Please use this identifier to cite or link to this item: http://hdl.handle.net/1822/57898

TitlePredicting completion time in high-stakes exams
Author(s)Carneiro, Davide Rua
Novais, Paulo
Durães, Dalila
Pego, José Miguel
Sousa, Nuno
KeywordsNeural Networks
Online exams
Random decision forests
Stress
Issue date2019
PublisherElsevier B.V.
JournalFuture Generation Computer Systems
Abstract(s)For the majority of students, assessment moments are associated with significant levels of stress and anxiety. While a certain amount of stress motivates the individual and improves performance, too much stress will have the contrary effect. Stress has therefore a fundamental role on student performance. It should be the educational organizations’ mission to understand the underlying mechanisms that lead to performance anxiety and provide their students with the best coping tools and strategies. In the present study we analyze student behavior during e-assessment in terms of mouse dynamics. Two major behavioral patterns can be identified, based on ten features that quantify the performance of the student’s interaction with the computer: (1) students who are able to sustain performance during the exam and (2) students whose performance varies significantly. Data shows that the behavior of each student during the exam correlates strongly with the time it takes the student to complete it. Several classifiers were trained that predict the completion time of each exam based on the students’ interaction patterns. Two of them do it with an average error of around twelve minutes. Results show that there are still mechanisms that can be explored to better understand the complex relationship between stress, performance and human behavior, that can be used for the implementation of better stress detection, monitoring and coping strategies.
TypeArticle
URIhttp://hdl.handle.net/1822/57898
DOI10.1016/j.future.2018.01.061
ISSN0167-739X
Peer-Reviewedyes
AccessOpen access
Appears in Collections:ICVS - Artigos em Revistas Internacionais com Referee

Files in This Item:
File Description SizeFormat 
Carneiro D 2019 vCT.pdf2,28 MBAdobe PDFView/Open

Partilhe no FacebookPartilhe no TwitterPartilhe no DeliciousPartilhe no LinkedInPartilhe no DiggAdicionar ao Google BookmarksPartilhe no MySpacePartilhe no Orkut
Exporte no formato BibTex mendeley Exporte no formato Endnote Adicione ao seu ORCID