2(3), May 1985, page 7

[conclusion of Jensen article and entire Bridwell and Ross article at top of this page]


A COMPARATIVE STUDY OF THE FIRST-GENERATION INVENTION SOFTWARE

Ellen McDaniel
Department of English
Texas A&M University
College Station, TX 77943-4227
(409) 845-9935 or 3452


COMPUTERS and COMPOSITION 2(3), May 1985, page 8

A comparative study of this "first generation" software has not yet been done. I have initiated that research effort evaluating existing invention and prewriting programs. I am reviewing the work of the invention software "pioneers": BURKE, TAGI, and TOPOI by Hugh Burns; SEEN by Helen Schwartz; WORDSWORK by Cynthia Selfe: COMPUPOEM by Stephen Marcus; QUEST and FREE by James Strickland; PREWRITE by MiMi Schwartz; and COMPOSITION STRATEGY by Behavioral Engineering. (I am still acquiring programs and hope to have some others evaluated by May.) I am comparing and evaluating these programs to learn what in them should be emulated or superseded in the CAI development that is going on now, much of which is being done by these same people. Even while recognizing that much of what these people have done will have to be discarded or substantially revised in current and future CAI design, I believe that my study is a tribute to these people's work as well as a serious assessment of it.

My study is based on evaluation criteria that I developed while working on a PLATO courseware development project last year. That project included the rehosting of courseware onto the IBM-PC. Among other things, I have reviewed the English software for ease of use, control of the student-computer interaction, best use of the computer's capabilities (speed, storage, graphics, function keys, etc.), inventiveness, organization, and degree of improvement over traditional means of teaching the same or similar materials. The evaluations show that the programs range from on-line workbooks that simply put print materials on-screen, to programs that make ingenious use of activities that rather painlessly and even comically get turned into writing.

This review is also done in light of three overriding concerns that I have for the present and future development of writing software. First of all, are there sound theoretical guides for our CAI development that have come from this first wave of programs, or is there just good advice--dos and don'ts? Secondly, do we know enough about the technology to employ its improvements in our software and to anticipate and align our efforts with its advancement into expert systems and artificial intelligence? Thirdly, what is the likelihood of using these or future English CAI programs in writing programs, especially large ones? These last two concerns are very much bound up in the nature and amount of support that universities are willing to give to CAI design, development, programming, testing, and administration--all time-consuming, collaborative, and costly endeavors.