The World is an Authoring System

by Steven Okonski

Originally Published in March 1996 in CBT Solutions Magazine

  • As Andy Rooney said, "Have you noticed authoring systems lately? They're springing up all over the place. There are authoring systems for graphics, the Internet, presentations, computer-based training, kiosks, advertising, performance support -- I could go on. Do we really need all these authoring systems? What are people doing with them?"


    Well, actually, Andy didn't say these things. But, he might be tempted to if he picked up a copy of almost any current computer magazine. In fact, his questions are quite valid. With the authoring system label being applied to most anything these days, how do you find one for CBT? That's what this article will help you do.


    The term "authoring system" came into use about fifteen years ago to describe "a software tool that assists in the creation of computer-based training." It took a while for the definition to spread: during that era, I recall talking with several people who (very logically) wanted to buy an authoring system to help them write a book. Apparently the definition of "word processor" also took time to spread. Even as recently as five years ago, "authoring system" generally meant something for CBT. Each year since then, the term has been more broadly applied. Recently I saw Visual dBase called "a database authoring system."

  • AUTHORING SYSTEMS ARE COOL

  • This is happening because, currently, being an authoring system is fashionable. It's cool to be an "author." You'll find "authoring system" in the IN column side of those what's- in/what's-out lists. So, the spin doctors prefix an adjective, and jump on the bandwagon: database authoring system, interactive multimedia authoring system, home page authoring system, and our veteran: CBT authoring system.


    There are differences between all these systems. While you might use Visual dBase to create a database, it would be a mistake to use it for CBT (well, you probably could, but I wouldn't want to be present during the unveiling to the client). So, how do you locate the systems suitable for CBT? To start, many of them are described and advertised right here. You might scan other magazines too, but none cover this market as well as CBT Solutions.


    For example, interactive multimedia (IM) authoring systems exist aplenty. Most IM systems can handle standard user input elements such as button clicks, key presses, and typed responses. IM systems are often suitable for creating projects such as business presentation slideshows, simple demo disks, or perhaps even kiosks. How do CBT projects differ? Like IM projects, CBT projects also obtain user input, but then they must study, scan, parse, analyze, judge, and record it. That means a CBT-capable authoring system needs features such as: test question generation, response judging (exact match, spelling allowance, phonetic match, word search, anticipated incorrect), automatic scoring, feedback, user log on and log off with password, built- in user bookmarks (so users can resume where they left off), user comments, multi-user network support, user rostering and class scheduling, Computer-Managed Instruction (CMI) data collection and analysis, and report generation. Whew!


    To illustrate the importance of these features, let's study an example of a fill-in-the-blank type question that appears in most CBT projects. Our sample question will ask the user to enter the name of the device that measures blood pressure. First, you, as author, could create the screen/page/frame that displays the question to the user. Depending on the authoring tool, you might employ menus, icons, objects, templates, flowcharts, or whatever. Since, in general, all authoring tools have a way to create the visual portion of such a screen/page/frame, the specific details of how are irrelevant for this discussion.

  • MAKING JUDGMENTS

  • What's more important is how to accept the user's response, and judge it (i.e., determine if it is correct). Now we're getting to the part where real CBT authoring systems can distinguish themselves from other types of authoring systems. A good CBT authoring system will let you, as author, specify the correct answer (sphygmomanometer) by simply typing it. There should be some kind of dialog box, property sheet, or whatever, in which you type the correct answer. You should not need to write any programming code whatsoever (such as an IF...THEN statement) to handle a common CBT operation such as this. At run time, the system should compare the user's response with the one you typed, and automatically worry about tedious details such as whether the user entered the response in upper- or lower-case characters, and with/without extra blank spaces. Several CBT authoring systems do this nicely; one that does not is not a real CBT authoring system.


    Let's make our example even more realistic. You might want to allow the user two extra tries at answering correctly, and allow some variance in spelling. You also want to keep score (perhaps this is the final exam). If the user's response is wrong, but in the ballpark (something like coagulometer or hemoglobinometer) you want to provide custom feedback that discusses why. Finally, in order to document the effectiveness of the training (and your budget!), you also want to permanently record the user's response to disk for later analysis. There are CBT authoring systems (not many though) that can do all these things without a single byte of programming. Oh, programming certainly has its place, but such common CBT operations as these should be built into the system so that your authoring productivity is maximized. Without them, you'll be writing lots of programming code, and it's likely the reading on your sphygmomanometer will rise!


    One way many authoring systems attempt to maximize your productivity is via templates. A template is a pre-built routine or page that you customize. For example, a page template might consist of an empty text area, an empty button and an empty picture container. You would simply fill in the text, put a caption on the button, and specify the name of the graphics file that contains the picture. The obvious advantage is that templates can save you a lot of time. One disadvantage is your project starts looking like the project created by the competition down the street, and vice versa. On the whole, templates are good, but like any good thing, they can be abused: some authoring tools now come packaged with thousands. Sometimes dozens of these templates are identical, except for a minor detail such as the color of the background! Sorting through so many templates nullifies the time savings; you might as well create the page from scratch. Plus, templates make authoring systems bulkier: some systems now hog more than 200 megabytes of disk space upon installation! The cost of hard disk space is declining, but it's not free. This is a subjective estimate: I'd say a good maximum number of pre-packaged templates is 100; any more than that is a negative. Perhaps the best solution is an authoring system that lets you take any page you created previously, and use it as the template for a new one. That gives you the best of both worlds: the time savings of templates, plus a way to make your project's appearance unique.


    All these built-in features and templates are ways that authoring systems attempt to minimize the amount of programming you need to do. Lest you begin to think that all programming is evil, let me say that some is necessary, if not downright advantageous, in almost all CBT projects. There is no way an authoring system developer can anticipate everything you are going to want to do in your project. There can never be enough built-in dialog boxes and templates and features to handle every instructional design style. This is the problem with icon-only type authoring tools. An authoring system without an embedded programming language is like a computer without a keyboard: pointing and clicking with the mouse sure is fun, at least for awhile. The ideal CBT authoring system has built-in handling for all the common CBT operations, and lets you insert programming code where necessary to accomplish those special things that will make your project stand out. Several systems allow this (you might see it advertised as visual authoring plus available programming).

  • THE PROMISE OF VB

  • So, what do you look for in programming? Basically, the authoring systems that offer an embedded programming language (i.e., one built-into the system) do so in either of two flavors: proprietary or standard. Proprietary means the authoring language is unique to that system; standard means the language is a version of a popular one, such as Basic or C. The advantage of a proprietary language is that the vendor can design it from the start to match the needs of the application. The disadvantage is that fewer people know the language; consequently, it is tougher to locate experienced authors, and labor costs will therefore be higher. Currently, thanks to Microsoft Windows and Bill, Visual Basic (VB) is the most widely used programming language. Millions employ VB proper, or VB for Applications (embedded within Microsoft Excel and Word). An authoring system containing a version of, or links to, VB virtually assures a steady stream of available help and snap-on components. Plus, Microsoft's recent announcement of VB Script (for the Internet) means VB programming skills will carry over to that up-and-coming environment as well.


    Speaking of the Internet, care to deliver CBT over it? That will be possible someday, but for the near future, it remains a goal rather than a reality. The primary bottleneck is not the availability of authoring tools, bandwidth, or, in bottom- line terms, the speed that stuff appears on the screen. Everyone has seen a Windows-based application that runs (crawls?) slowly. Now, imagine taking that application, and forcing everything to be transmitted over a modem; that's the Internet. OK, you can stop imagining now; I don't want to trigger nightmares. The unfortunate reality is the Internet, at present typical communication speeds, is way too slow for effective, serious CBT. Sure, you'll see some postage stamp sized multimedia demonstrations soon, and they do mark a starting point, but there is a long road ahead. How long is that road? As a benchmark, let's look at another media distribution mechanism: CD-ROMs. When did CD-ROM drives gain wide acceptance? Perhaps 1994? When did CD-ROMs first appear? About 10 years before that.


    Does this mean we won't see real CBT on the Net for another 10 years? My guess is it won't take that long; five years might be more realistic because of the high interest level in the Net. Most current modems transmit at a 14.4-Kbit/second rate, and 28.8-Kbit/second is becoming more widespread. ISDN, while an improvement, transmits only about five times faster than that. That's the rough equivalent of going from a 360K floppy diskette to a 1.44M diskette...nice, but not revolutionary. The new technology that holds the most promise is the cable modem. Cable modems that transmit at a 10-Mbit/second rate are now available...that's about the speed of an older, corporate LAN, and about right for serious CBT. Of course, you do need to be wired with coax, and your cable company needs to support this use, and... details, details!


    For the next several years, you'll still need to concentrate on delivering CBT via the computers people are currently using. Choosing an authoring system specifically designed for CBT will reduce your development effort and costs. Looking further down the road, it won't be communications technology that determines the ultimate usefulness and success of CBT on the Internet, the benefits (or lack thereof) will. What value does the Net add to CBT? A larger audience, perhaps. Easier accessibility and updating (after everyone is wired). What else? When veteran CBT developers see the Internet, they think of the PLATO system (from the University of Illinois and Control Data Corporation). PLATO did many of the same things being lauded on the Internet now, and it did them 20 years ago. As someone once said, "History repeats itself because no one listened the first time." Maybe it was Andy Rooney.


    Steve's favorite PLATO game was =empire=. Contact him at Intersystem Concepts, Inc., 410-531-9000, or intersys@insystem.com.