Plagiarism in five universities in Mozambique: Magnitude, detection techniques, and control measures

Definition and symptoms

Plagiarism is presenting as your own another person’s words, ideas, data, artwork or designs—unless considered common knowledge—without referencing the true author. In written works, this happens in three forms: (i) verbatim (word-for-word) copies without quotation marks and/or references (ii) paraphrases without references and (iii) “use of another author’s unique ideas, data or evidence (though not their words) without reference to the source” (Coughlin, 2013, p. 18). From these basic types are derived many, often overlapping subtypes and combinations. For example, mosaic plagiarism occurs when the work uses “bits and pieces from a source (or several sources), changing a few words here and there without either adequately paraphrasing or quoting directly” (Harvard College Writing Program 2013).

Copyright issues aside, reuse of one’s own work, especially if the source is cited and the target audience is largely different than that for the original work, is usually considered acceptable republication. However, if the target audience is mostly the same (e.g., via another scientific journal in the same discipline) or if the source is not acknowledged, the reuse gains a less respectable name, self-plagiarism, especially when large tracts of recycled text are involved (Collberg & Kobourov 2005, p. 92 and 94; Bretag & Mahmud, 2009, p. 193; Andreescu 2013, p. 775 and 796).

Another problem occurring in tandem with plagiarism is the abuse of secondary citations, a frequent practice among students and even some professionals. To avoid errors, accepted professional practice insists on use of the primary source of a citation or data except when access is impossible or extremely difficult. Even in this latter case, one must reference both the primary and secondary sources so readers can judge the wisdom of relying on the latter’s accuracy. In this regard, students frequently commit three errors: (i) taking information and references from secondary sources though the primary source is easily available; (ii) referencing an unread (unavailable?) primary source without mentioning the secondary source where, in fact, they got the information; and (iii) erroneously referencing secondary sources as if they were the primary sources. In all three cases, a superficial review by a professor might conclude that the formal aspects of source referencing were obeyed whereas, in fact, essential information had been omitted (Pecorari 2006, p. 14 and 17).

Gravity of the offense: perceptions, consensus building, and consistent interpretations

To determine the gravity of plagiarism, purposeful deception is the central issue and, if proven, is widely condemned (Pecorari 2013: KL441-42).^a However, purposeful deception is often omitted from academic codes of behaviour to avoid creating a loophole for real plagiarists to allege unintentionality. Instead, intentionality is only considered when pondering which remedial or disciplinary measures to apply to a specific incident (Devlin 2006, p. 48; Pecorari 2013: KL445-46). As Standler (2012) argues, since “copyright infringement occurs without proof of intent by the infringer,… intent should also be irrelevant to determination of plagiarism” (p. 7).

Significant differences of opinion also arise when professors are requested to compare specific texts against their sources and judge whether plagiarism has occurred. For example, in a survey asking 152 professors from five universities in the New York–New Jersey metropolitan area and 49 from the Internet to classify six passages involving progressively worse plagiarism, none of “the paragraphs yielded 100% agreement among respondents”. The professors’ criteria for determining plagiarism ranged from very lax to extremely rigorous (Roig 2001, p. 313).

Students too have divergent perceptions about what and how serious plagiarism is. For example, in two New York colleges where 316 undergraduate students were requested to judge 10 paragraphs, 76% “correctly identified the two paraphrased versions of the original”, but approximately 50% incorrectly identified “six of the plagiarized versions of the paragraph” as not having been plagiarized (Roig 1997, p. 117). Often, students admit to not even having a clear notion of what plagiarism is (Michalska 2012, p. 8). Various studies also confirm that most students do not deem plagiarism a serious offence (McCabe et al. 2012: KL832; Pupovac et al. 2010).

Professors not only hold diverse opinions about what constitutes plagiarism but also react in widely different ways when discovering it in a student’s work. Typically, when considering how to respond, professors and administrators consider “intention, extent, and scale of the offence” but the penalties they recommend vary greatly for the same offence (Flint et al. 2006, p. 149; Carroll & Appleton 2005). Moreover, faculty may be disinclined to cooperate with a university’s regulations requiring central reporting of incidents of plagiarism, if they (i) consider “the offence … as unintentional or caused by personal stress” (Paterson et al. 2003, p. 147), (ii) think the rules would impose unjustly severe penalties in a specific case (Pincus & Schmelkin 2003, p. 208), (iii) feel the university would not “deal with the instances adequately” (Sutherland-Smith 2003, p. 11), or (iv) deem “the misconduct investigations … too cumbersome” and time-consuming (Standler 2012, p. 78).

The problem, therefore, is more than just obtaining a clear definition of plagiarism but also of building consensus and getting both students and professors to apply the definition and appropriate disciplinary or pedagogical measures in a reasonably consistent manner. For example, “to promote systematic and consistent decision-making” about plagiarism, Curtin University of Technology in Australia identified “four dimensions or criteria: experience, nature, extent and intent” (Yeo & Chien 2007, p. 190). In the UK, due to growing concern about “inconsistency in the application of penalties for student plagiarism within higher education”, “the Higher Education Academy funded the Academic Misconduct Benchmarking Research Project, which confirmed vast variation between different institutions in the penalties available for student plagiarism” (Tennant & Rowell 2010, p. 3). After requesting 104 academics to identify important factors for judging the seriousness of and appropriate disciplinary and/or pedagogical response for hypothetical cases ranging from minor to major plagiarism, the project designed a “benchmark tariff for application of penalties for student plagiarism in higher education” (Tennant & Rowell 2010: 5–8 and 13). Though the system never gained widespread acceptance, it “foster[ed] the notion that criteria can be defined and applied for judging the seriousness of an offense” (Weber-Wulff 2014, p. 148).

Early and ongoing prevention

To reduce plagiarism, prevention is best. The primary focus should be on promoting writing and analytical skills, emphasizing the academic and economic benefits of proper techniques and good ethics, providing technological tools to train students to quote, reference and paraphrase properly, and establishing a private or public regulatory framework to ensure that universities, teachers and students cooperate and organize strategically on multiple fronts to prevent academic fraud, including plagiarism.

Administrators need to work with professors and students to build a strong, committed consensus that the academic goal of training original thinkers who have dignity and honour is good for students, universities and nations. Then, universities must teach new students about the necessity and techniques for properly referencing sources and provide instructional materials and referencing software so students and staff can clarify doubts and reduce errors. To eliminate the didactical confusion that arises when definitions and styles vary from professor to professor requires a consistent university-wide referencing policy that permits variations for specific disciplines (e.g., law, medicine, engineering) and, preferably, is reinforced by a sequence of writing skills courses or seminars during the degree course (Cleal 2005, p. 49; Ellery 2008, p. 513–14; Pecorari 2013: KL2358). Moreover, universities should firmly discourage (i) use of secondary sources when the primary sources are available and (ii) dependence on often unreliable sources such as Wikipedia and non-Ph.D. theses since these two practices are often signs of superficial research and poor professional judgment, frequently associated with plagiarism. To enforce this directive, a university might prohibit such practices unless a professor or supervisor grants an exemption due to the exceptional quality of the source.

Buttressing this consensus and regulatory framework, lecturers need to emphasize the importance of good writing and correct referencing (Pecorari 2013:KL2358) and then be manifestly vigilant about plagiarism, a responsibility that many shirk, minify or reject. Wherever many lecturers have such negative attitudes, administrators face a substantial challenge to work with the academic community to build consensus about definitions, proper citation techniques, disciplinary regulations, and the urgent need for lecturers to promote professional ethics, help students to improve their writing and referencing skills, and cooperate with the disciplinary structures for cases of serious plagiarism. To train students and detect plagiarism more easily, lecturers need TSR software and, when reporting significant violations, merit administrative support. Moreover, due to “the strong association between past and future academic dishonesty, … individuals who have participated in academic dishonesty are quite likely to offend again”. Universities, therefore, need a computerized central registry for incidents of cheating handled by a teacher or adjudicated by the university (Williams & Williams 2012, p. 106). The system should detect repetitive cheats and automatically inform the disciplinary structures (Devlin 2006, p. 50).

Universities should also mandate the publication of bachelor, masters and Ph.D. theses over the Internet, which is increasingly popular as more and more universities adhere to the Budapest Open Access Initiative for wide dissemination of peer-reviewed research results (Xia et al. 2012, p. 90; Weber-Wulff 2012, p. 33; BOAI Budapest Open Access Initiative 2002, BOAI 2012; Deutsche Nationalbibliothek 2009). Electronic publication facilitates public access and permits TSR software to scan uploaded theses as possible sources for plagiarized works. Though open-access publication may convince degree candidates that plagiarism is risky, it might threaten the economic interests of an author whose thesis presents potentially valuable discoveries, inventions or literary works. To safeguard such content, authors could be allowed to justify and request a temporary but renewable embargo on publication (Hawkins et al. 2013).

With exponential growth of the Internet, the ghost-writing of term papers, theses and dissertations became a burgeoning business threatening the credibility of the degrees issued by educational institutions where academic fraud is rife (Rooks 2006). To combat this, many US states have criminalized the direct or indirect sale of term papers, theses, dissertations or “other materials to be submitted for academic credit” (Berry 2010, p. 131–32). Some states permit universities “to request a court to enjoin a business from selling term papers, etc., to its students” (Standler 2012, p. 69). Despite many successful cases enjoining or closing down such companies, the laws and law suits have not halted the paper mills’ popularity or profitability. Nevertheless, such laws comprise an important part of a holistic anti-plagiarism strategy (Dickerson 2007, p. 46).

Other legal or regulatory measures may also help in contexts where some higher education institutions are strong and others are new and academically weak. National regulation can require all higher education institutions to (i) use TSR software to screen theses and, eventually, class papers, (ii) divulge theses on the Internet, and (iii) define and implement a holistic strategy against plagiarism and other academic fraud.^d And, to thwart use of class papers from prior years or other nearby higher education institutions, TSR software should be permitted to retain copies in their databases.

Pedagogical changes can also be useful. For example, Kerkvliet and Sigmund (1999), p. 341) found that “students taught by faculty, rather than graduate teaching assistants (GTAs), were 32% less likely to cheat. This is strong evidence that the use of GTAs is costly in terms of academic honesty”. Also, the reuse of “old assessment tasks and topics makes it easy for students to plagiarise” from prior papers (Ford & Hughes 2012, p. 184). Another “strategy is to avoid giving open-ended or generic topic assignments. Allowing students to write about any topic of their choosing (‘pick some topic related to this course’) or giving general topics (‘write on a social concern’) makes plagiarism appealing and easy” since online cheat sites provide thousands of papers on such topics (Rooks 2006, p. 4). Instead, professors should, when possible, assign projects involving field or laboratory research requiring original work thus reducing opportunities for plagiarism or, worse yet, the purchase of ready-made reports from “paper mills”. As for theses, oral examination committees must be keenly aware of their two critical functions: (i) to determine the quality of the work and (ii) to use detailed questioning and, in seriously dubious cases, perhaps even the Cloze test to confirm that the work is, in fact, that of the degree candidate (McKelvie et al. 2004, p. 9; Rooks 2006, p. 6; Standing & Gorassini 1986, p. 132).

Traditional or modified honour codes also yield measureable results though their implementation may encounter scepticism and resistance, especially from lecturers who lack honour-code experience (McCabe et al. 2003, p. 383). However, honour codes and shared governance make students partly responsible for dealing with cheating. This reduction represents “an important selling point to those non-code faculty who are reluctant to take on the unwanted responsibility of monitoring and policing student cheating” (McCabe et al. 2003, p. 383).

Resistance to honour codes and even to the entire effort to detect and discourage plagiarism may be intense if many of the lecturers graduated from institutions rife with academic fraud, as sometimes happens in countries with rapidly expanding university systems and many new, often small, and far from robust higher education institutions. In these conditions, some staff may feel personally threatened by proposals for a stricter code. To mitigate fears of possible denunciations and garner support may require an amnesty for past plagiarism while senior, highly published academics mount vigorous pedagogical efforts to ensure that all professors, teaching assistants and academics know the norms well. Thus, to confront such an academic crisis, the focus must be on the future, not the past.

Detection and consequences

Though, to reduce plagiarism, prevention must be primary, analysts generally agree that detection and remedial and disciplinary measures are also essential (Weber-Wulff 2014; Blum 2010; Pecorari 2006; McCabe et al. 2012). To detect plagiarism requires skills and technology complemented by clear agreed-upon administrative structures and regulations to guide professors confronting minor offences, and disciplinary committees judging serious cases. For example, how does one judge the seriousness of a case of plagiarism? What should the consequences be when deceptive intent is clearly present? Or when it is not? Teaching and minor reprimands seem best for new and junior students who are first or second offenders, with sterner measures reserved for senior and repeat offenders (Tennant & Rowell 2010; Yeo & Chien 2007). Definite consequences must also be mandated for plagiarism discovered in a thesis after the diploma is issued (Pavela 1999; Standler 2012). Moreover, when denouncing significant academic fraud, lecturers must be confident that their diligence is appreciated and appropriate disciplinary and pedagogical measures will ensue (Schmelkin et al. 2001, p. 6; Heckler et al. 2013, p. 244).

Detection measures and disciplinary consequences inspire proper practice and better learning. For example, in papers submitted by 664 students in one US university, 16.6% of the content was plagiarized if the use of TSR software not anticipated versus 9.8% after a warning (Heckler et al. 2013, p. 235). Similarly, among 569 internal and external business students at a New Zealand university, 31.4% committed plagiarism before the warning, whereas, once forewarned, only 21% did so on the next assignment (Walker 2010, p. 9). Thus, vigilance, detection and punishment—taken together—buttress the educational and normative approaches to prevention. Toward that end, participation by student leaders in educational efforts to deter plagiarism and, later, in disciplinary hearings fortifies the desire by students to learn to cite correctly and avoid plagiarism.