Evaluation Of E-learning Platforms: A Case Study

3y ago
11 Views
2 Downloads
447.56 KB
13 Pages
Last View : 1m ago
Last Download : 3m ago
Upload by : Konnor Frawley
Transcription

Informatica Economică vol. 16, no. 1/2012155Evaluation of E-learning Platforms: a Case StudyCristina POPAcademy of Economic Studies, Bucharest, Romaniacristinel19@yahoo.comIn the recent past, a great number of e-learning platforms have been introduced on the marketshowing different characteristics and services. These platforms can be evaluated usingmultiple criteria and methods. This paper proposes a list of selected quality criteria fordescribing, characterizing and selecting e-learning platform. These criteria were designedbased on e-learning standards. I also propose a mathematical model to determine theprobability that a student uses an e-learning platform based on the factors (criteria) thatdetermine the quality of the platform and the socio-demographic variables of the student. Thecase study presented is an application of the model and the input data, intermediatecalculations and final results were processed using SAS (Statistical Analysis Software).Keywords: E-Learning Platform, E-Learning Standards, Learning Object, LogisticRegression, Quality Criteria List, Univariate Analysis1IntroductionThe World Wide Web is a repository ofcontent (files, databases, datasets, images,video or audio clips, simulations, animations,etc.) of all known formats and standards. Theexcessively increasing load of information onthe Internet leads to an inevitable overload ofuseless information or information forcommercialization purposes. Teachers andstudents may not use this information fortheir educational need but rather as a globalnetwork for communication, interaction andsharing. Within the online context, the usercan be a content “producer” and “consumer”simultaneously [1], thus leading to a hugeamount of raw information, produced by ahuge number of heterogeneous users withoutany didactic reformation applied andincapable to support classroom learningdesign. In the education sector, there isalways a quality control procedure takingplace against the educational material of theschools from the Ministry of Education.Therefore in the classical media context, isalso need of multiple criteria and methods toapprove the quality of e-learning content ande-learning software.2 E-learning platformTraditional means of learning restrict thelearner to certain learning methods, at aspecific time and place whereas e-learningservices create wider horizons fororganizations and individuals who areinvolved in the learning process. Theseenvironments facilitate the delivery of thelearning materials so the learner can accessthem at home or at the office.The most part of contemporary e-learningplatform can be viewed as organized intothree fundamental macro components: aLearning Management System (LMS), aLearning Content Management System(LCMS) and a Set of Tools for distributingtraining contents and for providinginteraction [2]. The LMS integrates all theaspects for managing on-line teachingactivities. The LCMS offers services thatallow managing content of the units while theSet of Tools represents all the services thatmanage teaching processes and trators).An e-learning platform can be characterizedthrough the following management services: services for including and updating userprofile; services for creating courses andcataloguing them; services for creating tests describedthrough a standard; user tracking services; services for managing reports on coursefrequency and use;

156 services for creating, organizing andmanaging own training contents orcontents provided by other producers [3].3 E-learning standardsImportance and need of specifications andstandards are well known to all of us indifferent areas of activity. Standards imposecertain order providing more uniform andprecise access and manipulation to e-learningresources and data. There are number oforganizationsworkingtodevelopspecifications and standards such as: ADL,IMS, ARIADNE, IEEE, ISO etc to provideframework for e-learning architectures, tofacilitate interoperability, content packaging,content management, Learning ObjectMetadata, course sequencing and many more[4].The ADL (Advanced Distributed Learning)initiative “is to provide access to the highestquality learning and performance aiding thatcan be tailored to individual needs, anddelivered cost effectively at the right time andat the right place” [5]. The ADL isaccountable for the Sharable Content ObjectReference Model (SCORM), a well-knownand accepted standard for all users of elearning platforms. This standard consists ofthree separate specifications: Content Aggregation Model (CAM) forassembling, labeling, and packaging oflearning content. The basic units ofinterest in the Content Aggregation Modelare Sharable Content Objects (SCO) andContent Packages that are used to delivercontent Run-Time Environment (RTE) whichincludes Launch (describes how a LMSprovides Content Packages to the tioninterfacebetweenContent Packages and LMS duringexecution) and Data Model (LMS recordsthe result of interaction between learnerand learning object using data model). Sequencing and Navigation (SN) forsequencing and content navigation. Thismodule controls and monitors theinteraction between users and LMS. TheseInformatica Economică vol. 16, no. 1/2012specifications are based on IMSConsortium specifications.Instructional Managements Systems (IMS)Global Learning Consortium is a consortiumof e-learning solutions providers. Thestandard IMS focuses on the development ofXML-based specifications. Several IMSspecifications have become worldwidestandards for delivering learning productsand services: IMS Content Packaging specificationdescribes data structures that can be usedto exchange data between systems thatwish to import, export, aggregate, anddisaggregate packages of content [6]; IMS Learning Design specification allowsa wide range of teaching techniques inonline learning; IMS Meta-data specification describes alearning object and allows to specify anannotation to search these educationalresources efficiently; IMS Question and Test Interoperabilitydescribes a standard data model forrepresenting the test items and reportsevaluation results; IMS Learner Information Package is acollection of information about the learner(individual or group learners) or theproducer of learning content (teachers orproviders); IMS ePortfolio specification was createdto make ePortfolios interoperable acrossdifferent systems and institutions.Alliance of Remote Instructional Authoringand Distribution Networks for Europe(ARIADNE) has created a standards-basedtechnology infrastructure that allows thepublication and management of digitallearning resources in an open and scalableway. ARIADNE aims to provide flexible,effective and efficient access to large-scaleeducational collections in a way that goesbeyond what typical search enginesprovide[7].IEEE Learning Technology StandardsCommittee (LTSC) “is chartered by the IEEEComputer Society Standards Activity Boardto develop accredited technical standards,recommended practices, and guides for

Informatica Economică vol. 16, no. 1/2012learning technology [8].” The IEEE/LTSC isorganized in working groups to developdifferent aspects of learning technology.International Standardization Organization(ISO). A subcommittee of the worldwideoperating standardization body ISO, theJTC1/SC 36 committee, is working onstandardization issues in informationtechnology for learning, education andtraining in liaison with the IEEE/LTSC. TheISO/JTC1/SC36 committee is organized ology;learnerinformation; management and delivery oflearning, education, and training; qualityassurance and descriptive frameworks [9].I would also like to propose severalspecifications for the quality of e-learningcontent (Learning Object, LO):1. LO objectives – at the beginning of eachLO teacher should clearly define theobjectives, so the students should beaware of what they learn.2. LO should be designed by level ofdifficulty – the students have not the samelevel of understanding, therefore teachersshould design LO by level of difficulty(very advanced, advanced, average,beginner).3. LO should be completed within a certaintime (i.e. from 5 to 15 minutes) – thecontent of the LO should be limited to acertain period of time so students do notget bored.4. Glossary – new terms should have a briefexplanation in the glossary of each LO5. Recapitulation and summary – at thebeginning of each LO should be apresentation (recapitulation) of theconcepts that should be known for a betterunderstanding of the new content. At theend of the LO should be a summary of thelearning content. Student may choosewhether to read the entire content of theLO or just the summary.1576. Detailed feedback on learning progress student should review certain chapters,paragraphs, etc.; teacher should highlightthe positive aspects; student should accessexternal links for more information.4 Quality criteria listThe growing number of available e-learningsystems and the commercialization of thesesystems highlight the necessity of qualityevaluations of online published learningmaterials. Although quality evaluation oflearning materials in e-learning systems havebecome increasingly important, the actualevaluation standards and methods forinformation quality (IQ) in such systemshave not yet reached a consensus [10]. Theevaluation of e-learning systems is importantfor all the actors involved in the learningprocess. Teachers and students need toevaluate the benefits of using e-learning incomparison with the classical methods oflearning [11].Evaluation of e-learning platforms requiresevaluating not only the implementingsoftware package (Learning ManagementSystem), but also the e-learning content(Learning Object). Both pedagogical andtechnological aspects must be carefullyevaluated. The following quality criteriawere developed based on the e-learningstandards (i.e. Scorm, Learning ObjectMetadata, IMS Specifications, etc.).I outline below six basic categories for theevaluation of the Learning boration,accessibility/effectiveness,management of e-learning content and users,administration, tools and technology) andothers six categories for the evaluation of theLearning Objects (didactic and pedagogicalevaluation, metadata, content evaluation,multimedia presentation, evaluation of theusers, technology).

158Informatica Economică vol. 16, no. 1/2012Table 1. Quality Criteria ListLearning Management SystemLearning Object (LO)A. FunctionalityA. Didactic and Pedagogical EvaluationA.1 Sequencing and NavigationA.1 LO should be design on different levelsStructureof difficulty (very advanced, advanced,average, beginner)A.1.1 ParagraphsA.2 LO for different learning profileA.1.2 MenusA.3 LO should be completed within a certaintime (i.e. from 5 to 15 minutes)A.1.3 External LinksA.4 LO objectivesA.1.4 SitemapA.5 Recapitulation LOA.1.5 Search EngineA.6 Summary LOA.1.6 Smart NavigationB. Communication/CollaborationB. Learning Object Metadata [12]B.1 EmailB.1 General (i.e. title, description, keyword)B.2 ForumB.2 Life Cycle (i.e. version, status)B.3 ChatB.3 Meta-Metadata (i.e. identifier, metadataschema)B.4 Web-blogB.4 Technical (i.e. format, size, location)B.5 WikiB.5 Educational (i.e. interactivity type,learning resource type, interactivity level)B.6 WhiteboardB.6 Rights (i.e. cost, copyright, description)B.7 Relation (i.e. kind, resource)B.8 Annotation (i.e. entity, date, description)B.9 Classification (i.e. purpose, description,keyword)C. Accessibility/EffectivenessC. Evaluation of the LO contentC.1 Access Status (free, payment,C.1 Free-of-errormixed)C.2 Multilingual ContentC.2 RelevanceC.3 Compliance to W3CWAIC.3 AccessibilityStandardsC.4 Plug-ins neededC.4 Credibility/ValidityC.5 Users feedback for evaluation ofC.5 Updatede-learning platformC.6 Easy of manipulationD. Management of e-learning contentD. Multimedia presentationand usersD.1 Progress report for usersD.1 Balance between textual and visualelementsD.2 Grade bookD.2 Attractive content presentationD.3 Progress report for Learning D.3 Entertainment games

Informatica Economică vol. 16, no. 1/2012Learning Management SystemObjectD.4 Export reports (i.e. Excel, PDF)E. AdministrationE.1 User registrationE.1.1 StudentsE.1.2 TeachersE.1.3 AdministratorE.1.4 Other users (i.e. parents)E.2 Templates for different userinterfaceE.3 System settingsE.4 Management of user groupsE.5 Backup SystemE.6 System MaintenanceE.7 Other modulesF. Tools and TechnologyF.1 The e-learning platform can beaccess by a standard browser (thebrowser displays all the multimediacontent)F.2 Friendly user interfaceF.3 DownloadinformationspeedF.4 Technical characteristicsoflarge159Learning Object (LO)D.4 Educational gamesE. LO for evaluationE.1 Different items for evaluation (i.e.multiple choice, true/false, free text, emptyspaces, drag and drop-matches)E.2 Initial evaluation (before the learningprocess)E.3 Final evaluation (at the end of thelearning process)E.4 Feedback on learning progressE.4.1 Students should review certainchapters, paragraphs, etc.E.4.2 Teachers should highlight the positiveaspectsE.4.3 Students should access external linksfor more informationF. LO TechnologyF.1 Reusability - a single LO may be used inmultiple contexts for multiple purposesF.2 Interoperability - LO may be used bydifferent e-learning platformsF.3 LO can be aggregated – LO can begrouped into larger collections of content,including traditional course structuresF.4 LO are self-contained – each LO can betaken independently5 The mathematical model used for theevaluation of e-learning platformsThe evaluation process consisted of thefollowing steps: Construction of the sample (samplerequirements, model performance, modeldevelopment); Fine classing and univariate analysis ofdata; Multivariate analysis – linear regressionand logistic regression; Correlation analysis; Validation of the model.5.1 Construction of the sampleVariable whose value I wish to predict iscalled the criterion or the dependent variableand the variable whose value is used topredict the criterion is called the predictor or

Informatica Economică vol. 16, no. 1/2012160the independent variable. In this case, thecriterion variable is: using an e-learningplatform to meet certain quality criteria isenough for better understanding, learningand assessment knowledge and the predictorvariables are the quality criteria list(described in table 1) and socio-demographiccharacteristics of the student.I used a survey to identify the training needsof the users. Example of question in thesurvey: Do you consider the user’s feedbackGroupGoodBadIndeterminateimportant for the evaluation of an e-learningplatforms ?Users may answer: Yes, I agree; No, I disagree; I don’t know.I say they are ‘good’ those who answer yes, Iagree, ‘bad’ those who answer no, Idisagree, and ‘indeterminate’ for those whoare undecided. The goal is to build a modelto discriminate between good and bad.Table 2. GB classificationDefinitionYes, I agreeNo, I disagreeOther responseSample requirements: Quite recently, in order to resemble with areal situation; Representative for the target population; To contain a sufficient number of bad, aminimum of 4%Model performance: the event to bepredicted is the probability that an user’sanswer is good. It is necessary to exclude allthose undecided, for a good discriminationbetween good and bad.Development and Hold-out sample: Thedatabase will be divided into two, respectingthe original proportions (weights 70% - 30%or 80% - 20%): The base development, used for the modeldevelopment; The base used for the validation of themodel.intervals that undergo analysis and fromwhich inferences can be drawn about theimportance of a characteristic in thedevelopment.There are many methods to determine anoptimum number of intervals (e.g. Sturgesmethod), but I consider enough that eachinterval to contain about 5% - 10 % from thebase. Non-numeric variables will be groupedseparately and analyzed in the same manner(e.g. gender, year of study, job, etc.). Thepurpose of the univariate analysis is toidentify all the variables that can beconsidered as suitable predictors of theprobability of a student being Good.I calculate WoE (Weight of Evidence) whichindicates that it is necessary to groupmultiple ranges into one.5.2 Fine classing and univariate analysis ofdataConsists in amalgamating observations into aset of ranges or intervals to produce statistics(e.g. good/bad odds) that could not beproduced for individual observations (as oneobservation is either good or bad). It is theseA method of excluding variable that is notrepresentative is given by Information Value,IV.IV (% good %bad ) *WOEk k % good WOE ln %bad krepresents number of groups.

Informatica Economică vol. 16, no. 1/2012Power of explanationLowMediumGoodVery good161Table 3. Measures of explanatory powerInformation ValueGini Index 0.02 10%0.02 to 0.110% to 20%0.1 to 120% to 30% 1 30%Gini Index is calculated by comparing thecumulative number of goods and bads byscore. Graphically, it is the area between thetwo lines on the curve (XYW) expressed as apercentage of the maximum possible (XYZ).The two axes on the graph are cumulativepercentage of goods (y-axis) and cumulativepercentage of bads (x-axis).YCum %GoodsWZXCum % BadsFig. 1. Gini IndexGini Index is calculated as follow: gi cumulative percentage of good at agiven score; bi cumulative percentage of bad at agiven score; Sn the n-th score in the scoredistribution.Using simple geometry, the area under thecurve for a given score is defined as:1Ascore (bi bi 1) * ( gi g i 1)2The total area of (XYZ) minus the total areaof (XYW) is:Ag Sn Ai S2iThe area of triangle (XYZ) is equal to:1AT (100 *100) 5,0002The Gini coefficient is then calculated as themodulus of:g ( AT Ag )ATThe result is between 0 and 1, as aproportion.The Information Value measure is calculatedas follows:n g .B g b I i i . log i i 1 G B bi .G where G and B are the total number of goodand bad respectively5.3 Multivariate analysis – linearregression and logistic regressionGeneralizing, the term Regression is used tocharacterize the way in which themeasurement of an unobserved (ordependent) variable Y changes according tothe measurements of one or more differentevents (the independent variables xi, i 1, 2, ). The purpose of a regression analysis is

Informatica Economică vol. 16, no. 1/2012162to quantify the relationship between thedependent and independent variables.Linear regression: in linear regression theobjective is to find an equation that links thelatter to the former through a linear function:Y 0 1 X 1 . n X n iThe coefficients i represent the weights toapply to the value of the independentvariables to estimate the dependent variableY; the term i is the error term, the differencebetween the actual and the predicted valuesof Y. The coefficients are determined so as tominimize the sum of the squared errors i(Ordinary Least Squares criterion), but thereare some other robust methods in presence ofoutliers in data.Logistic regression – in logistic regressionthe unobserved variable Y is a Bernoullianrandom variable whose possible values are 0and 1. The probability that Y can assume thevalue 1 depends on the regressors setxi (i 1,2,., n) :P(Y 1 X x) ( x),(1)The procedure for estimating such aprobability is based on the comparison (oddsratio) between the probability of an eventhappening and the probability that it does nothappen:odds( x) P(Y 1 X x)P(Y 1 X x) ( x) ,(2)P(Y 0 X x) 1 P(Y 1 X x) 1 ( x)The natural logarithm of the odds (logit) is alinear function of the regressors xi:ln odds( x) 0 1 x1 2 x2 . n xn ,(3)Combining formulas (2) and (3) and sol

Informatica Economică vol. 16, no. 1/2012 155 Evaluation of E-learning Platforms: a Case Study Cristina POP Academy of Economic Studies, Bucharest, Romania cristinel19@yahoo.com In the recent past, a great number of e-learning platforms have been introduced on the market showing different characteristics and services.

Related Documents:

Working on digital labour platforms 7 Over the past decade, labour platforms - digital platforms that connect workers with work - have emerged as a new trend in the world of work. Connecting predominately self-employed workers with clients in need of services on an on-demand basis, platforms have proved capable of transforming how, when

employed worker (Lozano and Reilly, 2018). Casilli points out that the concept of digital labour addresses four areas that give rise to a typology of analysis: on-demand platforms, microwork platforms, online social platforms and smart platforms.3 In the case of on-demand platforms, they

Evaluation The purpose of evaluation in the context of learning is to develop understanding carried out by educators. The learning evaluation objectives are learning activities that include teaching objectives, dynamic elements of education, implementation of learning, and curriculum [16]. In the evaluation of learning, several procedures must

tion rate, evaluation use accuracy, evaluation use frequency, and evaluation contribution. Among them, the analysis of evaluation and classification indicators mainly adopts the induction method. Based on the converted English learning interest points, the evaluation used by the subjects is deduced for classification, and the evaluation list .

POINT METHOD OF JOB EVALUATION -- 2 6 3 Bergmann, T. J., and Scarpello, V. G. (2001). Point schedule to method of job evaluation. In Compensation decision '. This is one making. New York, NY: Harcourt. f dollar . ' POINT METHOD OF JOB EVALUATION In the point method (also called point factor) of job evaluation, the organizationFile Size: 575KBPage Count: 12Explore further4 Different Types of Job Evaluation Methods - Workologyworkology.comPoint Method Job Evaluation Example Work - Chron.comwork.chron.comSAMPLE APPLICATION SCORING MATRIXwww.talent.wisc.eduSix Steps to Conducting a Job Analysis - OPM.govwww.opm.govJob Evaluation: Point Method - HR-Guidewww.hr-guide.comRecommended to you b

Section 2 Evaluation Essentials covers the nuts and bolts of 'how to do' evaluation including evaluation stages, evaluation questions, and a range of evaluation methods. Section 3 Evaluation Frameworks and Logic Models introduces logic models and how these form an integral part of the approach to planning and evaluation. It also

Today: Model Selection, Evaluation, Learning from Imbalanced Data Reinforcement Learning Ensemble methods (e.g., boosting) Learning with time series data Learning with limited supervision, other practical aspects (e.g., debugging ML algorithms) Intro to Machine Learning (CS771A) Model Selection, Evaluation Metrics, Learning from Imbalanced Data 2

BAB 6 LEMBAGA JASA KEUANGAN DALAM PEREKONOMIAN INDONESIA KOMPETENSI INTI 3. Memahami, menerapkan, menganalisis pengetahuan faktual, konseptual, prosedural berdasarkan rasa ingin tahunya tentang ilmu pengetahuan, teknologi, seni, budaya, dan humaniora dengan wawasan kemanusiaan, kebangsaan, bakat dan minatnya untuk memecahkan masalah kenegaraan, dan peradaban terkait penyebab fenomena dan .