Jack Jobst and Billie Wahlstrom in their article, "How `Friendly' Should Effective Software Be?" addressed the need for user-friendly software but questioned the degree of friendliness in terms of its conduciveness to learning. They raised the consideration that, depending upon the student, the teacher's attitude may be more conducive to learning than the user-friendliness of a program (Jobst & Wahlstrom, 1984). Regardless of the accuracy of their observation--their point in general seems to be well taken--their conclusion rests upon the assumption that the computer in CAI is the instructor. This concept of software-instructor seems to indicate CI versus CAI. If, however, we regard the computer not as the primary instructor but as an "assistant" or "tutor," then we may be able to provide a more accurate fit between audience and instructional strategy.

As an instructor in a two-year, open-admissions college, with the assistance of two

p. 8

programmers who were peer tutors, I rewrote the grammar section of the English curriculum to include the use of the computer. Recognizing that there is a difference between teacher instruction and machine instruction, our software was programmed to simulate a "tutor," not a teacher. A tutor, as an objective person, exerts no pressure to perform but provides guidance for future study. Tutors reinforce classroom instruction and allow time for practice to identify problem areas. Following tutoring sessions, students generally have an awareness of performance and a responsibility to pursue what is needed. The purpose of our software-tutor complied with this definition; it reinforced the classroom instruction by providing tutorials for drill and practice, and time to master the concepts covered in the main course.

The group who used the software-tutor was composed of college students enrolled in regular core courses but placed in remedial English on the basis of a placement test and a writing sample. However, we did not want to attach the stigma of remedial instruction to the software and wanted to acknowledge the students as regular college students. Therefore, the software program, functioning as a tutor, was not overly friendly-cute and was not hostile. The students were instructed that the software was a supplement to instruction. They understood the persona exhibited by the software epitomized a "tutor" not a teacher. These tutorials for drill and practice were supported by me and human peer tutors.

The concept of our tutor persona was carried throughout the software program in the following

p. 9

ways. No prior computer knowledge was required on the part of the students; to use the software, the students did not need to know access codes and passwords. When the students entered the computer room, the peer tutors (programmers) had already accessed the software program and made sure a title page was displayed on the terminal. The student users were instructed to press the "enter" key, and the computer responded by asking them to type their names. The computer then displayed a list of drills in a menu format and asked the students to select one by entering a number between one and five. Following this activity came a brief explanation of the particular grammar drill. Once the students understood the requirements, they pressed RETURN and the drill began. There was no time limit established for answering or reviewing. The students always gave the command to advance to the next question.

After students were shown a sentence and asked to identify a word or part of a sentence applicable to the drill chosen, they responded by typing in an answer. Phrases, verbs, subjects and objects had to be typed in entirely to be properly identified. However, in order to save time and key strokes, the program accepted some abbreviations as well. Multiple subjects and objects could be answered in any order.

If a student's answer was correct, the program would continue. If the answer was incorrect, the machine would display a message much like the following:

Incorrect Answer
Correct Answer is ______

p. 10

Students proceeded to subsequent exercises by pressing the RETURN key.

The students regarded this software-tutor as friendly for two reasons. First, it provided them with the option to sign off at any time. The students could answer as many or as few questions as they wished and still receive a score. The drill ended when an EXIT command was given.

Second, the program provided students with specific feedback, but allowed them to evaluate their own performance. Upon exiting the exercise portion of the program, each student received a printout to be used as a study guide. On the printout appeared the student's name, the date, the time, and the number of the drill that had been completed. Each drill was listed along with the number of correct answers. If there were twenty answers in a specific drill and a student scored five correct answers, he or she could see that further work was indicated in that area. In addition, the program also provided an overall view of performance. It listed the total number of questions answered, the total score, and the percentage grade.

The results of the drills were stored in a file so that I could obtain a hard-copy printout of each student's performance. I used this printout to determine the strong and weak points of the entire class.

The use of the software-tutor to supplement instruction provided by the teacher and peer tutors created an incentive that affected students' performance. The students who used the software tutor received immediate feedback on their grammar and usage errors on the machine and clarification by the teacher and human tutors. These

p. 11

students had the opportunity to think about their mistakes before proceeding to additional instruction. Students did not generally repeat incorrect responses on the computer as often as they did on paper.

Our success in this project depended on our ability to emphasize the concept of "tutor" consistently throughout the curriculum: software-tutor, peer tutors at the terminal, and peer tutors in the classroom for writing. Using this approach--beginning with the software--worked well for our students. There was no intimidation. The students still had the responsibility for their own learning and the chance to communicate with the teacher for instruction. They became eager to learn and wanted CAI applied to other courses. Based on the positive responses of the students using the software tutor, non-remedial students asked to use the computer drills for review. All students understood the concept of CAI versus CI. Further data on our investigation involving the software-tutor is available upon request.

The purpose of a software package is the first question to be answered when establishing the persona of program. The second factor is the age and the level of the user. On the basis of these variables, the relationship between the persona and the learner can be evaluated more accurately and with greater success.

p. 12

Jobst, J., Wahlstrom B. (1984, November). "How `Friendly' Should Effective Software Be?" Computers and Composition, 2 (1), p. 5.

Claire Redmond teaches at the University of Rhode Island in the College Writing Program, Department of English. Cheryl Lawrence and Frank Villani teach at Framingham State College.