Advanced Computer Based Education on the World Wide Web

Patrick W. Brewer
Elizabeth J. Gibson
Ajay Dholakia
Mladen A. Vouk
Donald L. Bitzer
Dept. of Computer Science
North Carolina State University
Raleigh, NC 27695-8206.

Abstract

With the wide spread use and support of the World Wide Web (WWW), many people have begun to look at the WWW as a new tool for education. While the WWW has made a great leap in providing a mechanism for the dissemination of information, it is not designed to be an educational tool in the classical sense. In this paper, we discuss many features that an advanced educational environment requires, and we examine how these features relate to the capabilities of the WWW tools. In an advanced learning environment, one of the most important components is the ability of the environment to track individual students and adapt to their knowledge-transfer needs. Therefore, we pay particular attention to how this tracking can be adequately implemented in the WWW environment.

Keywords

WWW, Education, Computer-Based Teaching, Computer-Based Training, Computer-Based Learning, Tracking

1. Introduction

The current methods for education and training usually involve direct instructor presentation of materials at set times. Although a more flexible computer-based education (CBE) or computer-assisted instruction (CAI) has been an option for many years [Alpert70, Anderson85, Brandon88], it is the advent of the World Wide Web (WWW) and the initial deployment of the National Information Infrastructure (including the leading-edge North Carolina Information Highway - NCIH), that re-kindled a widespread interest in network-based education (NBE). NBE takes many forms and many issues need to be explored in evaluating a distance education system [Horn94].

Currently there is a flood of "educational" material appearing on the WWW. Unfortunately, a common problem is to confuse the browsing-oriented library-type approach with teaching. In a browsing environment, course material is made available over the network, but the user can expect little or no interactive feedback or help with the material. A better alternative is a "virtual workshop" approach where limited on-demand user-initiated help may be available, but the evaluation of user progress and encouragement of group interaction. However in this case, feedback is either non-existent or very limited. Even better is to implement "virtual classroom" systems designed to provide: 1) a method for interactive and adaptive query, response and tutoring, and 2) a scheme for tracking and measuring learning and providing progress feedback. Very few systems like that are readily available. An example of a non-WWW operational system of this type is NovaNET [Alpert70, Jones92, Jones83]. To the best of our knowledge, there is no WWW-based system that falls into this category. However, the WWW is being actively researched as a means for providing on-line distance education [Butts94], [Hurley94], [Perron94], [Dwyer95], [Campbell95], [Ibrahim95].

In order to have a good understanding of what an advanced educational environment requires, one should turn to the existing research on computer-assisted instruction (CAI). An excellent treatise on the foundations of designing CAI can be found in [Steinberg91]. Designers of CAI draw on research from many disciplines and on practical experience. Steinberg makes an important point that computers are a medium for presenting instruction rather than a method for instruction. Thus, CAI can draw on already known techniques of traditional instructional design theory. Some guidelines for developing CBE software are given in [Overbaugh94], and evaluations of instructional software are available in, for example, [Gros94], [Reiser94], and [Wulfekuhle94].

Good CBE courseware requires the following WWW support.

Proactive Environment: Browsing tools tend to be designed as relatively passive devices where most, if not all, interactions must be initiated and guided by the user. This raises difficulties in implementing system-initiated, active knowledge testing and feedback required in all good CAI systems [Calriana93].
Student Identification: It is very likely that a student interacts with a CBE systems over several sessions. For example, any non-trivial course is likely to consist of a number of lessons that need to be taken over a period of time. In order for the CBE system to provide meaningful guidance and instruction to an individual student, the system must be able to keep a history of that student's actions and progress. Therefore, the system must be able to identify the student in order to keep a correct record of that person's history.
Syncronization: Multimedia features require synchronization of audio, animation, and video components with each other and with text. One may include audio or video in a WWW page and expect the student to click on it at the appropriate time, and then wait until the audio or video segment is finished. However, research has shown that the student behavior will typically not be as expected [Jones94], and a CBE system must have the facility for imposing certain actions.
Quality of Service (QOS): A CBE system must provide a certain level of performance to be useful. For instance, experience shows that consistent response delays in excess of 100 ms are detrimental to student learning. Furthermore, research shows that once a student starts a one hour lesson, the probability of getting through that lesson without any problems (including network delays and content retrieval problems) should be above 0.95 [Rindos95]. Throughput and jitter are two other QOS parameters that need to be addressed.

In this paper, we discuss the issues related to the implementation of advanced educational features in the context of the WWW. In Section 2, we provide a brief overview of the educational underpinnings. In Section 3, we discuss the desired CBE features in detail. Section 4 focuses on the tracking, and the conclusions are given in Section 5.

2. Educational Basis

The use of the computer as an effective medium for delivery of courseware is forcing a rethinking of the entire course development process. Roblyer [Roblyer88] views this process as comprising three stages: instructional design, preprogramming development, and programming. He suggests that instructional method and lesson content must be fully developed before the actual programming takes place.

Rowntree [Rowntree88] proposes a systems approach to the development of courseware which involves the following steps: identify course aims and objectives, develop necessary learning experiences, evaluate the effectiveness of learning experiences, and improve the experiences in the light of evaluation. Thus, Rowntree's approach is iterative as opposed to the linear approach suggested by Roblyer. It is clear that instructional design is an important step, and this is particularly the case in CBE. In fact, since courseware can be considered as a very high level language program, courseware developers would be well-advised to draw on lessons from software engineering.

The educational foundations of instructional design are very clearly and succinctly summarized by Gagne in his nine events of learning : gaining attention, informing the learner of the lesson objective(s), stimulating recall of prior learning, presenting stimuli with distinctive features, providing learning guidance, eliciting performance, providing feedback, assessing performance, and enhancing retention and learning transfer[Gagne79]. Focusing on instructional design for computer-based instruction, Overbaugh [Overbaugh94] proposes four domains in which he groups Gagne's nine events: instructional set, teaching strategies, student performance, and issues specific to computer-based instruction. Campbell et al [Campbell95] propose the use of Gagne's nine events for instructional evaluation of courseware material that already exists in the form of lecture notes for regular classes. They also propose an organizational level evaluation as per the ASSURE model: Analyze learners, State objectives, Select media and materials, Utilize media and materials, Require learner participation, and Evaluate and revise.

A common element of all of these models, and the one we will focus on in this paper, is active learner participation via feedback and tracking. Overbaugh [Overbaugh94] stresses the importance of feedback and guidance in CAI software. Overbaugh states that "The way the feedback is presented facilitates or debilitates further learning". In order for a system to provide more than basic feedback and guidance, it must first gather input from the user on which to base the feedback. The need to gather input and provide feedback requires the educational system to be remarkably different from the passive browsing environment that is the usual WWW environment. It was noted by Ibrahim and Franklin that this need conflicts with the basic design of the WWW [Ibrahim95].

3. Mapping CBE onto WWW

We organize the issues involved in the mapping between CBE features and WWW capabilities into three categories: issues with solutions, issues for which no ready solutions are suggested, and issues which only indirectly relate to the WWW.

Issues and Solutions

Since the WWW was designed to be a reactive system, the WWW tools themselves take no action, they only react to user inputs. This is appropriate to the type of work for which the WWW tools were designed. However, in educational courseware, one may require more control. Some examples are trying to provide animations of a process to facilitate better understanding of that process, or putting time limits on answering test questions.

Related to the proactive issue is the ability to control when text (and/or alternative media) is displayed to the user, for emphasis of a point. This concept is like that of an instructor hiding part of an overhead slide until he/she begins to speak about the hidden topic. This is possible in presentation software.

To provide guidance to students as suggested, a CBE system needs to track data on which it is to base its guidance decisions. The WWW's statelessness is a major issue in trying to add advanced educational features, such as guidance, to material on the WWW. Ibrahim and Franklin suggest one solution to this problem [Ibrahim95]. Later, this paper describes another solution to the same problem, based on the security features of http servers.

The WWW enables easy hyper-media document construction. Unfortunately, using the multi-media aspect of the WWW brings forth the compatibility issue. Using digital-audio recordings to explain or enhance points in a lesson is desirable for CBE. However, while the WWW text is accessible to major types of desktop computers, the audio formats different hardware will handle is not standard.

Issues Without Readily Implementable Solutions

These issues are synchronization and caching. As stated, the synchronization issue is partially related to the reactive nature of the WWW. There are many types of synchronization one can discuss when referring to a multimedia environment. Here, the issue addressed is one of ensuring that audio and/or video segments are run at appropriate times. In a case study dealing with students in a distance learning class without access to local instructors, Jones and Petre found that "students are likely to approach instructional materials in unexpected, even haphazard ways . . ." [Jones94]. Synchronization is an important issue in education because the alternative media is being used to convey a concept to the student. If the media segment is not played within the expected context, it may not make the intended point. Worse, it may actually lead to misunderstandings or learning wrong concepts. In many situations, the widespread use of features similar to those used in HotJava may clear up this issue.

Caching is helpful in improving the speed of the WWW access, and in lowering the internet traffic level. However, caching can create a problem with educational courseware on the WWW. The potential problem arises when trying to track a student's progress through a lesson. Local caching by the browser causes only the first access to a page to be logged at the server site. The server will not record any further user access to that page while the page is in the browser's cache.

A simple solution is to turn off caching while using an educational site which performs tracking. An even better solution is a form of smart caching to cache some objects and not cache others. For example, an instructor might want to know how many times a student jumps to a glossary as a way to indicate if a student is having vocabulary problems. However, the instructor may not be interested in how often the student loads in-line images. This could be important for speed and traffic because images are likely to contain more bits than the text on the page. This "smart caching" will have to be implemented as a feature in the browser and possibly in the server (to allow the author to designate which objects are cacheable and which should not be cached).

Related Issues

The issues in this category impact the use of the WWW for educational purposes, but they are not directly controlled by the WWW community. They include page layout and design, interface standardization, certification of educational material, copyrights, quality of service, and so on. These issues although important, are not within the scope of this paper.

4. Tracking and Implementation

Using Security Features to Provide Tracking

Tracking is an important part of an advanced educational environment. It encompasses all activities that have a meaningful input in the evaluation of student knowledge acquisition and retention. Traditionally the WWW does not recognize any connection between two requests, it is effectively a stateless environment. This statelessness presents a problem in tracking individual student actions. In the past, some modifications were needed to the http server in order to maintain user identity (e.g., [Eval], [Ibrahim95]). However, server modifications carried the responsibility of supporting the code. Creating separate processes for each user could cause load problems on server machines. Fortunately, security concerns led to a more palatable solution to the problem. The current version of NCSA's http server supports security (access control) features which enable identification of users. In our experiments we have been using NCSA's Version 1.4.1 http server.

The NCSA server logs the file name and the machine (its IP address) for each request. But, this information is not detailed enough for one to extract the desired tracking data. However, when the access control features are configured for the site, the http server will also log the identity of the individual users. By analyzing the access log, it is possible to trace student actions. Protected CGI programs can identify users through the REMOTE_USER environment variable. Many people have recommended using fill-out forms in order to give on line quizs and to interact with the student. Using the above method, the data from the forms can be stored in the student's tracking history by CGI programs.

Platform Independent Audio

Using the audio to supplement the textual content of the lesson material is very attractive. The problem is that unless the students access the material from the same type of hardware platform used by the lesson author, there will be a compatibility issue.

Currently though, it may be possible to determine which type of audio format to send to a user by examining the environment variable HTTP_USER_AGENT. A CGI program can determine which browser a student is using, provided that the browser gives its platform information in this variable. It is then possible to decide which audio format to send to the user. In that context it is interesting to note that NCSA's Mosaic browser for X-windows version 2.5 does not include platform information in the HTTP_USER_AGENT variable, but it does make use of the HTTP_ACCEPT variable. On the other hand Netscape's Navigator version 1.1N provides the platform information in HTTP_USER_AGENT ( it returns Mozilla/1.1N (Macintosh; I; PPC) ), but it does not provide a detailed list in HTTP_ACCEPT. An added complication is that the server machine must keep multiple copies of the same audio saved in the different formats. A better solution would be a canonical audio format and/or dynamic transformation filters. This idea is expandable for use in making decisions of video-files formats which should be returned to the requesting browser.

5. Summary and Conclusion

In this paper we examined the characteristics that WWW needs to provide in order to support CBE. We believe that there is enormous potential for educational use of the WWW. But until the WWW readily provides support for proactive instruction, student identification, synchronization, and QOS guarantees, the WWW will not be able to compete with specialized network based CBE systems.

One of the most important features is the need to track individual student responses and accesses to the educational material. The reason is that only through detailed knowledge of students activities can the CBE system evaluate student knowledge and retention. Without that information, CBE cannot provide adequate feedback and cannot adapt its actions to the needs of the students.

We proposed several solutions and raised issues in the case of limitations that would require intervention at the level of WWW tools. The authors are currectly in the process of implementing a large scale networked based educational system that supports WWW based distance learning.

Acknowledgements

This work was supported in part through the NSF MRA award ACS-9418960. We thank Larry Lee and Ken Flurchick of MCNC-NCSC for their participation in discussions related to this work.

UNIX is a trademark of X/Open Company, Ltd. MS Windows is a trademark of the Microsoft company. Macintosh is a trademark of Apple Computers.

References

[Alpert70]: D. Alpert and D.L. Bitzer. Advances in computer-based education. Science, 167:1582-1590, 1970.
[Anderson85]: J.R. Anderson, C.F. Boyle, and B.J. Reiser. Intelligent tutoring systems. Science, 228:456-467, 1985.
[Brandon88]: P.R. Brandon. Recent developments in instructional hardware and software. Educational Technology, 28:7-12, 1988.
[Butts94]: D. Butts, C. Reilly, M. Speh, and J. Wang. {WWW} and the global network academy. Proc. First Intl. World-Wide Web Conf., 1994
[Campbell95]: J. K. Campbell, S. Hurley, S. B. Jones and N. M. Stephens, "Constructing educational courseware using NCSA Mosaic and the World-Wide Web", Proceedings of the Third International World-Wide Web Conference, Darmstadt, Germany, April 10-14, 1995 (http://www.igd.fhg.de/www/www95/papers/52/www3.html).
[Clariana93]: R. B. Clariana, "A Review of Multiple-Try Feedback in Traditional and Computer-Based Instruction",Journal of Computer-Based Instruction, Vol. 20, No. 3, Summer 1993, p. 67-74.
[Gagne79]: R. M. Gagne and L. J. Briggs, Principles of Instructional Design, Second Edition, New York: Holt, Rinehart and Winston, 1979.
[Gros94]: B. Gros and J.M. Spector. Evaluating automated instructionall design systems: A complex problem. Educational Technology, 34(5):37--46, 1994.
[Dwyer95]: D. Dwyer, K. Barbieri, and H.M. Doerr. Creating a virtual classroom for interactive education on the {Web}. Comp. Net. {ISDN} Sys., 27(4):897--904, 1995. Proc. Third Intl. World-Wide Web Conf.
[Hurley94]: D. Hurley, A.D. Marshall, S.N. McIntosh-Smith, and N.M. Stephens. Courseware for parallel computing using {Mosaic} and the {World Wide Web}. Proc. Second Intl. World-Wide Web Conf., 1994.
[Horn94]: D. Horn. Distance education: Is interactivity compromised? Performance and Instruction, 33(9):12-15, 1994.
[Ibrahim95]
Bertrand Ibrahim and Stephen D. Franklin, "Advanced educational uses of the World-Wide Web", Computer Networks and ISDN Systems, Vol. 27, No. 4, April 1995, p.871-877
[HotJava]
HotJava's homepage is http://java.sun.com/.
[Jones92]: L.M. Jones and D.J. Kane. Network use in central management of large university physics courses. J. Computer-Based Instruction, 19(3):77-81, 1992.
[Jones83]: L.M. Jones, D.J. Kane, B.A. Sherwood, and R.A. Avner. A final-exam comparison involving computer-based instruction. Am. J. Phys., 51(6):533-538, 1983.
[Jones94]: A. Jones and M. Petre. Computer-Based Practical Work at a Distance: A Case Study. Computers in Education, 22(1/2):27-37, 1994.
[Overbaugh94]: R. C. Overbaugh, "Research-Based Guidelines for Computer-Based Instruction Development", Journal of Research on Computing in Eduation, Vol. 27, No. 1, Fall 1994, p. 29-47.
[Perron94]: D. Perron. Learning on the {WWW}: A case study. Proc. Second Intl. World-Wide Web Conf., 1994.
[Server] Information about NCSA's Version 1.4 server can be found at http://hoohoo.ncsa.uiuc.edu/docs/Overview.html.
[Roblyer88]: M.D. Roblyer, "Fundamental problems and principles ofdesigning effective courseware", Instructional Designs for Microcomputer Courseware, D.H. Jonassen (ed.), Hillsdale, NJ: Lawrence Erlbaum Associates, 1988, p. 7-34.
[Reiser94]: R.A. Reiser and H.W. Kegelmann. Evaluating instructional software: A review and critique of current methods. Educational Technology Research and Development, 42(3):63--69, 1994.
[Rindos95]: A. Rindos, M. Vouk, S. Woolet, J. Hines, and J. Lester, "ATM Technology Enabling Educational Applications Across the North Carolina Information Highway," TELECOM 95, Geneva, Switzerland, October 1995.
[Rowntree82]: D. Rowntree, Educational Technology in Curriculum Development, London: Paul Chapman, 1982.
[Steinberg91]: E. R. Steinberg, Computer-Assisted Instruction: A Synthesis of Theory, Practice, and Technology, Hillsdale, NJ: Lawrence Erlbaum Associates, 1991.
[Wulfekuhle94]: N. Wulfekuhle. Selecting a hypermedia authoring program for {CBT}. T.H.E. Journal, 21(7):77--80, 1994.
[WWW]: Information about the World Wide Web can be found at http://www.w3.org/hypertext/WWW.TheProject.html.