At present we are just starting to use an AppleShare LAN (Local Area Network) for the Macintosh computers in our new facility. We have a long way to go before this equipment gets near an optimal use. This document contains specifications which are an attempt to help us move up the data ladder beyond stand alone microcomputing. Since I have taught courses in which networks were routinely used, I will specify the kinds of general courseware functions which ought to be available to any course using this technology. Some of these will require the addition of a modest amount of hardware. Most require software, some of which will probably need to be written for us, rather than purchased. With these specifications, I hope we can begin planning on how to acquire them.
+--------------------+
+--------->|: EMAIL |<---------+
| +--------------------+ |:
| +--------------------+ v
| | COURSE MATERIALS | +----------------+
| | syllabus | |: | :
| +----->| schedule |---->| STUDENTS |--+
|: | | policies | | | |:
|: | | assignments | +----------------+ :|
|: | | etc | 5 |: |: :|
v |: +--------------------+ |: | : | |:
+-------------+ +--------------------+ |: |: | |:
| |<-->| INSTRUCTION |<------+ |: |: |:
| TEACHERS | +--------------------+ |: |: |:
| | +--------------------+ v | :|
| |<-->|: |: +--------------+ | :|
+-------------+ | |<--| ASSIGNMENT | | :|
5 | GRADEBOOK | | COLLECTION | | :|
| | | +--------------+ v :|
| | | +------------------+ :|
| | |<--| TESTING | :|
| +--------------------+ +------------------+ :|
| | | :|
| v v :|
| +--------------------+ :|
+----------- | EVALUATIONS |<------------------------+
+--------------------+
Figure 1.
If you have never taught a course with networked computers, you have worked with information that is essentially limited to the levels for transmission of data on paper.
Much of the labor in teaching lies in handling all the paperwork and there is little hope of tools which amplify your abilities until you get to networks. Teachers working with paper usually process a few 100 kilobytes of text a day, depending somewhat on grade level. Teaching with stand alone microcomputers is very comparable to textbook data levels (a few megabytes). At these levels, stand alone computing is a supplemental device to teaching. After learning computers with floppy disk use, your teachers may stand back and wonder if it was worth it. What is gained in speed is often lost in setup time or being restricted to the machine rooms, and they end up doing pretty much the same as before. Working with diskettes is like choosing workbooks or textbooks. Often it is just faith that computers will be needed in the future that sustains their efforts, rather than tangible differences in what they can do as an instructor.
Once connected through networks, data volume easily rises to 100s of megabytes and eventually gigabytes. At this level, the quality of data flow exceeds anything you would attempt to handle on paper. This does not mean you can just get more students per teacher ... it mostly means that teacher and each student exchange more information in the process of instruction. Measurements of learning rarely or never attempted can become routine.
The beginning use of a network is usually electronic mail (email). At a starting level, students and teachers are given network accounts which permit them to send simple messages to each other. These messages are time-stamped, location-stamped, and logged by the system so there is no residual doubt as to when they are sent, received, and processed. Email systems are easily obtained through commercial sources.
When first used, email establishes a method of objectively communicating. The delivery is nearly instantaneous on a local level and almost as quick even on a global scale. The speed of transmission resembles telephone or radio services, but the construction of a textual message allows editing and self-reflection to enter the process. The result then has the advantages of more thoughtful construction than face-to-face communication, but the speed of delivery found in electronic communications.
As a teacher I have used email with students for solving individual problems with course curriculum delivery. If one or a few students have personal difficulties in meeting some requirement, they can post a description of their problem which I can then respond to in a regular timely manner. This is better than the old method of playing telephone tag or office sitting to communicate. Their initial requests and the teacher's replies are automatically logged and filed for future reference if later problems arise. If many students start requesting the same explanation of some feature of an assignment, it quickly becomes apparent that something was insufficient in the initial material. After a few iterations of a general correction process like this, the production of course materials directly improves.
Some instructors claim that anonymous delivery of email is a beneficial utility to have in course design. This feature allows students to bring up issues which might be sensitive and they would not do so to an instructor who controls their final grades. This might be a helpful option, but not as a main problem solving agency. In the solution of some course problems, email serves as a special contract between instructor and student for special assignment considerations. In such cases, anonymity must be set aside. The use of pseudonyms should be reserved only for course transactions which are tangential to the process of instruction.
A course in a LAN usually has some disk space set aside for its instructional materials. Students are given read-only privileges to this area where they find copies of the course syllabus, assignments, statements of grading policies and so on. They can pull copies of this material to their local work stations anytime they wish and process it any way they want (usually print hardcopy or save to their own floppies). Any changes in course materials, made once by the teacher, are immediately accessible to all in the course. Usually there is one short file which is displayed to your students when they login into their LAN accounts. This serves for course announcements, informing them when new material is put online or when old material has been updated.
Besides reading materials, software can be placed in the course areas which students can pull local and execute. Often this may be a sample solution to an assignment or some exercise instead of textual explanation. I have put course software used for grading some of their assignments on line so the students could apply it to their homework before submission and see results while they can still correct it. Similarly, testing software can allow students to practice test themselves, relieving some grade anxieties.
In general, much of the work in teaching comes from duplicating and moving paper in the preparation of course material. Once freed from this burden, more effort can be spent in creating and drafting course work since distributing the result is instantaneous. In working with paper, most documents went through less than 7 draft copies. In electronic media, documents typically go through 20 draft copies, resulting in superior quality of final form. There is considerable advantage when instructional materials are produced and distributed in purely electronic form, making printed paper an optional secondary process.
Instruction in technology needs some specialized hardware. Some of this is available in our present Mac classroom design.
The first requirement is a teaching station which has central visibility from student seating. Overhead projection facilities of the teacher's monitor is used for direct instruction in software use and the display of data files. This permits a teacher to show students what operations of the mouse cursor and the keyboard are needed for various software.
The second need is that the instructional station has immediate access to the video output of any of the student workstations. A teacher should be able to observe what any student is doing at their workstations, without leaving the instructional station. This is the LAN equivalent of how teachers used to walk through the rows of desks observing students working at their desks. This removed instructors from their position of central visibility, violating the prime ergonomics of classroom teaching. Walking behind the students, peering over their shoulders at the monitors is more isolating than desk touring and lowers systematic observation of student performance. A teacher should be able to systematically step through a series of student displays without leaving the central teaching station. If one student is having difficulty, the teacher should be able to simultaneously place a pointer on the video they are both watching. If other students have come upon a common problem or produced a unique result, their monitor displays can be routed to the overhead projector through the teaching station. I have an unconfirmed report that Falloron Software has a product, called Timbuktu, which does this something like monitor sharing. In which case, a software solution may be available.
One of the main advantages in teaching through a LAN is the handling of homework assignments in electronic form. A considerable part of a teacher's job is correcting student work. The power of this technology to amplify what a teacher does can not be brought into play until assignments are routinely processed in network form.
On the starting level, collecting student assignments through a network is relatively simple. Part of the disk space made available to the course is set aside so students have write-only privileges in this area. They only have to copy their assignment files into this directory and their homework is time stamped by the file copy process. Since the students are not given read or deletion privileges in this area, they cannot see or change the homework submitted by others.
On better levels, software for assignment collection registers each homework submission into a database or spreadsheet which students and teachers can check for classwide statistics on each assignment. This verifies for the student that each file has been submitted for grading. The teacher can see the completion rates for each assignment or student.
Automatic grading of some assignments becomes possible when all homework is submitted in a standard network format. In some cases an assignment in computer form can be completely scored with results immediately available to the students and logged in a gradebook spreadsheet or database. In other cases, partial evaluations can be made in a form that is completed by the teacher who enters the remainder of the scores to the gradebook.
As an example of a partial grade evaluation, let us consider the assignment of a term paper. Once the student has submitted a file, software can provide an immediate summary of whether the essay fits the assigned length characteristics, the readability grade level of the text, and frequency of specialized vocabulary items expected in the assignment. Spell checking can automatically flag most of the common errors in the text (although students should know how to spell check before submission). Software can check that standard reference forms (like APA or Turabian) have been followed, and the thoroughness of library work can be evaluated by comparing citations to a list of major works expected for the assignment. Citation ratings can be collated for the entire class then each student's effectiveness in library research can be rated against the others who submitted this assignment, besides the standard reference list expected by the instructor. Once these technical qualities have been graded in the assignment, a copy is presented to the teacher for reading and evaluating the nontechnical qualities, like continuity, opinion or logical thought sequences. The teacher is then able to focus mainly on the humane aspects of writing since the technical qualities are scored independently.
Once we move beyond the idea of a term paper being merely a linear sequence of words on paper, we will have more hooks upon which automatic processing can hang. If the assignment were submitted in a hypertext format, the teacher (or other student readers from the class) could tag any portion of the student's text with comments that are available on a level separate from the original. Also the complexity of the hypertextual structure of the student's work is subject to automatic grade evaluation ... this is an indication of how much effort the student put in completing the assignment. A flat text from a word processor would score rather low compared to a deep text with a rich structure of inner connections already explored by the students.
Records of course grades can be stored in a standard spreadsheet format which students are given read-only privileges. This permits students to access the gradebook and see their progress. In the courses I have taught, a special front end to the gradebook spreadsheet is used so the student can only consult their own grades and general class statistics.
Testing is an area where LAN deployment can make substantial improvements. Like the example above on student term papers, a test or quiz should become more than just a fixed sequence of questions and responses. Most of the test generation software that exists today is oriented toward testing on paper or some intermediate form (like scanning bubble sheets or stand alone floppy disk test administration). The network provides an immediate collation of student responses into a course data pool. Feedback of student performance can be done on a question-by-question basis, or deferred until test completion. Either way, the labor in test scoring need not be constrained by teacher time requirements.
Besides the mechanics of test administration and scoring, test generation can provide item response analysis for local teacher-made questions. The first time you give a new course, you may not have anything other than personal intuition on how difficult test questions are. By the time you have given the course three or four times, you should have exact statistics on item difficulties and comparison values on students and class groups. Few teachers take the time to perform these statistical calculations on their testing when they have to put their effort into paper testing and scoring. In LAN testing these features should become automatic standard processes.
Once item response statistics begin developing, you have the options of pursuing nonlinear testing sequences, not available or very clumsy for paper formats. All the students in the course do not take the same set of test questions. The software can generate unique tests from an item bank for each student, matching items for the same difficulty levels. This prevents students from cheating by sharing response sequences with each other. Also, if a student shows a weakness in one area of testing, additional questions of same or weaker difficulty from that area can be presented. This prevents accidental errors or nonsystematic gaps in knowledge from overly affecting final results. Interactive testing formats should become standard once LAN operation performs the item response analysis and item banking tasks. The teacher is freed from administration and scoring details to focus on question writing which most clearly represents course content.
Evaluation of instruction is at a very primitive level. Elementary and secondary teachers are rarely evaluated by student response and university instructors are subjected to a one-size-fits-all student evaluation which ignores course content. Not only does this amount to a popularity contest which rewards easy courses, it misses valuable information which ought to correlate to specific course functions which can be augmented through instruction.
Evaluations are usually done with students completing Likert attitude responses, ranking statements on a 5-point good-to-poor scale. As such, these items can be subjected to a statistical analysis comparable to the questions in testing software. Not only should the instructor and physical properties of the classroom be evaluated, but each major course component should get a rating level. The textbooks, readings, assignments, exams, and major elements of the curriculum content each receive an attitudinal evaluation. Unlike the instructor and classroom, these elements can be compared to the data of technical analyses found in the LAN testing and grading components of the course. The volume of data in making such comparisons is rarely attempted by hand on paper.
When I taught a networked course, these evaluations told me when the textbook began slipping out of date with other course material and was serving fewer students than previous semesters. Some students who scored lower on certain assignments of a known complexity level (measured by grading software) rated those assignments poorer. Often, when the complexity was within their grasp (as indicated by testing), this indicated that either they did not apply themselves to task or there was something internally wrong with the assignment. This was factored out by students who scored highly on the same assignment ... they rank it low when something is internally wrong that does not show up in complexity measurement. In general, the students who score high and rank a component low are pointing at the weaknesses, while students who score low and rank a component high are pointing at the strengths of a course. All elements of curriculum, especially content areas, need to receive a uniform evaluation of student-perceived difficulty. This level of analysis should all be done automatically, giving an instructor something concrete to work on when improving coursework.
This is all secondary, indirect evaluation ... after the fact. Students who received the instruction are asked how they responded. This we have inherited from the teaching paradigms which were confined to paper measurements. Perhaps soon the networks will provide us with a large enough framework in data handling that we will be able to evaluate instructional design directly: doing evaluations based on what a teacher does or how a curriculum is structured rather than how students react. There is nothing wrong with measuring student reactions ... but if that is all, we are missing two-thirds of the process in instructional design.
In general, a LAN which is isolated to itself is only slightly better than a stand alone microcomputer. The central problem is how to get software and data in and out of the thing so it can be used productively. The expensive way to go about this is to purchase data through floppy disks. As medium for data distribution, floppy disks are not very different from textbooks: the main expense is in the distribution of the diskette itself, not the production of the information.
Connection to wider area networks (WAN) is a solution to some of this. When data is distributed through purely electronic means, the cost is reduced by orders of magnitude. WAN data delivery costs much less than a penny per kilobyte and is charged through central university funds. Compare this with floppy disk data delivery which is roughly $5-$10 per kilobyte which you pay for directly. Computer networking is also the least expensive option for long distance communication.
Our facilities were built with fiber optic cables to the central campus computer services. This means that the main expense of connecting to the rest of the university has already been accomplished. Data from any part of the campus should be readily accessible from our Macs. Our mainframe computer is connected into global networks, so information from any part of the earth would be equally accessible.
These connections require a terminal emulation card, some emulation software, and the cabling at a cost of about $650 for each microcomputer.
Courseware functions require that students and faculty have routine contact with the LAN hardware. The easiest form of contact is to call the LAN over the telephone lines and deliver or extract your data by file exchange. This is much easier than going to the LANroom and sitting at one of the machines. Mainframe connections could simulate this indirectly, by allowing users to phone in there. Direct phone service to our LAN would accomplish several other functions directly at very small cost.
Our university has very poor modem service through the mainframe. All of the standard file exchange protocols have been disabled in favor of a proprietary protocol which is available primarily for one manufacturer's equipment. Direct phone service is needed for common file exchanges.
Bulletin Board Service (BBS) software is readily available for educational services. Our students and faculty should become familiar with BBS operation and what educational uses they have. Direct phone service is needed for this.
Direct phone service to a LAN is not very expensive. Modems cost around $100 dollars plus the cost of a phone line to each. Two or three such setups are probably adequate for initial needs.
Dr. Stan F. Kulikowski, II,
stankuli@UWF.bitnet
Research Scholar
Educational Research and Development Center
College of Education
University of West Florida
11000 University Parkway
Pensacola, Florida 32514-5753
904-474-2800
Fax: 904-474-3126
Return to GLOSAS News Contents for this issue.
URL: http://library.fortlewis.edu/~instruct/glosas/teach31.htm
March 1993
GLOSAS NEWS was orinally posted to the WWW at URL: http://library.fortlewis.edu/~instruct/glosas/cont.htm by Tina Evans Greenwood, Library Instruction Coordinator, Fort Lewis College, Durango, Colorado 81301, e-mail: greenwood_t@fortlewis.edu, and last updated May 7, 1999. By her permission the whole Website has been archived here at the University of Tennessee server directory of GLOSAS Chair Dr. Takeshi Utsumi from July 10, 2000 by Steve McCarty in Japan.