Object-oriented Instructional Design and Applications to the Web

Authoring and delivery environments for Web-based instruction should be based on information and knowledge constructs and not on the manipulation of data. This higher level approach can be realized using the object-oriented paradigm, not just as applied to software but also to instructional design. This note summarizes our work in this direction and points out how some classical problems with hyperlinks can be analyzed from this viewpoint. We also look at extensibility, an important feature lacking in existing on-line authoring and design systems, and illustrate how an object oriented approach can support this as well. Hyperlinks, Data, and Information When thinking about pedagogy it is useful to distinguish among data, information, and knowledge. In (Robson 1999) we illustrate these epistemological categories by looking at tabular data that is to be included in a document or lecture. The main point is that forcing an author to enter data into table cells via a GUI fixes the form in which the data is displayed. This makes it impossible to specify context sensitive rendering methods and thereby inhibits the ability of an author to think in terms of connections among ideas. Instead the author is forced to spend time on the format used to display data. A table or graph is a means to convert data into information and the problem we encounter is that the authoring system confuses the two. Another fundamental example of this same phenomenon is the notion of a hyperlink itself. It has been pointed out by that standard hyperlinks carry with them the same flaws that GO TO statements carry in programming languages. To quote from (Lennon & Maurer 1996), There is certainly no significant data model for WWW, and in many ways first generation systems may rightly be likened to novice programs that used too many GO TO statements. Manifestations of these flaws are the prevalence of broken links, the inability of users to annotate documents, and the difficulty of creating symbolic links by traditional means. The computer science analysis of the hyperlink problem is that hyperlinks should be object-oriented instead of procedural in nature. This approach is taken by Kappe, Maurer, and Tomek in their work on the Hyperwave system developed at the Institute for Information processing and Computer supported new Media in Graz, Austria (Maurer 1999). Whereas we agree with the principles behind the Hyperwave solution, from our perspective the underlying problem lies in the confusion of data with information. In the minds of authors, hyperlinks are often symbolic and informational. For example, an instructor might think about inserting a link to a brief biography of Linus Pauling into a Web page discussing Vitamin C. The instructor may or may not have a specific biography in mind. The correct way to insert such a link is symbolic. The URL should be “link that displays a brief biography of Linus Pauling” rather than something like http://bio.linus.pauling.edu. By hard-coding the URL of a link, information is replaced by its instantiation and becomes data. This is not what we want. It is interesting to note that Vannevar Bush's memex machine (Bush 1945), which is considered to be a visionary description of the Web made long before its invention (Berners-Lee, 1995), is a good example of turning data (the pages of the machine) into information (trails through the pages) and knowledge. Below is a quote from the memex article that describes the use of the machine. Note how the owner's goal is described using the works "interest" and "study", which implies a quest for knowledge, and that this knowledge is constructed by finding data (books and articles) that clearly represent information and linking this information. The owner of the memex, let us say, is interested in the origin and properties of the bow and arrow. Specifically he is studying why the short Turkish bow was apparently superior to the English long bow in skirmishes of the Crusades. He has dozens of possibly pertinent books and articles in his memex. First he runs through an encyclopedia, finds an interesting but sketchy article, leaves it projected. Next, in a history, he finds another pertinent item, and ties the two together. Thus he goes, building a trail of many items. Occasionally he inserts a comment of his own, either linking it into the main trail or joining it by a side trail to a particular item. When it becomes evident that the elastic properties of available materials had a great deal to do with the bow, he branches off on a side trail which takes him through textbooks on elasticity and tables of physical constants. He inserts a page of longhand analysis of his own. Thus he builds a trail of his interest through the maze of materials available to him. We observe in (Robson 1999) that learning is the process of acquiring or constructing knowledge and that many components of teaching deal primarily with information. As in the Vannevar Bush description, a key component is the forming of cognitive relationships among pieces of information. Yet the vast majority of current on-line educational resources, the current versions of memex, operate strictly in the data domain. This has many consequences. One is that too much effort is spent on form and format and too little on pedagogic substance. This contributes to the large cost of developing Web courses or course components at Universities (Arenson 1998 and Harapnuik et.al. 1998). Another consequence is that authors and designers cannot interact with learning environments in a language built around concepts of teaching and learning. This is typified by the need to make hard links instead of symbolic cognitive links. But the worst effect is that teachers and authors are using technology that dictates and limits on-line pedagogic methods instead of using technology which both serves existing pedagogy and encourages the development of new instructional strategies. We will discuss that now. The Limits of Commercial Environments. Confusing data with information can make more work for the author, but the real problem with the functional nature of current authoring environments is the lack of extensibility. Towards the end of Bush's memex article (Bush, 1945) he speculates that Wholly new forms of encyclopedias will appear, ready made with a mesh of associative trails running through them, ready to be dropped into the memex and there amplified. The Web reflects this in its ability to integrate new sites, but the tools that exploit the Web for pedagogic purposes are limited to the functionality that the creators envisioned. To illustrate this, suppose that we are using a commercial course management system that offers all of the standard amenities: user types, quizzes, navigation, record-keeping, chat, email, and so on. Suppose further that we would like to use this on-line environment to do some new things, for example to (A) display student progress towards a learning objective as a bar chart, (B) pair students and have them exchange papers and critique each others papers, or (C) have different students see different examples on the same Web page depending on their scores on a particular quiz. These are all things that have been done and that represent sound pedagogy. (A) was suggested to the author by a high school guidance counselor and was implemented in a mathematics Web site. The implementation uses nothing more complicated than an image tag with the width parameter read in from a database. (B) is something we have done using QuestWriter software developed at the author’s home institution (Bogley et. al. 1996) In our implementation, the computer takes care of all of the management including pairing off students, delivering electronic versions of papers and comments, and sending email reminders when a student falls behind on an assignment. (C) is an example of adaptive pages. Adaptive pages have been implemented in a number of academically developed environments (see, for example, (Interbook 1999) and the work referenced there). Despite all of this, the creators commercial environments have not included any of the functionality represented by (A), (B), or (C). It might be possible to bend the environments to implement the associated pedagogical ideas, but it would be very hard to do so neatly and impossible to truly add new functionality. An Object-oriented Model If our goal is to create an environment that operates on a higher level than data manipulation and that permits users or designers to add new functionality, then we need a new underlying model. In (Robson, 1999) we propose such a model. We start with an object-oriented approach to instructional design that is independent of the medium in which it is realized but works very well for the Web. At its core is the notion of a learning object. Examples of learning objects include things like "explanations", "assessments", "lectures", and "applets". Anything that an instructor or student might use as a path to learning is a candidate for a learning object. In our model the designer (or author) defines learning objects and manipulates their methods and properties. Methods might include ways of rendering an object in various media, and properties might include other objects. An important property of an object is a relationship to another object, often rendered as a hyperlink in the Web context. This allows us to incorporate the symbolic and cognitive notions of links discussed earlier in this note. In (Robson, 1998) we illustrate how this object-oriented model can be implemented in a primitive way. We discuss a few Web sites with ad-hoc authoring environments built using a simple database and Web integration software (Robson & Whitesitt 1999). In these environments the author can associate an external link to the content of a page, or a learning objective to a lecture, by using a pull-down list of available links. Although this is just a first step, it brings out the underlying philosophy that the role of the author is to specify relationships among objects and not to worry about how those relations will be rendered.