41September/October 2001� EDUCAUSE r e v i e w40 EDUCAUSE r e v i e w � September/October 2001

Marie Groark is Principal Consultant for PricewaterhouseCoopers. Diana Oblinger is Senior Fellow, EDUCAUSECenter for Applied Research, and Professor of the Practice, Kenan-Flagler Business School, University of North Carolina at Chapel Hill. Miranda Choa is Management Consultant for PricewaterhouseCoopers.

Pl agiarism is not new. However, the May 10, 2001, headline “U of Virginia Hit by Scandal over Cheating” on the front page ofthe New York Times rekindled debates on the topic of cheating. Thefollowing day the Chronicle of Higher Education carried an article ti-tled “Cheating Is Up at Amherst College, Data Suggest.” And a fewdays later USA Today ran an editorial, “Cheating Thrives on Cam-pus, As Officials Turn Their Heads.”

and AcademicAnti-Plagiarism Tools,

Term Paper Mills,

By Marie Groark,Diana Oblinger, andMiranda Choa

© 9

/01

Ste

ve M

cC

rac

ken

Integrity

There is no current estimate of the numberof sites, although some lists of Internetpaper mills are maintained by academicg ro u p s ( e . g. , < h t t p : / / w w w. c o a sta l .edu/library/mills2.htm>). These sites at-tract secondary school students as well ascollege and university students. They arealso not exclusive to the United States.

The growing number of term papermill sites on the Web attest to their popu-larity among students.

■ AP Business wire reports that traffic tothese sites exceeds 2.6 million hits permonth.

■ Cheater.com has 72,000 members andis growing by a few hundred per day.

■ With 9,500 papers in its database, theEvil House of Cheat reports 4,000 visi-tors a day.

■ Scho olsucks.com, which claims10,000 visits to its site per day, reportsbeing profitable “from Day 1.”13

Institutional Attitudes toward Academic DishonestyAlthough academic dishonesty is be-lieved to have increased in the last twodecades, it is not clear that the number ofinfractions reported by professors hasrisen as well. In a survey of 800 faculty

members who were asked why they ig-nored possible plagiarism violations,professors cited inadequate administra-tive support as a primary factor.14

Research by Donald McCabe has in-dicated that there is an inverse correla-tion between the rate of plagiarism andthe emphasis on academic integrity by in-stitutions or instructors.15 Thus a growingnumber of institutions are addressing ac-ademic integrity through honor codes,pledges, and discussions of ethics. Onepolitical science professor at OaktonCommunity College, for example, giveshis students a six-page letter spelling outhis expectations of them, as well as hisobligations to them. In the first page heasks: “Would you want to be operated onby a doctor who cheated his way throughmedical school? Or would you feel com-

fortable on a bridge designed by an engi-neer who cheated her way through engi-neering school? Would you trust your taxreturn to an accountant who copied hisexam answers from his neighbor?”16

Once an instructor suspects plagia-rism, it can be a laborious process provingthat plagiarism has actually taken place.Instructors may need to comb through oldpapers and primary and secondary re-sources and compare the suspicious paperto these sources. Tracking down a stu-dent’s sources and proving plagiarism cantake days. Those who have used an auto-mated plagiarism tool cite the streamlinedprocess as one of the primary advantagesof the tool. But most important, papersplagiarized from the Internet and identi-fied by an anti-plagiarism tool often pro-vide an open-and-shut case.

Tools to Ensure Academic IntegritySeveral vendors provide tools designed toensure academic integrity (or to identifycases of suspected plagiarism). These toolscan be differentiated on the basis of severalfeatures. For example, some vendors offerusers the opportunity to download theirproprietary software and grant a one-person license to use the software as ananti-plagiarism device for an unlimited

number of users. Those vendors that donot have downloadable software use anInternet-based application. Fees are alsohandled differently by different vendors.Some charge users by the paper; otherscharge for a specific time period; a fewchange a one-time-only fee. Another im-portant distinction among many vendorsis the databases they search. Vendors’search engines typically match submis-sions against a proprietary database con-taining company-specific content. Thesedatabases may contain other student sub-missions and papers from term papermills or other sources. In addition to data-bases, some vendors’ search enginesmatch submissions against content foundon the Internet. Finally, the amount oftime it takes to process a search varies aswell. Times range from less than an hour to

two days. The following examples of spe-cific vendor products illustrate how theyfunction, as well as which features differ.

iParadigms has developed a search en-gine, iThenticate, that is able to clearlyidentify matching texts between two text-based documents of any language or size.The search engine examines a chunk oftext, eliminates words that are too com-mon, and turns the other words intonumbers. It also converts Internet con-tent into numbers. Consequently, it cancompare patterns of numbers from thesample with patterns of numbers fromthe Internet. To do this, the search engineuses a number of complex mathematicalalgorithms. iParadigms’ search engine isused at Web sites that monitor plagiarism(e.g., Turnitin.com) and at a Web site fo-cused on media such as movies andmusic (Slysearch.com). The initial site,Plagiarism.org, now serves as a Web re-source about online plagiarism.

Originally, Plagiarism.org was createdto screen for Internet plagiarism at theUniversity of California at Berkeley. Itdrew national attention when a studyfound that up to 15 percent of studentswere copying online material and using itin term papers. When demand for theanti-plagiarism technology exceeded the

site’s ability to provide it, a second site,Turnitin.com, was created. This site nowprovides access to the iThenticate tech-nology through a Web front-end.17

Turnitin.com is iParadigms’ portal for registered users of the company’s services. Registered users may includeboth faculty and students. Students whohave completed assignments and wantto be sure that they have not inadver-tently missed citations can submit pa-pers to Turnitin.com. Faculty may alsodirect students to submit papers to thesite. At Turnitin.com, students or in-structors submit papers that will betested against the Internet or proprietarydatabases for plagiarism violations.

Th e c o m p a ny b e l i e v e s t hat t h e recirculation of term papers is a keysource of plagiarism; therefore, its

43September/October 2001� EDUCAUSE r e v i e w

Figures from around the country aredrawing attention to the issues of cheat-ing, plagiarism, and academic integrity:

■ At the University of Virginia, 122 stu-dents were accused of cheating onterm papers in introductory physics;half may face expulsion or loss of de-grees awarded in earlier years.1

■ Cases of suspected cheating and pla-giarism at Amherst College averagedfive a year from 1990 to 1998 but in-creased to sixteen in 1999 and nine-teen in 2000.2

■ Reported occurrences of academic dis-honesty at the University of California–Berkeley doubled between 1995 and1999.3

■ In a recent survey conducted by Don-ald McCabe, founder of the Center forAcademic Integrity at Duke University,72 percent of high school students re-ported one or more instances of seriouscheating on written work, and 15 per-cent of students reported submissionsof papers obtained “in large part” froma term paper vendor or Web site.4

■ A study by the Center for AcademicIntegrity found that almost 75 percentof college students own up to someform of academic dishonesty.5

■ At Penn State, despite the fact that fac-ulty had discussed the consequencesof cheating with 63 percent of the stu-

dents surveyed, 17 percent of the stu-dents said they had cheated on testsand 44 percent said they had cheatedon class assignments.6

The amount of cheating appears to beincreasing. For example, at medium-to-large universities, the percentage of stu-dents who said they collaborated on as-signments even though it was notpermitted increased from 11 percent in a1963 survey to 49 percent in 1993. Forthirty-one small-to-medium institutions,unpermitted collaboration increasedfrom 30 percent to 38 percent between1990 and 1995.7

Furthermore, the ease with which information can be copied from the Web and the emergence of term paper ven-dors or “mills” on the Internet are likelyadding to the growing problem of plagiarism. For example, a neuro-biology professor at the University of California–Berkeley found that 45 of 320students in his class had plagiarized at leastpart of their term paper from the Internet.Nearly 15 percent of his students plagia-rized even after they had been warned thathe would use anti-plagiarism technology.8

In a recent survey commissioned by Knowledge Ventures, an education

technology company, more than 90 percent of academic administrators andfaculty interviewed said that academic

integrity is an issue on their campus.Most were unable to pinpoint the extentof the problem, the source of the prob-lem, or whether specific departments orstudent groups were more at risk. In ad-dition, of those who stated that aca-demic integrity is an issue, 83 percentsaid that it has become more of an issueover the last three to five years, prima-rily due to the use of the Internet as a re-search tool. Compounding the effects of the Internet are difficulties in provingviolations and a reluctance to report violators.9

Term Paper MillsTerm paper mills existed long before theInternet. Companies who sell term papers have advertised on campus and inmagazines such as the Rolling Stone forseveral years.10 With the advent of Inter-net technology, though, the number ofplaces where papers are available hasgrown and the ease with which paperscan be obtained has increased. Some ofthese Web sites are operations set up by students while others are for-profitventures.

At term paper mills, students can di-rectly purchase pre-written papers. Somesites offer free services or make money

through advertising. Others act as an ex-change—a student must submit a paper toget a free paper. Most term paper millscharge a fee, ranging from about $5 to $10per page. Students may pay an additionalfee for immediate e-mail delivery (e.g.,$15). Other sites will write a customizedpaper for a much higher fee.

In most states, it is illegal to sell papersthat will be turned in as student work.11

Thus many for-profit sites post dis-claimers saying that the informationshould be used only for research pur-poses and should not be submitted as astudent’s own work. The companies willbill a student’s credit card using an unrec-ognizable company name.

Experts estimated that more than 70term paper mills were in operation in early1998, up from 28 at the beginning of 1997.12

42 EDUCAUSE r e v i e w � September/October 2001

In most states, it is illegal to sell papers that will be turned in as student work.

database contains student papers, papersposted online, material from academicWeb sites, and documents indexed bymajor search engines. Its database con-sists of 800 million Internet pages andmore than 100,000 papers. Student pa-pers submitted by registered users arealso archived to the database. Thus, extended use of the service will build aninstructor’s archive of papers and will en-sure that students cannot easily recyclepapers from previous classes.

If an instructor is interested in “test-ing” a paper for plagiarism, he or she submits it for processing through the pro-prietary search engine. Once the processis complete, an originality report indicatesthe probability (in terms of a percentage)of whether the paper was plagiarized. In-structors can click on links to direct themto the source of the possibly plagiarizedmaterial. This process takes twenty-fourhours on average but can take up to two

days, depending on the length of the textand the level of demand.

Turnitin.com was designed to provideusers with a simple process for submit-ting papers for a plagiarism test. A usercompletes a short form to submit thepaper. The user identifies institution, de-partment, course, name, and ID number.Then the user may submit the paper fortesting by pasting a text-only documentonto the Web page.

The company has developed a four-tier pricing structure: an individual in-structor can opt for a plan that provides100 originality reports for $100 per se-mester. A department with no more than50 classes can choose a plan that provides500 originality reports for $650 per se-mester. A small institution with no morethan 150 classes per semester can get5,000 originality reports for $1,750 per se-mester. Finally, a large institution can re-ceive an unlimited number of originalityreports for $4,000 per semester.

Another vendor is EVE2 (Essay Verifi-cation Engine), anti-plagiarism softwarethat instructors may license for a one-time fee and download to their harddrives. A user may submit a paper in .txtform, and the software will search the In-ternet, including term paper mills, for re-sults. Although the software is targeted atall educational levels, most users seem tobe high school teachers.

As noted, an instructor must convertpapers to .txt format and submit the textto the software (which has been down-loaded to his or her hard drive). EVE2then examines the papers and makes alarge number of searches of the Internetto locate “suspect” sites. Once suspectsites have been located, EVE2 visits all ofthese sites to determine if they containwork that matches the paper in question.

Searches require fifteen minutes to

two hours for processing. This time is inpart linked to the power of the user’s harddrive. Once the search is complete, the in-structor is given a full report on eachpaper that contained suspected plagia-rism, including the percentage of thepaper plagiarized, an annotated copy ofthe paper showing all plagiarism high-lighted in red, and links to the plagiarizedsites. The report does not distinguishwhich sentences have been plagiarizedfrom which sites. There is a fifteen-dayfree trial of the software, after which timea customer may purchase a license to thesoftware for $19.99.

New companies are planning to enterthis market as well. One is KnowledgeVentures, headquartered in Boston,which underwrote portions of the re-search for this article. Knowledge Ven-tures is developing a suite of tools andmiddleware components for use in text-to-text matching applications. CiteMaster,

the first application to incorporate thesetools, will use supercomputer processingpower to compare submitted text docu-ments against the content of a proprietaryacademic database of textbooks, journals,Web content, and student-submitted pa-pers. Through the use of customized algo-rithms and a parallel-processing plat-form, CiteMaster should be able to detectboth verbatim and inexact text matchesand to return results in real time. For eachstudent-submitted paper, a dynamicallygenerated “report” will be returned to aWeb browser, including an HTML recon-struction of the original query documentand relevant statistical information aboutthe matches found. Matching or “suspect”sentences will be clearly identified in thetext, with their proper citations listed ac-cordingly. CiteMaster is in the early stagesof development, however, and is not yetcommercially available.

Lastly, other vendors focus on the factthat patterns of words in documents can beused for more than detecting plagiarism.

44 EDUCAUSE r e v i e w � September/October 2001

An instructor interested in “testing” a paper forplagiarism can submit it for processing through a proprietary search engine.

However, not all plagiarism can be at-tributed to the increased use of the Inter-net. In a study of 4,500 students fromtwenty-five high schools, 54 percent hadused the Internet to plagiarize. But the re-search also showed that the majority ofthose cheating would have plagiarizedwithout the Internet. Only 6 percent ofthose who submitted plagiarized workhad relied solely on the Internet.19 As

technology continues to change, so do theoptions for cheating. Information may bestored on calculators, for example, and e-mails can be sent via palm-held devices.

Will plagiarism increase as today’shigh school students enter college? If ex-isting attitudes are an indicator, it might.In a survey by Who’s Who among AmericanHigh School Students, 84 percent of studentssaid cheating was common in their highschools. According to studies by DonaldMcCabe, less than half of the students sur-veyed felt that copying a few sentences

without a citation was wrong. In addition,22 percent of students in the study turnedin an assignment done by their parents.Another poll found that 66 percent of stu-dents said that cheating “didn’t seem like abig deal.” Sixty-six percent of their par-ents agreed.20 In addition, in focus groupsinvolving high school students, manyagreed with the following statement:“Many of our teachers are clueless when it

comes to the Internet, the material youcan find on the Internet is of sufficientquality to submit on your assignments,and paper topics are usually so broad thatyour teachers are not at all likely to recog-nize a source you might use.”21

Although the anti-plagiarism toolscur rently available can search for instances of suspected plagiarism, thisplagiarism-screening approach is not al-ways integrated with other activities,such as the discussion of ethics and over-all academic integrity. Identifying in-

stances of suspected plagiarism withoutproviding a more developmental remedymay not provide a long-term solution. Forexample, a major factor determiningwhether or not a student will cheat is theacademic culture of the specific institu-tion that he or she attends. At the fewerthan one hundred institutions with ro-bust academic honor codes in place—in-stitutions where students pledge not to

cheat and where they play a major role inthe judicial process—significantly fewerincidents of cheating occur. Size of theinstitution does not seem to matter. Re-cent evidence shows that cheating can bereduced, even on large campuses, withstrong honor codes and a focus on aca-demic integrity.22 When students are fre-quently reminded that their institutionplaces a high value on academic integrity,and when this value is reinforced by thecampus community, the incidence ofcheating is lower.

47September/October 2001� EDUCAUSE r e v i e w

For example, WordCHECK Systems pro-motes KeyWORD software as helping

■ legal researchers identify, archive, andtrack legal decisions by specific judgesand court jurisdictions;

■ grant-writers track keyword patternsin more effective grant proposals;

■ political scientists find keyword pro-files of political communications made

by candidates and officials in speeches,press releases, and opinion papers;

■ corporate information officers iden-tify, archive, and track keywords in e-mail to ensure the protection of pro-prietary information; and

■ market researchers develop betterqualitative information by trackingkeywords found in focus groups andsurveys.18

Implications for Higher EducationThe amount of time that students spendonline has steadily increased, with theb u l k o f t hat t i m e b e i n g s p e n t o n academic pursuits. With the growth in students’ use of the Internet and theattitude that information should be“free,” it is perhaps not surprising thatincidents involving plagiarism have alsoincreased.

46 EDUCAUSE r e v i e w � September/October 2001

Another poll found that 66 percent of students said that cheating “didn’t seem like a big deal.”

Campuses have responded to the aca-demic integrity issue by taking the fol-lowing actions:

■ Increasing the emphasis on the institution’shonor code. This may include requiringthat freshman sign the honor code ordiscussing the issue with students.One university gives all incoming stu-dents a CD-ROM that provides anoverview of intellectual property andcopyright issues.

■ Forming student groups to monitor and promote the honor code. A handful of institutions have created student groupsthat are responsible for promoting aca-demic integrity on campus. Thesegroups may make presentations aboutthe importance of academic integrity tostudents, for example, or may conveneworkshops about academic integrity.

■ Rewriting student guid elines. Many campuses have rewritten studentguidelines to emphasize academic integrity.

■ Establishing an academic integrity office.Some campuses have formed offices ofacademic integrity. These offices mayprovide an annual review of the statusof academic integrity on campus, de-velop a forum to address issues ofacademic integrity, and/or oversee afaculty-student council designed toresolve incidences of academic dishon-esty as they arise.

■ Convening faculty workshops aimed at moreeffectively addressing the issue in classrooms.Faculty may be unaware of the growthof academic dishonesty on campus orof specific techniques that can be usedto combat this issue. Many campuseshave convened workshops and lec-tures specifically designed to teachfaculty how to identify and deal withacademic dishonesty.

■ Utilizing technology-based tools. Somecampuses use these tools to find andverify suspected cases of plagiarism.

It is not just cheating by students thatappears to be on the rise today. The pres-sures and opportunities for dishonest be-havior are increasing in many academicand professional contexts, ranging fromusing term paper mills to falsifying cre-dentials to fabricating research results.Academic integrity is a fundamental ten-

ant for higher education. Both higher ed-ucation and society benefit when collegeand university integrity standards pro-vide a lifelong foundation.23 Althoughtechnology tools have a role to play in thismission, higher education must find thesolution to academic integrity problemsthrough an integration of technology,policies, and personal interaction. e

Notes1. Diana Jean Schemo, “U of Virginia Hit by Scandal

over Cheating,” New York Times, May 10, 2001.2. “Cheating Is Up at Amherst College, Data Suggest,”

Chronicle of Higher Education, May 11, 2001, A11,<http://chronicle.com/weekly/v47/i35/35a01103.htm> (accessed July 12, 2001).

3. “Cheating Thrives on Campus, As Officials TurnTheir Heads,” USA Today, May 21, 2001.

4. Donald L. McCabe, “Student Cheating in AmericanHigh Schools,” May 2001, <http://www.academicintegrity.org/index.asp> (accessed July 12, 2001).

5. See <http://www.academicintegrity.org/cai_research.asp> (accessed July 12, 2001).

6 . S e e < h t t p : / / w w w . s a . p s u . e d u / s a r a / p u l s e /academic.shtml> (accessed July 12, 2001).

7. See <http://www.academicintegrity.org/cai_research.asp> (accessed July 12, 2001).

8. Verne G. Kopytoff, “Brilliant or Plagiarized? Col-leges Use Sites to Expose Cheaters,” New YorkTimes, January 20, 2000.

9. This survey was conducted in February 2001 byPricewaterhouseCoopers on behalf of KnowledgeVentures.

10. Peter Applebome, “On the Internet, Term PapersAre Hot Items,” New York Times, June 8, 1997.

11. Ibid.; see also Ronald B. Standler, “Plagiarism in Colleges in USA,” <http://www.rbs2.com/plag.htm#anchor333347> (accessed July 15, 2001).

12. John N. Hickman, “Cybercheats: Term Paper Shop-ping Online,” New Republic 218, no. 12 (March 23,

1 9 9 8 ) : 1 4 , < h t t p : / / w w w 2 . b c . e d u / ~ r a p p l e b /Plagiarism.htm> (accessed July 23, 2001).

13. Kendra Mayfield, “Catching Digital Cheaters,”Wired News, February 29, 2000, <http://www.wired.com/news/print/0,1294,33021,00.html> (accessed July 12, 2001).

14. “Why Professors Don’t Do More to Stop StudentsWho Cheat,” Chronicle of Higher Education, January 22,1999.

15. “New Research on Academic Integrity: The Successof ‘Modified’ Honor Codes,” College AdministrationPublications, <http://www.collegepubs.com/ref/SFX000515.shtml> (accessed July 12, 2001).

16. Bill Taylor, “Integrity—Academic and Political: ALetter to My Students,” <http://www.academicintegrity.org/pdf/Letter_To_My_Students.pdf>(accessed July 12, 2001).

17. See <http://www.iparadigms.com/plagiarism.html> (accessed July 17, 2001).

18. S e e < h t t p : / / w w w . w o r d c h e c k s y s t e m s . c o m /applications.htm> (accessed July 17, 2001).

19. Cara Branigan, “Rutgers Study: Web Makes Student Cheating Easier,” eSchool News Online,June 1, 2001,<http://www.eschoolnews.org/showstory.cfm?ArticleID=2638> (accessed July 19, 2001).

20. “A College Test: How to End Epidemic of CheatingStudents,” Philadelphia Inquirer, April 22, 2001.

21. “New Research on Academic Integrity: The Suc-cess of ‘Modified’ Honor Codes,” College Administra-tion Publications, <http://www.collegepubs.com/ref/SFX000515.shtml> (accessed July 12, 2001).

22. Ibid.23. See Center for Academic Integrity, “The Funda-

mental Values of Academic Integrity,” October1999 <http://www.academicintegrity.org/Values.asp> (accessed July 12, 2001).

This piece is adapted from a sampleresearch bulletin produced for the new EDUCAUSE Center for AppliedResearch (ECAR). For more infor-mation about ECAR, visit <http://www.educause.edu/ecar/>.

48 EDUCAUSE r e v i e w � September/October 2001