Pi­lots of EU-backed £6 mil­lion tool to tackle on­line cheat­ing re­port suc­cess

THE (Times Higher Education) - - CONTENTS - Anna.mckie@timeshigh­ere­d­u­ca­tion.com

A e7 mil­lion (£6.2 mil­lion) tool de­signed to stamp out cheat­ing in on­line as­sess­ment is pre­par­ing for launch af­ter suc­cess­ful Euro­pean tri­als.

The Adap­tive Trust- based E-as­sess­ment Sys­tem for Learn­ing (Tesla), funded by the Euro­pean Union, com­bines anti-pla­gia­rism soft­ware with fa­cial-, voice- and key­stroke-iden­ti­fi­ca­tion tech­nol­ogy.

More than 27,000 stu­dents at seven Euro­pean uni­ver­si­ties, in­clud­ing the UK’s Open Univer­sity and the Open Univer­sity of Cat­alo­nia (UOC), have now taken part in tri­als of the tool.

Denise White­lock, pro­fes­sor of tech­nol­ogy-en­hanced as­sess­ment and learn­ing at the UK’s OU, said that Tesla had “had a good re­sponse” and had proved to be a ro­bust anti-pla­gia­rism tool.

It will be de­vel­oped into a prod­uct avail­able for pur­chase by uni­ver­si­ties, al­beit with a “ba­sic” ver­sion likely to be made avail­able free of charge, af­ter the re­search project con­cludes in March.

A key weak­ness of many ex­ist­ing on­line as­sess­ment sys­tems is that they may al­low a stu­dent to en­list a sub­sti­tute to do work in their name by giv­ing that per­son their user­name and pass­word, or may al­low a stu­dent to en­ter their de­tails to start an exam and then have some­one else take it for them.

Us­ing tech­nol­ogy to ver­ify the iden­tity of peo­ple sit­ting on­line ex­ams – be it through their ap­pear­ance, the sound of their voice, their way of typ­ing or their man­ner of writ­ing – could im­prove the rep­u­ta­tion of on­line qual­i­fi­ca­tions. It could also re­duce on­line providers’ costs by spar­ing them the need to con­duct face-to­face as­sess­ments.

Anna Elena Guer­rero, the project co­or­di­na­tor at the UOC, said that stu­dents ini­tially found the fa­cial recog­ni­tion soft­ware to be “the most in­tru­sive” el­e­ment of the tool. How­ever, those in­volved in more than one round of pi­lot­ing of­ten ended up chang­ing their per­cep­tions. “In the be­gin­ning they felt that fa­cial recog­ni­tion was very in­tru­sive, but later that de­creased,” she said.

“Mainly stu­dents saw Tesla as an ad­van­tage, be­cause they can save time on com­mut­ing or they feel it makes their de­gree stronger,” Dr Guer­rero said. “It will also be good for com­pa­nies that are go­ing to hire these kinds of peo­ple [who take on­line cour­ses] to know that we can iden­tify au­thor­ship and that the qual­i­fi­ca­tions are ro­bust.”

Pro­fes­sor White­lock said that uni­ver­si­ties could es­chew bio­met­ric iden­ti­fi­ca­tion in the tool, but added that it might be best to em­ploy it while mak­ing it “very clear where this data will go”. “Stu­dents are used to sit­ting in an exam room; we need to ed­u­cate peo­ple about the soft­ware and ex­plain the ad­van­tages,” she said.

Last year, re­searchers at Ari­zona State Univer­sity re­ported suc­cess in de­tect­ing cheat­ing in on­line as­sess­ments by track­ing whether stu­dents tilted their heads to read ma­te­rial off-screen and whether they paused be­fore an­swer­ing ques­tions.

An­other OU re­search project ex­plored us­ing we­b­cams to mon­i­tor stu­dents’ en­gage­ment and emo­tions via their fa­cial ex­pres­sions or eye move­ments, in a bid to im­prove re­ten­tion and course de­sign.

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.