Complexity Theory: CALL @ The Edge of Chaos by Stephen A. Shucart 

venturing out... with Mitchel Resnick by Scott H. Rule 

Email Options for Students by Paul Daniels 

Officer Reports: 
Elin
Swanson
CALL News
Scholarship
This and That
Conferences
Call for Papers
English Drama

In a recent JALT CALL thread, Layton and McLaughlin write of evaluating learning environment attributes which are close to our development interests:  

1. extent of interaction of student with subject-matter (through book, hypermedia, classroom, discussion, etc.) 

2. extent of interaction of student with subject-matter expert (which could be a teacher, a computer, or whatever) 

(I'll address the CALL aspect of their note in order to give some input from a development point of view.) 

Qualitative Aspects  
These are all elements noted by the writers. 

1. Presence 

2. Quality of student work 

3. Quality of student enjoyment 

First off, the CAI material, the courseware, has to have functionality that can record students' interactions. What sort of record keeping? Presence is recorded by the study session time start/finish as copied from the PC's on-board calendar/clock. If specific interactions require completion in a set time frame (say ten seconds to answer a true/false comprehension question, and the student does nothing, the program can simply ask the student 'Are you there?' in order to trigger the student's attention to task. Failure to answer this question can lead to the lesson timing out and quitting. The quit time is then recorded. But presence itself is of little real use unless considered in relation to quality of work. It is possible to record number of student clicks, right/wrong answers, to weight the difficulty of various questions/problems and so on and to create an algorithm that results in some sort of scoring mechanism. A mere percentage of right/wrong may not be adequate if the courseware is itself truly interactive in adopting itself to challenge student's levels, that is, if the difficulty or frequency of questions and syntactical challenges varies with individual student usage. 

Secondly, those courseware and interactions themselves must be valid. Two requirements here; (1) the courseware has to match curricula objectively understood as pedagogically valid and set by the teacher, (2) the interactions must exploit the material in ways that teach (as opposed to simple repetitive or mechanical game clicking). 

On the third, I cannot accept the notion of 'student enjoyment' as a measure of the learning environment. Students must be challenged and engaged. Their intelligence must be respected, not insulted. O yes, they must be taught as well. If these factors are developed sufficiently, enjoyment, whatever that creature might be, follows in train. 

Mechanical Aspects 
Record keeping is indispensable for evaluation, to ensure teachers that students are on task and not wasting their time. It is also useful in encouraging students to see the results of their work and to compare the current with past results. The network environment is the ideal control factor for computer material. But our experience so far has been that networking systems in use at most schools lack sufficient 'robustness' to enable the steady streaming of student access and record keeping from, say, 40 workstations, to the teacher's computer or server. Will NT 4.0 solve the network problems? Record keeping options also include storage to the workstation hard disc, to a floppy disk, printed out from the system, or exported to a standard data formats for use in FileMaker or Excel. 

DynEd's Courseware Management System represents an early effort at a fully integrated evaluation system, but results have been mixed, with some failures as well as successes. DynEd is now rebuilding record keeping systems from the ground up in a three stage development path, (1) solve the networking tech issues, (2) address the class management/admin issues, (3) resolve the student evaluation issues. All developers need input from teachers just as text publishers do - only more so. In the case of evaluation schemes, teachers must advise developers regarding which specific modes of interaction need recording. To what depth of detail? What statistical or comparative tools will teachers use with class or group data? Against what (if any) benchmark testing standards should student evaluations be rated? Each of these are separate development issues as well as legitimate testing concerns. 

Interaction with 'subject matter expert' 
Two thoughts are useful here. The ‘expert’ may be defined as the underlying engine powering the courseware. This engine is the most direct statement of the developers’ concern to teach. The engine may be inflexible and static or, as described above, dynamic. In this sense the ‘expert’ is hidden or latent. However, as processing power is increasing, it is now possible to transform the latent power of the engine into an on-screen tutor. This can become a visual presence that advises the students as they progress through a lesson, giving tips on language, or culture or even on how to study. How this ‘expert’ is characterized, or the options available for characterizing it, will prove interesting to ‘student enjoyment.’ Some students may prefer to not have such a character, especially if it is personified as too ‘cute’ or in some other way contrary to their personal tastes. 

Consistency of Environment 
One final note, for student records to mean anything, they must be in the context of a consistent courseware environment. This principle has supported all of our development work and is one of our basic foundations. All of our courseware permits valid evaluation in a similar manner. Inconsistent evaluation lacks validity. The creation of wide variety of materials within a consistent, internally valid environment is quite a challenge, you will agree. Networking that environment is very dependent upon the local hardware, whether it is properly installed and strung together, and whether the networking software is reliable and robust. 

You will agree that consistency of environment is vital when we move to the next topic for consideration, disc-less, on-line delivery of courseware. As always, it is the needs of the student and teacher that will determine the shape and contents of that environment and courseware.