simulation
implementation of the web based testing
currently, the web based testing is implemented through five basic techniques or combinations of them. these techniques are html links, html/cgi forms, scripting language, plug-in, and java. each of the techniques has its strengths and limitations. the techniques are as under:
the easiest and the earliest technique is the implementation of testing through html pages. in this implementation, the question is presented on a web page. the possible responses are defined as active areas on the web page and these active areas are linked to other web pages. each of the linked web pages will provide evaluation of a particular response, associated feedback, and the further link to a new instructional material or a question. the linked page will also provide a method of computing the score and displaying it. the main advantage is the implementation simplicity and the facility to keep count of the score. the system does provide a higher level of security as the student has no prior idea of the result or the consequences of selection of a choice. it is possible to compute the score and display it for every choice. since the question is on a web page, each of the questions is encoded by the hypertext and since the feedback is a web page, it has to be anticipated and coded on the web page, and thus the feedback has to be predefined. however, the wbe requires a new page access and transfer for every student response, and this could be time consuming. since active areas are linked to the response web pages, it is best suited for click-selection oriented test items such as single answer multiple choice, true false, and pointing.
the second and the most well established method for web testing is the implementation of test questions through a combination of html forms and cgi-compliant evaluation scripts. it is currently being used in a number of commercial and university-grown systems and test development is relatively simple and can be done with the available html authoring tools. it derives its flexibility through the display of question in a form format through html and then process the information captured through form, using cgi scripts. the processing can be as simple as comparing a word to a complex database searches and associated processing.
the student receives the question in the form of html form and responds to it by filling appropriate spaces on the form. when the form is submitted, the information goes back to the server, where the cgi script extracts the necessary information, uses it as data for complex processing and returns the value to a web page that is to be communicated back to the student. thus a part of the message is dynamically generated depending upon the input, using server side programming.
the returned page may provide a means of computing the score and displaying it. the main advantage is the conceptual simplicity and the ability to handle a wider variety of questions. the system does provide a higher level of security as the student has no prior idea of the result or the consequences of selection of a choice. sensitive information that is required for test evaluation such as question parameters, answers, feedback may be safely stored on the server side preventing students from accessing it. the server-side evaluation makes all assessment time functions, such as recording results, grading, providing feedback etc. easy to implement. a single server-side evaluation script could perform all these functions. it is possible to compute the score and display it for every choice. since the feedback is generated as a web page, it is processed and thus is not be predefined.
the html forms – cgi script are useful for a variety of questions and often can be implemented in many ways. they are best suited where the student has to input data into specified fields as in the case of form. selecting a single answer in multiple choice, and selecting true/false are implemented through radio buttons. responding with a short answer like single word or multiple words, multi answer multiple choice questions, naming, direct numerical answer etc. are implemented by capturing data through the input fields provided on the form. advanced questions such as matching pairs or ordering can also be implemented using forms. in some cases check boxes are also useful. server side map technology can be used for questions involving pointing and dragging. it is possible to carefully frame the question and get replies for questions that need to specify a procedure or steps in problem solving. the hidden fields can be imaginatively used for many interesting purposes like explorations, communicating information to cgi script etc.
the system requires the return of the form, initiate cgi script, access the data, process data and transfer the web page back to the student with the result, feedback etc. this process could be time consuming. the cgi-based questions do not work when a user's connection to the server is broken or very slow. it is well suited for presenting basic types of tests. more advanced types of tests as well as more interactive types of tests such as drag-and-drop tests can not be implemented with pure sever-side technology. authoring questions with server-side evaluation is a bit difficult as the question and its evaluation is spread out between html presentation and a cgi evaluation script.
the third development is using javascript for test generation. since javascript is embedded into html document, the web browser runs it on client-side. there is also a server side version of javascript. the use of javascript for question delivery and evaluation provides features that are similar to the ones provided by form and cgi script. however, javascript being a programming language, supports more advanced, open ended, and interactive questions. unlike form and cgi script, javascript question works in standalone mode as the question text, information needed to evaluate the question and to provide feedback, and the evaluation program come bundled together as a single unit, when javascript is used. this provides ease in authoring the tests. when client-side javascript is used, in principle the students have access to all the information of the test and solutions.
using client-side javascript, it is not possible to record student grades. hence javascript would be useful to implement self-assessment tests than for grading tests. a combination of client side programming that provides the necessary flexibility in choosing a variety of interactive questions, and server side programming with javascript or cgi script that provides facility to evaluate, record, and provide feedback offers powerful combination. the client side execution has the advantage of using a high degree of interactivity in the questions and independence from the server.
the fourth technique is to use plug-in technology such as macromedia’s shockwave. using these tools, animations are run on the client machine. shockwave provides features that need to be exploited for animation based questions.
the fifth technique is to make use of java. since java is a complete programming language with platform independence, java combines the best features of form/cgi shockwave and javascript. it provides smooth communication between the client side application and server side application.
certain types of questions such as essays, full-fledged designs, case studies, problem solving etc. cannot be handled by any of the above techniques easily. in such situations the following two strategies are suggested. in the first case the system marks the questions that require the attention of the teachers and informs them. it corrects the rest of the questions. however, with large number of students, the process becomes very taxing for any teacher. in the second approach, the answer scripts of students are sent to a fixed number of peers along with an ideal answer prepared by the teacher. the peers evaluate the answer script with the help of the ideal answer script sent by the teacher. the answer script that is to be evaluated is initially assigned a score of 100 marks. for every missing point, for every wrong point, the student gets –10%. for any additional point written and pointed out by the peers, the student gets a +10%. the number of peers chosen may be between 3 to 5. each of the members of the peer groups will get 1% for every mistake or additional point they find. from correcting any single answer script, a student may get a maximum of 5%. however since every student corrects 3 to 5 answer scripts, the maximum marks a student can earn will be 15% to 25%.
the seventh and the last of the techniques is to use an intelligent domain expert on the server side with a cgi- like interface. this provides the necessary intelligence to the system. limited experts could be built on server side using java.
conclusions
the success of wbe, as an alternative educational system, depends upon the interaction and the evaluation that is built into the learning process. the evaluation system has to be designed looking at the learning and certification needs at different stages of the learning process. the evaluation system has to respond to the challenge by using a wide range of test items as each type of test item has a specific role to play. the current technologies that are available for testing have adequate capability to support the learning process and are to be harnessed.
references:
1. bloom, b.s. (ed.). (1956). taxonomy of educational objectives. handbook: cognitive domain. new york: mckay
2. skinner, b.f. ( 1968). the technology of teaching. new york: appleton-century-crofts.
3. gagne, r.m. (1985). the conditions of learning ( 4th ed.). new york: holt, rinehart and winston.
4. murray w. goldberg and sasan salari(1997) an update on webct (world-wide-web course tools) - a tool for the creation of sophisticated web-based learning environments , proceedings of nauweb '97 - current practices in web-based course development.
5. meena kumari (1998): “ instructional design practices” , csi news letter
6. meena kumari et al. (1999): “ making technology based learning effective: a unique online learning architecture” communicated to the common
wealth learning conference and to be presented in west indies in nov 99.
