Thoughts on the Paperless Classroom
Our PerspectiveWe do instructor-led technical training in corporate classrooms. Starting in the late 1990s, we augmented our training deliveries with web-based materials. Our intent was not to replace the classroom with the web as a kind of computer-based training (CBT), but simply to mitigate some of the logistical headaches of printed training workbooks. In this paper, we tell what we've learned from our pragmatic, real-life experience. We describe how we got real value and increased quality by using the web. What we've discovered is that it's not a perfect solution but it does offer some remarkable opportunities.
The VisionOne day in 1996 we were getting ready to teach a class. This is how we make our living, and the scenario was a familiar one, oft repeated. We were struggling yet again with assembling workbooks, preparing foils for the overhead projector, making sure we had the current version and that all the components were in sync, the files for the hands-on exercises were the right ones, etc., etc. And then it dawned on us: why not put all this stuff on the web? End the logistics headaches forever! The paperless classroom! -- what a concept!
Here are the benefits we anticipated:
But we soon discovered that a truly paperless classroom is not entirely practical.
- Save trees -- electronic media are so much more politically correct. What happens to all the pounds of paper we hand out in the classroom? They gather dust on our students' shelves until the day they go out with the office-cleaning trash. End the waste!
- A new, high-tech presentation -- on-line, multimedia, color, zing! No more drab paper. Project a state-of-the-art image for our company.
- Easy updates
- Simplified logistics -- no workbooks to maintain, print, ship, or distribute
- Simplified coordination with authors: they just edit on (or upload to) the website; no more frantic mailing of galleys, etc.
- No foils or overhead projector; use the data projector
- Fabulous for teaching programming -- compile and test in the same environment as the didactic material itself
- With a click of the mouse in a browser window, students can take tarballs and zipfiles; we need no longer to distribute data on diskettes, by e-mail, or even by ftp.
- Hyperlinks can point both to source code and to executables. With a click of the mouse, students can run programs and scripts for themselves without having to follow confusing instructions from the instructor and without having to passively watch a demo (and we all know how demos go... or don't go!)
- Helper apps (i.e. whichever utility is being discussed) can be invoked via hyperlinks instead of having to hunt for icons and opening files in the apps.
ObstaclesHere is what we found when we attempted to use web pages as our course materials with no printed workbooks.
- Students do want paper! They need margins in which to take notes. They use highlighters and Post-Its. These are part of the mental process by means of which they "make the material their own."
- A tangible object -- the printed workbook -- satisfies a subtle but genuine psychological need. It's important to convey the sense that from our classes students obtain something tangible.
- Copyright protection becomes an issue. Password-protecting web sites is of limited value; passwords don't remain secret, yet changing them frustrates past attendees.
- Students can't carry internet-connected computers around as easily as they can books, e.g. to read them on the bus.
- People read 25% slower on-screen. Scrolling is annoying.
- Web performance is vulnerable to network traffic, to hardware failures, and to software configuration problems on the servers and in the classroom.
- Reliable access demands high availability of servers. This adds to the pressures on our IT/IS people.
- We love using the overhead projector and relinquish it reluctantly. With the overhead projector, we can scribble on the foils, use color, underline key points, etc., facing our audience the whole time. With the data projector, we can highlight key points with the mouse and, if we're lucky with the orientation of the PC, we needn't turn away from our audience. This is not quite as nice, but we can live with it.
Paradigm ShiftsThe web is not a replacement for paper. It doesn't act like paper and we don't use it like paper. It's a new paradigm. It demands that we rethink how we present information.
The web is a hierarchical rather than a linear medium. Printed books are linear; they start at page one and proceed in obvious sequence; page two follows page one and so on.
Information on the web, by contrast, is organized hierarchically. In our web-based courses, for instance, the table of contents is the top-level node, with each chapter branching off from it. The sections within each chapter are, in turn, branches off that node, and so on to arbitrary depth.
Thus, navigation becomes an issue; without an accurate mental model of the structure of the information, it's easy to lose one's place and get confused. Windows proliferate on the screen and it becomes tricky to find one's way to the next topic or to a previous topic for review. We've learned that it's important to establish in our students' minds an accurate mental model of the structure of the material before attempting to navigate through it with them in the classroom. Early in each class, we have them open their browsers and point them to the top-level page, and make sure they're comfortable the idea of the hierarchical structure and with the two styles of browsing (same-window and new-window).
The BACK button is a simple and obvious but often inadequate navigation mechanism. It's tempting to build something fancier, such as a columnar frame on the left-hand side of the window containing a clickable table of contents. But we've concluded that this would be counterproductive and have resisted the temptation. Frames have numerous annoying behaviors. Good navigation mechanisms are notoriously difficult to engineer and add yet another degree of complexity to the material. Better to keep things simple.
Another important feature of any navigation scheme is to provide a sense of progress through the material. Psychologically, this sense of progress is necessary for students' satisfaction with the class and with their own performance. This is simple to accomplish with linear media such as books; having arrived at page 100 in a 200-page book, one sees at a glance that one is half-way through. On the web, this is harder to do. The change in color of a hyperlink that's been visited partially achieves this, but there's no practical way to present all the hyperlinks visibly at a glance.
We've Ditched the Paper; Why Not Ditch the Classroom, Too?Cut costs! Save Time! Less travel! These are worthy goals, and a variety of technologies can help achieve them: videotapes, CBT, web-based training, and distance learning (at your desk, or in high-tech meeting rooms) offer partial solutions.
But the traditional classroom remains the most popular and, we claim, the most effective way to improve workers' skillsets. Why?
The primary reason, we believe, to our chagrin, has nothing to do with the quality or content of what we do in the classroom. It's simply the luxury of time. Hours a student spends in the classroom are hours away from his or her desk, away from the ringing telephone, away from the distractions of e-mail, away from the boss's demands, away from the to-do list, and away from the fires that need extinguishing. In the classroom, one has permission -- a mandate, even -- to focus on one's own growth. Outside the classroom, this luxury is rare.
There are other reasons, too, why the classroom remains the most effective means for technical and professional growth. Unlike inanimate mechanisms (videotape, CBT, web) instructors possess exceptional technical experience as well as "people skills" so they can:
These all increase the relevance and effectiveness of training, and contribute to its value.
- Evaluate progress and help the student focus
- Relate the course material to a student's specific work goals
- Adapt and fine-tune the material (both the pace of delivery and the topics covered) to the needs of the individuals in the room
How Do You Know It's Working?This is one of the stickiest issues training professionals deal with. It's difficult to measure whether learning has taken place in any meaningful or lasting way or, even more important, whether real productivity gains have been achieved in return for the corporate investment.
The most common -- virtually universal -- measurement tool is the post-class evaluation form. We rarely teach classes in which these are not demanded. They do serve a useful management function, alerting training coordinators when disasters occur, such as logistical problems or quality issues. But they fail utterly to provide any meaningful measure of productivity gains.
Dubbed smile sheets, "eval" forms (on-line or on paper) report little more than the attendees' emotional state at the time of filling out the form. Some vendors (e.g. Learning Tree, one of the largest) even reward instructors with bonuses for high scores. Instructors quickly learn to "teach to the evals," working the class up to a fever pitch for the moment at which the forms are distributed. The customer's real goal -- productivity -- is irrelevant at this point. The numbers tell little more than how much the attendees liked the instructor's personality. Yet, training organizations often use the eval results as the primary decision criterion for whether to continue or modify a training offering. Lacking other data, what else can they do?
These issues -- training effectiveness and how to evaluate it -- are little impacted by whether the classroom is "paperless."
What's FabulousFor teaching programming and other technical software-related subjects, the web offers capabilities unimaginable with conventional media. For example, the materials for a web-based programming class can include hyperlinks to source files and to executable versions of same, via differently-named filesystem links. Suppose the web server is configured to execute (via the mechanism termed Common Gateway Interface, CGI) any file named with the extension .cgi and to treat as raw text to be delivered verbatim to the browser any file named with the extension .txt. In both the Unix and Windows environments, one file can have two names simultaneously, by means of a filesystem link (termed "hard" or "symbolic" in Unix and "shortcut" in Windows). Thus, a file containing a Perl script could be named both tryme.cgi and tryme.txt:
The web can also automate many of the routine chores of training administration:
Many training organizations have implemented some of these functions on their intranets, but we have yet to see the medium exploited to its full potential and look forward to the opportunity to implement such a system. Detailed discussion of training administration automation is beyond the scope of this paper, but the issue of on-line workbooks is central.
- course catalog maintenance and distribution
- class signups
- attendance recordkeeping
- class evaluation forms
- interdepartmental recharging (via software interfaces with in-house accounting)
- workbooks, files for labs and exercises
- links to related resources, documentation, and examples
CompromiseOur business processes are honed by experience. We have learned that the web is not a panacea but use it for its strengths.
Thus, we satisfy the students' demand for paper. Though we're still not entirely paperless, it's a big step in that direction.
- We do print a workbook -- a subset of the pages on the site: chapter summaries, labs, interesting source code.
- We print it just-in-time.
- Page numbering is tricky. Browsers number every web page "1 of 1" when printing. But they also print the filenames so we incorporate the page numbers in the filenames, e.g. 1-1.html (chapter 1, page 1), 1-2.html (chapter 1, page 2).
- To simplify the task of printing the workbooks, we create a directory containing hard links to the subset of pages to be printed. The clerical staffer's task thus consists simply of clicking each hyperlink and the print button.
- By putting these hard links in a distinct directory and by generating the footer with a server-side include, we generate an alternate footer suitable for printing. The footer we use for web pages to be seen on the screen is larger and more complex and contains hyperlinks and other things we wouldn't want on each printed page. In print, the footer need be only a copyright notice.
- Printed workbooks require a cover and a copyright page for printing. These things have no place in the on-screen web version. So we put them only in the print-list directory.
Conclusion: Alas, Some Paper is UnavoidableStudents do demand at least a little paper. They need margins to take notes in. They need something tangible to take home from the class.
Also, we have learned that even long after class, students wishing to refer to material presented in class, in preference to finding a bookmark in a browser, find it most natural to reach for a book on a shelf.
So workbooks do serve an essential purpose. Until our culture changes so that people are truly comfortable with completely electronic media, we'll never completely dispense with paper.
- Measuring Return on Training at Merck & Co.
- Journal of Applied Psychology
- The Educational Technology Page of the Virtual Library. Maintained by the TECFA, an academic unit active in the field of educational technology, in the School of Psycology and Education in FPSE, the University of Geneva, Switzerland.
- A good list of software resources for web-based instruction.
- Another list, called Web-Based Training Information Center.
- Reading from paper versus screens: A critical review of the empirical literature, Dillon, A., Ergonomics, 35 (10), 1297-1326 (1992).Do people really read slower on-screen than on paper? Here is some research that reports no differences between paper and monitors in accuracy, comprehension, fatigue, nor the number of eye movements made by subjects. However, the finding that is probably quoted most often is that silent reading from paper tends to be reliably faster (20-30%) than reading from monitors. This observation, which was reported by several studies in the 1980s, is reported in several current text books and is generally considered still to be true.
- WebNet JournalThe web site of a quarterly print magazine subtitled "Internet Technologies, Applications & Issues" with emphasis on education in academic and corporate environments.
- A Comparative Evaluation of Display Technologies for Reading, Gujar, A. U., Harrison, B.L., and Fishkin, K.P., Proceedings of the Human Factors and Ergonomics Society 42nd Annual Meeting, 527-531 (1998).Subjects read from a variety of different formats: a sheet of paper lying on a desk, paper displayed in a frame without glass, paper displayed in a frame under glass (to evaluate possible reflection problems), a monochrome monitor, a 21" color monitor, and a large image projected on a screen using an overhead transparency. The researchers found no statistically significant differences in reading time or detected errors. Nonetheless, the subjects rated reading from paper significantly better than reading from the other 5 conditions. Conclusion: the new, high resolution monitors allow users to read as fast from screens as they do from paper. It will be interesting to see how long it will take before this fact is generally known, and accepted as true by usability professionals.
Dan Keller Technical Services
2248 International Bl., Oakland
California, USA 94606