-
delj, 1989).                                             ribus, those who have spent a greater amount of
  Regarding the pressures to publish (and possibly       time in the Academy are more likely to be insti-
plagiarize), U.S. reward structures are very clearly     tutionalized into its norms and practices and
institutionalized and biased strongly toward re-         thus less likely to plagiarize. Of course, they may
search and peer-reviewed publication (Fairweather,       also become cavalier and complacent, yielding a
1993). Because management education began in             subsector of habitual plagiarizers, as demon-
the United States and only later spread to Europe,       strated by the examples cited here. However,
Asia, Latin America, and Africa, we would expect         expectancy valence theory argues that expecta-
the normative forces to be strongest and the will-       tions and valences together determine a person’s
ingness to deviate from the norm weakest in the          motivation to undertake a particular behavior
United States, followed by those countries that          (e.g., to plagiarize/not plagiarize). Plagiarism
were early adapters of the management education          may be thus affected by expectations of publica-
model. Given the characteristics of institutional-       tion, the value attached to the publication, and
ization, we would expect developed countries of          the need for publication to obtain a critical net-
the western hemisphere, such as those in North           work or position. Obtaining tenure is clearly an
America and Europe, along with Australia and             important threshold, as it allows the individual
New Zealand, to demonstrate the most universal           to maintain professional status and increases
and well-established compliance to professional          individual motivation. Moreover, expectations
rules and norms. In contrast, countries that have        may also be related to beliefs about the pressure
only recently adopted the institutions of peer re-       to publish, amount of competition for publica-
view (e.g., Eastern Europe, Latin America, Asia,         tions, and success of getting one’s past plagia-
and Africa) will be less likely to comply with pro-      rized work published. Thus, individuals with
fessional norms. Based on the above, we hypothe-         greater incentives (e.g., untenured scholars look-
size:                                                    ing for tenure; doctoral students looking for jobs)
Hypothesis 1: The incidence of plagiarism will be        will be more likely to plagiarize than others. We
                higher in newly institutionalized        therefore hypothesize:
                (noncore) countries than those from      Hypothesis 2: The incidence of plagiarism will be
                more established (core) countries.                      higher for untenured or junior schol-
We recognize that not all individuals face the                          ars than for tenured or senior
same incentives or constraints, so there is a dif-                      scholars.
ferential in willingness to engage in risk-taking           Although scholars from noncore countries are
behavior. While some individuals may have ad-            more likely to plagiarize, we believe that this rela-
opted a laissez-faire approach to scholarship            tionship will be stronger for untenured and new
(habitual plagiarizers), others may calculate the        scholars. The incentives attached to a publication
risks versus the incentives for their specific and       coupled with the noncore country’s failure to en-
highly contextual and conditional situation. For         force or detect plagiarism imply that authors from
example, many doctoral students as well as               these countries will be more likely to plagiarize.
some faculty members are only subsidized to              We specifically argue that untenured and new
attend conferences for which their papers are            scholars’ temptation to plagiarize may be espe-
accepted. Attending conferences is an important          cially strong when they believe that the institu-
aspect of socialization and advances the oppor-          tional norms of ethical research are virtually non-
tunity to develop important professional net-            existent, allowing unethical behaviors such as
works. Networks and conferences are particu-             plagiarism to go undetected. Past research has
2012                                            Honig and Bedi                                            109

indicated the influence of incentives to plagiarize      will be more likely to plagiarize. This relationship
and the risk of getting caught in influencing un-        may be further exacerbated by institutional norms
ethical behaviors (Houston, 1983; Michaels &             and policies that fail to effectively address and
Miethe, 1989; Tittle & Rowe, 1973). Houston (1983),      even implicitly condone unethical behaviors such
for instance, found that the risk of getting caught      as plagiarism. An academic environment where
acted as more of a deterrent to cheating for high-       the norms of ethical research are relatively less
performing students who had little incentive to          developed may heighten a nonnative English au-
cheat than for low-performing students. Accord-          thor’s willingness to plagiarize, thus increasing
ingly, we propose the following hypothesis:              the likelihood of plagiarism. The following is hy-
Hypothesis 3: Author status moderates the rela-          pothesized:
               tionship between an author’s coun-        Hypothesis 5: Education moderates the relation-
               try and incidence of plagiarism. The                    ship between an author’s country
               positive relationship between non-                      and incidence of plagiarism. The
               core country and incidence of pla-                      positive relationship between non-
               giarism will be stronger for unten-                     core country and incidence of pla-
               ured or junior scholars than for                        giarism will be stronger for authors
               tenured or senior scholars.                             receiving their degrees from non-
   Further, rule violation may be differentially                       English speaking countries.
related to the abilities of the deviant. High-
ranked, high-impact publications in manage-
ment are universally English language publica-
                                                         Gender and Plagiarism
tions (Harzing, 2010). Language skills play an
important part in the editorial and research pro-        Another factor that could play an important role in
cess, and articles that are not clearly written or       predicting the incidence of plagiarism is gender.
have grammatical errors are often desk rejected          Kelling, Zerkes, and Myerowitz (1976), in their risk
or otherwise turned down. Good English lan-              as value theory, propose risk taking as a strongly
guage skills are important for rephrasing, sum-          valued masculine tendency that motivates high
marizing, and citing other individuals’ work             levels of risk taking among males. Femininity, on
without resorting to cut and paste plagiarizing.         the other hand, is stereotypically associated with
We maintain that a scholar with excellent Eng-           lower levels of risk taking, as females tend to be
lish language reading and writing skills has less        more concerned about the negative effects of their
of an incentive to plagiarize than someone who           behavior on others (Robbins & Martin, 1993). The
finds the language particularly difficult. Al-           masculinity–femininity distinction is crucial to un-
though we cannot measure the language profi-             derstanding the specific gender roles assigned to
ciency of our sample, we can deduce, to a certain        males and females (social role theory; Eagly, 1987).
extent by proxy, the level of English expertise.         Masculinity, for instance, has been associated
We assume that those who studied in English-             with independence, self-assertiveness, aggres-
speaking countries are more fluent in English            siveness, toughness, and competitiveness (Eagly,
than those who did not, and will thus have a             1987; Gerschick & Miller, 1995; Lee & Owens, 2002).
comparatively easier time communicating and              Femininity, on the other hand, has been associated
writing in English; most would have at least             with more expressive and communal personality
completed their dissertation in English, which is        traits such as compassion, sympathy, nurturance,
arguably a major demonstration of language               sensitivity to others, and high moral standards
proficiency. We hypothesize this relationship as         (Chang, 2006; Franke, Crown, & Spake, 1997; Powell
follows:                                                 & Greenhaus, 2010). These gender roles shape at-
Hypothesis 4: The incidence of plagiarism will be        titudes and behaviors; research has shown that
               higher for scholars that received         individuals engage in behaviors that are consis-
               their highest degree from a non-Eng-      tent with the prevailing gender stereotypes (Adler,
               lish speaking country than for those      Laney, & Packer, 1993; Deaux & LaFrance, 1998).
               receiving their highest degree from       Viewed through the lens of risk as value theory,
               an English speaking country.              this suggests that males will be more likely to
   Furthermore, education may also affect the rela-      engage in risky behaviors. Doing so would be con-
tionship between an author’s country and the inci-       sistent with their gender belief system, which es-
dence of plagiarism. As argued above, authors            tablishes risk taking as an admirable masculine
who are educated in a non-English speaking coun-         trait and enhances their self-esteem by winning
try but are expected to write and publish in English     praise and recognition from others (Clark, Crock-
110                                Academy of Management Learning & Education                             March

ett, & Archer, 1971; Shapira, 1995; Wilson & Daly,        (Hofestede, 1980) will exhibit lower levels of pla-
1985). This assertion is also consistent with the         giarism than male scholars from noncore coun-
social role theory that characterizes men as more         tries, many of which are higher in collectivism
thrill seeking and individualistic, and therefore,        (e.g., China, Korea, Taiwan).
acknowledges aggressive and risk-taking behav-               Furthermore, we argue that although gender dif-
iors as a part of the male gender role (Eagly, 1987;      ferences in plagiarism may exist between different
Whitley et al., 1999; Zuckerman, Kuhlman, Thorn-          cultures, overall plagiarism levels will be higher
quist, & Kiers, 1991). Indeed, a meta-analytic study      for men than for women in most cultures. One
on gender differences in risk taking found that           reason for this is that risk taking is an attribute of
compared to females, males were more likely to            masculine psychology and the cultural differences
engage in a wide variety of risky behaviors such          will be unable to entirely eliminate the risk-taking
as drinking, using drugs, intellectual risk taking,       tendency (Byrnes et al., 1999). Thus, men (who are
and risky experiments, to name a few (Byrnes,             more inclined to take risks than women) will be
Miller, & Schafer, 1999). There is also evidence of       more likely to engage in risk-taking behaviors (Ar-
higher levels of academic cheating (Finn & Frone,         nett, 1992). In his study of 105 Malaysian students
2004; McCabe & Treviño, 1997); student plagiarism         and 96 Australian undergraduate students study-
(Lambert, Ellen, & Taylor, 2003); financial risk tak-     ing in Australia, Egan (2008) found that tendency to
ing (Weber, Blais, & Betz, 2002); and drug use and        plagiarize was higher among Malaysian males
gambling (Zuckerman & Kuhlman, 2000) among                than Malaysian females. Furthermore, Malaysian
males. On the basis of the above theory and re-           students studying at offshore campuses of the uni-
search, we expect parallel gender differences in          versity in Malaysia were more inclined to plagia-
plagiarism among academics such that male                 rize than Malaysian students studying in Austra-
scholars will be more likely to engage in plagia-         lia. Based on the above theory and research we
rism than female scholars:                                propose that:
Hypothesis 6: The incidence of plagiarism will be         Hypothesis 7: Gender moderates the relationship
                greater for males than for female                         between an author’s country and in-
                scholars.                                                 cidence of plagiarism. The positive
   In addition, gender may also moderate the rela-                        relationship between noncore coun-
tionship between country and plagiarism. Specifi-                         try and incidence of plagiarism will
cally, the relationship between noncore country                           be stronger for male scholars than
and plagiarism is likely to be stronger for male                          for female scholars.
scholars from noncore countries than female schol-
ars and males from core countries. Arnett’s (1992)
theory of broad and narrow socialization suggests
that the level of individual risk taking is influ-        METHODS
enced by individual factors, such as level of sen-        Organizational Context:
sation seeking, as well as cultural factors, such as      The Academy of Management
ethical rules, autonomy, and so forth. An individu-
al’s sociocultural background (e.g., ethical norms        The AOM, the largest body of academics dedicated
and beliefs, etc.) emphasizes or deemphasizes the         to the study of business management issues, was
sensation seeker’s inclination to take risks. Gender      founded in 1936 with a formal constitution estab-
has been shown to affect risk-taking behavior,            lished in 1941. Today, the Academy has approxi-
with men preferring greater risks, resulting from         mately 19,630 members from 104 nations in 25
overconfidence (Barber, Barber, & Odean, 2001). It        thematic divisions; it sponsors four prestigious
has also been shown that cultures vary in per-            peer-reviewed academic journals, and hosts an-
ceived risk; it was found, for example, that the          nual meetings attended by over 10,000 people
Chinese were lowest perceived risk averse and             (AOM website). Increasingly, AOM has drawn an
highest perceived risk seeking, as compared with          international audience, as scholars from business
German, Polish, and United States respondents             schools around the world participate in its annual
(Weber & Hsee, 1998). In particular, collectivist cul-    conference and submit to its journals. In 2009, in-
tures were found to have cushioned individuals by         dividuals representing 78 countries participated in
providing more acceptance, allowing for higher            the annual meeting, representing nearly one half
risk taking (Weber & Hsee, 1998). We may thus             of the 8,380 persons in the program. Approximately
anticipate that male scholars from core countries,        one third of the universities that sent more than 30
such as the United States, Australia, and the             participants to the annual meeting were interna-
United Kingdom, that are higher in individualism          tional. With respect to plagiarism and authorship
2012                                             Honig and Bedi                                             111

credit, the Academy (AOM, 2010) has a very specific       that appears in it. This assumption of collective
policy, as follows:                                       responsibility is not new and is popular in medical
                                                          research where coauthors are increasingly re-
   4.2.1. Plagiarism                                      quired to share full responsibility for the content
   4.2.1.1. AOM members explicitly identify,              regardless of their contribution (Nayak & Maniar,
   credit, and reference the author of any data or        2006). The American Physical Society, for instance,
   material taken verbatim from written work,             has a formal policy for coauthors, stating that “co-
   whether that work is published, unpublished,           authors who are accountable for the integrity of
   or electronically available.                           critical data reported in the paper, carry out the
   4.2.1.2. AOM members explicitly cite others’           analysis, write the manuscript, present major find-
   work and ideas, including their own, even if           ings in the conference or provide scientific leader-
   the work or ideas are not quoted verbatim or           ship as bearing responsibility for all of a paper’s
   paraphrased. This standard applies whether             contents” (Dalton, 2002). This position is also sup-
   the previous work is published, unpublished,           ported by two important aforementioned ethical
   or electronically available.                           standards set forth by AOM, that “AOM members
   4.2.2. Authorship Credit                               ensure that authorship and other publication cred-
   4.2.2.1. AOM members ensure that authorship            its are based on the scientific or professional con-
   and other publication credits are based on the         tributions of the individuals involved” (4.2.2.1) and
   scientific or professional contributions of the        that “AOM members take responsibility and credit,
   individuals involved.                                  including authorship credit, only for work they
   4.2.2.2. AOM members take responsibility and           have actually performed or to which they have
   credit, including authorship credit, only for          contributed” (4.2.2.2).
   work they have actually performed or to                   However, it must be noted that despite the spe-
   which they have contributed.                           cific recommendations of the ethical code of con-
   4.2.2.3. AOM members usually list a student            duct of AOM and others, not all scholars subscribe
   as principal author on multiple-authored pub-          to this particular view. They may, for example,
   lications that substantially derive from the           consider their coauthorship as a partnership, with
   student’s’ dissertation or thesis (Academy of          responsibilities delegated according to expertise
   Management).                                           or preference. It may be the case that individuals
                                                          are surprised and even incapable of determining
                                                          the extent of plagiarized content when joining a
Sample
                                                          research team. The notion of professional trust and
Our sample consisted of all empirical as well as          expertise has been well-socialized into our norma-
nonempirical papers presented at the Interna-             tive scholarly view, and the mere idea that a col-
tional Management (IM) division of the 2009 annual        league would “cheat” or “cut corners” may shock
meeting of the Academy of Management. The In-             many in our profession. However, ignorance is no
ternational Management division was selected              excuse for not following requirements, and we
based on its cross-cultural focus and significant         have clear and explicit rules regarding intellectual
representation at the Academy (2,988 members as           property, citation, and collective work and respon-
of August 12, 2010). The International division rep-      sibility. This study reflects our concern that in-
resents papers from different countries and covers        stances of ethical slippage are occurring with con-
a wide variety of areas, such as organizational           siderable frequency. As scholars, we believe that
behavior, international business, strategic man-          the Academy has an obligation to maintain the
agement, and organizational theory. The IM divi-          highest standards regarding intellectual property.
sion, therefore, represents a kind of microcosm of        In short, while we recognize that scholars with nor-
the overall research presented at the Academy and         mative views may be uncomfortable assigning
helps extend the generalizability of our findings to      equal responsibility for violations by all author-
other divisions and areas. In total, 279 papers (rep-     ship team members, taking a perspective of shared
resented by 636 authors) that were available on-          responsibility, which is supported by our ethical
line were selected for use. Each coauthor was con-        guidelines, is the best way to reduce unethical
sidered individually, and we utilized conventional        behavior in our field.
research norms, asserting that all authors share
equal responsibility for their presented work (Dal-
                                                          Coding Scheme
ton, 2002). Thus, regardless of their contribution, if
the coauthors take credit for the presented work,         Two independent raters coded the studies on mul-
then they are also responsible for the plagiarism         tiple dimensions, such as percentage of plagia-
112                                Academy of Management Learning & Education                          March

                                                          ized/developed countries (i.e., North American and
This study reflects our concern that                      European nations, Australia, and New Zealand).
instances of ethical slippage are                         The country from which the authors received their
occurring with considerable frequency.                    highest educational degree was a dummy variable
                                                          coded as 1 ⫽ English-speaking country and 0 ⫽
                                                          otherwise. Finally, gender was a dummy variable
                                                          with males coded as 0 and females, 1. Variables for
rism and paper characteristics. In order to assess        which the above information was unavailable
the reliability of this coding, a random subsample        were left unclassified.
of 57 studies (i.e., 20%) was independently coded by
the second author. Agreement was obtained on 49
of 57 comparisons, yielding a reliability coefficient     Procedure
of 86%. Any initial differences in coding were re-
solved by way of discussion and a more careful            We used the on-line plagiarism detection service,
examination until agreement was reached.                  Turnitin, to check papers for plagiarism. Turnitin
                                                          checks a paper for its originality by comparing it
                                                          with billions of Internet pages (live as well as
Dependent Variable                                        cached), previously submitted student papers, pe-
The dependent variable in this study is the per-          riodicals, journals, and on-line publications. Once
centage of plagiarism, defined as the ratio of pla-       a paper has been uploaded on Turnitin, the web-
giarized words to the total number of words in a          site generates an originality report that indicates
paper. The reference sections were excluded from          the percentage of matches between the submitted
the total word count. A section was considered            paper and the existing database. The website then
plagiarized if the author(s): (a) copied the entire       creates an exact replica of the submitted paper,
section from another paper without proper ac-             except that any text that is copied is color-coded
knowledgment, or (b) copied the section from an-          and linked to its original source. However, this
other paper with proper acknowledgment but left           color-coding can be deceptive, as it only specifies
out the quotation marks or page numbers, thus             the use of external sources without identifying
giving the impression that the work was para-             whether these sources have been properly cited. In
phrased. Since the focus of this study was on indi-       addition, legitimate use of statistical phrases and
viduals plagiarizing others’ work without appro-          other descriptive terms are also highlighted as
priate acknowledgment, we adopted a more                  plagiarized. To overcome the above limitations, we
conservative approach toward self-plagiarism. If          manually checked the highlighted sections for ap-
authors used sections from their own previous
work or cited the primary source, then it was not
considered plagiarism.                                                      TABLE 1
                                                            Mean Number of Words Plagiarized, by Author
                                                                         Characteristics
Independent Variable
                                                                                                 Mean Number
We coded for the author’s academic status as fol-                                                  of Words
lows: 1 ⫽ student, assistant professor, lecturer, re-                                      N      Plagiarized
search assistant, nonacademic (essentially, non-
tenured or junior scholars) and 2 ⫽ associate             Status
                                                            Untenured or Junior Scholars   304       76.67
professor and professor (essentially, tenured or se-
                                                            Tenured or Senior Scholars     279       86.68
nior scholars). However, we should clarify that             Total                          583       81.55
while in many countries, and particularly in North        Country
America, associate and full professors are tenured          Noncore Countries              129      203.09
faculty, in a limited number of systems it may be           Core Countries                 501       66.64
                                                            Total                          630       94.63
possible to hold the position of untenured associ-
                                                          Gender
ate professor. Further, although tenure does not            Male                           379      117.30
exist worldwide, senior scholars typically enjoy            Female                         206       66.03
greater prestige and receive more resources. We             Total                          585       90.23
next coded for the country in which the author’s          Education
                                                            Non-English Speaking Country   189       96.32
university was located: 1 ⫽ established (core)
                                                            English-Speaking Country       319       67.24
countries and 0 ⫽ otherwise. For the purposes of            Total                          508       78.06
this paper, core countries were defined as western-
You can also read