Developing A Computer-Assisted Reading Instruction Program

Mark W. Simpson Sheila Bolduc-Simpson

Abstract: CARI is a computer assisted reading instruction program that was designed by two English Language teachers for the purpose of providing supplementary and relevant material for their students in three subskill areas. The program was written for intermediate level adults learning English in a business environment. The reading subskills include skimming, scanning and guessing. The paper describes in detail the content and structure of the program and outlines the steps involved in its development. Following the description are comments and suggestions for teachers on how to evaluate effective courseware. KEYWORDS: ESL, ESP, Basic, Super Pilot, reading software, software design, design and development, skimming, scanning, guessing, tutorial. Computers and learning are fast becoming inseparable associates in the field of modern education. David Linowes (1983) says they are causing a revolution in education as profound as that caused by the printing press (p. 1). More than just another teaching tool, computers are being used as teaching and learning partners. English language teachers are beginning to take advantage of what computerassisted instruction (CAI) has to offer to their students. CAI refers to the interactive use of computers. The computer presents material to and receives, analyzes, and acts upon responses from each student on an individual basis. The two kinds of computer assisted instruction are mainline instruction and adjunct instruction. The former method delivers complete courses. The latter provides units supplementary to classroom lessons (Hallworth and Brebner, 1980, p. 18). CAI is no longer a luxury available only to a few, but is not viewed by many as a necessary supplemental source of quality education (Poirot, 1980, p. 2). Background The revolution in computer assisted instruction began in the late fifties and early sixties. At that time CAI was introduced in selected sites across the United States. These sites were generally at the university level. In the mid-sixties the move was made to introduce CAI to pre-college students. The purpose was to increase skill levels. However, the idea of using computers in the classroom was still experimental. By the 1970's the microcomputer system was introduced. This resulted in lowering the cost of CAI to the point where most school districts could afford its use (Poirot, 1980, p. 2). Now, in many school districts in the States CAI is an integral part of school curricula. Problem Unfortunately, as with any new and powerful learning medium, we are at the early stage of learning how to make full use of the medium's capabilities (Bork, 1980, p. 59). This is evident in the development of software. Teachers who want to use effective teaching software are exposed to a large amount of software of little quality. Much of it does not take advantage of the capabilities of the expensive hardware that schools are investing in (Linowes, 1980, p. 3). Many of the programs are still

at the level of asking the student to answer a question and then telling him if the answer is right or wrong. Because the capabilities of hardware have hardly begun to be exploited, the majority of programs on the market are simple, unimaginative, or merely gimmicky. Most of the computer curricula developed for the micro are canned versions of traditional instructional methods and materials. They also have not been field tested for use in computer-assisted instruction (Thomas, 1981, p. XIV). This problem of poor software is the most troublesome aspect of microcomputers. Because computer programmers often are not educators, software suitable for the needs of teachers is not being produced. For every educational program there are ten games on the market. The teacher or school administrator has to wade through the morass of
34

programs labeled educational to find something that could possibly be used in his particular curriculum. Also, many educators are not informed enough about computer-assisted instruction to create the market for quality software programs. This problem is particularly acute in the area of English language learning. There is an absence of software in the form of instructional lessons and other related information. Jenks (1981) proposes a major internationally funded project that would collect, edit, code and store ESL instructional materials at all levels and for all currently perceived purposes. From this computer-based instructional program that would be custom selected for a particular learner (p. 223). This is indeed a worthwhile project to consider for the future, but the problem still remains for today's English language teachers. How might they develop or help develop relevant CAI software? Purpose Two educators working for a large oil company in Saudi Arabia teaching English for Specific Purposes (ESP) decided to tackle the problem. Their goal was to develop a reading skills program for their high intermediate level students. The purpose of this article is to describe to other English language teachers how educators with a little computer programming experience came up with an English reading skills CAI program for their Saudi students. Description of the Program The Subjects The software program was designed for Level C Professional English Language participants working for the Arabian American Oil Company (ARAMCO). All participants are graduates from universities in the Middle East where Arabic is the medium of instruction. These employees have been slotted by the company to fill future openings in managerial positions. They attend classes seven hours a day for twelve weeks in order to complete one level. There are four levels in the program, ranging from Basic to Level C. Level C participants possess English language abilities that fall into the high intermediate category. In a twelve week period participants in the program have between 150 and 175 hours of reading and writing instruction. Hours spent in the computer lab in this same period of time range from twelve to twenty-four. The term Professional English refers to English for Specific Purposes in a business-related oil industry-centered context. Instructional Objectives Level C was selected because of the program designers' experience teaching the Level C reading and writing course over the past two years. Level C was also chosen because of the diversity of skills and materials which the designers had at their disposal. The designers focused on three reading subskills.

They viewed these as essential for Level C students in order for them to acquire expected final level competency. The three subskills were skimming, scanning, and guessing. Skimming is quick reading for the general idea of a passage. It is an activity that is particularly useful when the reader does not have time to read the whole article, letter or passage (Clarke and Silberstein, 1979, p. 55). Scanning is similar to skimming in that the reader has to force himself through a selection at an initially uncomfortable rate, but the search is more focused since the information needed is very specific (Clarke and Silberstein, 1979, p. 55). The third subskill is guessing the meaning of a lexical item. The vocabulary attack strategy the program designers used was obtaining meaning from form, function, and context. The specific instructional objective the two teachers had in mind for the student in Level C was to reinforce the reading curriculum with a related CAI program. The terminal objectives, as outline by the Curriculum Unit at ARAMCO, for Level C are stated as follows: 1. Identify the purpose and intent of documents. 2. Identify the reason or purpose of reading a particular document. 3. Ignore irrelevant information. 4. Determine the degree of relevance of a particular document to job related needs. Under these terminal objectives more detailed objectives are stated. These specifically refer to the skimming, scanning, and guessing subskills in which the student must have a minimal competency before completion of the level. The CAI program was designed to capsulize the reading strategies taught in the classroom, in addition to giving the student practice in using these strategies with materials he himself could choose. Timing is an integral feature of the program and is part of the goal of achieving competency in the reading subskills. Although tips are provided for the student in order to help him skim, scan, and guess, the only way to improve these subskills is to force him to read more rapidly. The program provides the skimming, scanning, and guessing items before the text is presented. Clarke and Silberstein (1979) concur that teachers must provide the questions and coach the student through the text. Later, the student will be expected to form his own questions before he begins the text, makes predictions, and pushes himself to read quickly (p. 55). If the student were given the opportunity to skim and scan at his own rate, he would spend an inappropriate amount of time reading every word in order to find the correct answer to the question. This strategy defeats the purposes for which skimming and scanning are intended. Timing the exercises forces the student to work quickly. Once again, the two goals of the program are: 1. to provide the student with a step by step reading skills CAI program that directly relates to what he is learning in the regular reading curriculum, and 2. to force the student to work
35

through a selection of his choice as quickly as possible without sacrificing accuracy. Instructional Prerequisites

The program assumes that the student is familiar with the purposes of the three reading subskills: skimming, scanning and guessing. The student should also be able to apply these subskills to materials that range from 10th to 12th grade reading levels (Fog and Fry indices used). This program requires prior classroom exposure to the subskills before it can be used by the student. Program Limitation The program was designed for the Apple II computer series and was written in Applesoft Basic. Although much of the current educational software is written in Basic, there are more versatile and efficient languages that one can use when programming. The tasks of devising tabulations of correct and incorrect answers and assisting the student when wrong answers are given were not included because of programming complexity and lack of disk space. The more efficient language, Super Pilot, has programmed routines for tabulating correct and incorrect answers along with more extensive computer-student dialog capabilities. Super Pilot was specifically designed for teachers wishing to author their own CAI software and includes a powerful lesson, graphics, and sound editor. This language is being considered by the writers for use in the development of reading and writing software programs. Another program limitation was with the hardware itself. The designers' Apple II could only print in upper case letters. Because students may have difficulty reading a screen of text with the same-size letters, the designers decided to use a commercially available program which allows for upper and lower case capabilities. Also, in future programs, another commercially available high resolution graphics program will be used. Content and Structure The name of the program is CARI and stands for Computer Assisted Reading Instruction. The student is introduced to the program when he sees the name of the program gradually appear on the screen. He listens to a few bars of Bach's Minuet in G as the letters C-A-R-I are displayed. The student's gaze follows the computer as it spells out what the letters signify. From the very start, CARI is attempting to make reading enjoyable by providing interesting visual and sound displays. The designers were cautious, however, in their use of graphics or sound effects to reinforce correct answers. While external reinforcers may be useful, they slow the pace of a lesson and give no direct support to the skills being studied. Programs that result in an uncomfortably slow pace waste valuable instructional time, particularly when the reinforcers are irrelevant to reading (Geoffrion and Geoffrion, 1983, p. 45). The corpus of the program consists of three subskill activities and three integrative activities. The materials range from company letters to excerpts from the company office manual. The types of material are listed below. 1. An ARAMCO internal letter 2. An ARAMCO external letter 3. An ARAMCO office manual excerpt 4. A textbook passage 5. An "Arab News" article 6. An article from ARAMCO World Magazine

In choosing the material to be used in the readings, several points had to be kept in mind. The first and foremost was to provide the student with appropriate reading material. Appropriate here refers to material that is relevant to the student's particular interests and needs. Mackay and Mountford (1979) state that when the reading selection is appropriate, there is an increase in motivation on the part of the student (p. 121). If the information contained in a text is seen as relevant, the student will more actively participate in the reading experience. A second factor that had to be considered was the reading difficulty level. Fog and Fry readability indices were used, in part, to determine this. The material ranged from tenth to twelfth grade. While useful in principle, readability estimation indices have many weaknesses in theory and ease of use. The largest weakness is the belief by many that readability is a well-defines construct that can be accurately measured. At best, readability formulas are crude judgements of reading difficulty (Geoffrion and Geoffrion, 1983, p. 104). Because of the writer's skills approach to reading, rhetorical concerns such as cohesion, clarity and simplicity of style and number of inferences required were also taken into consideration. Using intuition and the writers' knowledge of what the student had been exposed to previously in the regular reading curriculum, the writers examined syntactic complexity and lexicon in the various texts before deciding on a certain one. A third factor in choosing a text was length. Since all of the texts, with the exception of one, were unmodified, it was necessary to choose texts that were short enough to fill approximately two to five screens on the computer. The writers needed to give the student enough material to skim and scan, but not overwhelm him with a seemingly endless series of pages. The material is presented one screen at a time. The old teletype programs scrolled information onto the screen, printing one line at a time in a continuous roll. Lathrop and Goodson (1983) recommend not scrolling. The student should be presented with one screen, or page, of information at a time (p. 60). The writers used the onescreen-at-a-time technique in programming CARI. The fourth factor was variety. All too often in ESP curricula, there is the tendency to bombard the student with material in his field of study with no attention
36

paid to the style of the matter nor its particular focus. The writers selected different kinds of material focusing on business and education-related topics. Not only are the topics different, but the types of reading material are too. The student can choose to read a letter, an office manual excerpt, an article from a newspaper, or an excerpt from an English language textbook. The texts were picked so as to give the student a little more information than he already possesses. The intent of the writers was to teach the skills involved in reading, not the subject of the text itself. Texts should be avoided that are so advanced that the student is confused by both the language and the complexity of the information itself (Mackay and Mountford, 1979, p. 122). Once decisions were made on the materials to include in the program, a format had to be established for the specific subskill areas. The first menu presented to the student is called a Reading Subskills Menu. There is a choice of three options. The student can work on skimming, scanning or guessing. Once he selects the subskill he wants to work on, an attractive graphic is displayed on the screen which spells out the words skimming, scanning or guessing. An exercise instruction page follows the graphic. Four steps are outlined for each of the subskills. Words to help the student remember the four steps are highlighted in white. The student is then asked to review the four steps by typing n the four key words. Audio and visual reinforcement is immediately provided for each correct answer. If the student fails to provide the four key words, he must return to the instruction page and review the steps. The second menu is called the Text Menu. The student sees a list of three different types of texts from which he chooses one. The student is allowed a lot of flexibility in the selection and the sequence of tasks. He is even allowed four options

for the amount of time he wants to spend on each screen of text. Time ranges from ten to forty seconds per page of text. Once he had decided on a time, a question about the text will appear on the screen. In the case of skimming, the computer will tell the student that he will be asked to find the subject and the main idea of the text. For scanning, the computer will tell him he will be asked four questions. One of the questions will appear on the screen for the amount of time the student has already requested. Then the text will come up. This is also timed at 10, 20, 30, or 40 seconds per page depending on the student's preference. At the end of each text, the program will ask him if he is ready to answer questions. If the student indicates that he is not ready, the program will give him two options. He can choose a slower time or go back and review the steps for the particular skill he is working on. If he inputs that he is ready to answer the questions, one of the questions will appear on the screen followed by four options. The student is allowed as much time as he wants at this point to input his answer. Immediate feedback is given for both correct and incorrect answers. If the student answers incorrectly, he can choose a slower time and the question-text-ready-question sequence will begin again. The kinds of questions that are asked depend on the subskill involved. For skimming, the information required of the student is in the form of two questions. The first asks for the subject of the text. The second asks for the general idea. Each of the questions has four options from which to choose. The scanning questions are much more focused and concentrate on minute points. Since the search is more focused, the information needed is very specific. The questions asked are ones that require the student to find a date, a number, a place or a specific answer to a general comprehension question. Again, there are four choices, one of which is correct. The questions in the guessing subskill were more difficult to devise. Clarke and Silberstein (1979) in their article on psycholinguistic principles in the ESL reading class remark that vocabulary exercises are perhaps the easiest to write and the easiest to abuse. The three types of vocabulary attack strategies they discuss are obtaining meaning from context, from morphological analysis, and from monolingual English dictionaries. They feel that of the three strategies, guessing vocabulary from context is perhaps the most important. In context work, there are not only semantic clues of which students can be made aware, but also syntactic clues which can help (p. 57). The designers of the program attempted to incorporate in the questions techniques that would make the student aware of guessing strategies and lead him to the correct answer. The three questions can be categorized as: 1. function, 2. context, and 3. guessing. The function question involves looking at the vocabulary word in the white box in the passage and determining if it is a noun, verb, adverb, or adjective. The context question tells the student to look around the word in the white box. The program then asks the student to find a word that may help him guess the meaning of the new word. The part of the passage in which the vocabulary word appears is displayed on the screen once again and this time the student must actually type in the word he feels will help him guess the meaning of the new vocabulary item. This is the only place in the program where the student needs to input something more than a letter, number, or his name. To avoid lucky (vs. intelligent) guessing, the student is then asked to input another word that may help him guess the meaning of the new word. The words that the program has stored as appropriate responses exceed the number the student is required to input and include both semantic and syntactic clues. Appropriate feedback is provided for correct and incorrect answers. Spelling mistakes are counted as inappropriate responses, and repetition of a correct response is
37

not permitted. The final question asks the student to guess the meaning of the word. Hopefully, by this time, the student is ready to make an educated guess. The last reading activity on which the student can work is what the writers have termed an "integrative exercise." The student is not allowed to work on this until he has done at least one exercise in each of the subskill areas. The integrative activity consists of a choice of three texts. Once the student has selected a text, he works through the text in the three subskillsskimming, scanning, and guessing. He must pick out the subject of the text, the general idea, four answers to questions requiring specific information and two guessing exercises. Each of the guessing exercises contains three parts: function, context, and guessing. The purpose of including an integrative activity in CARI was to allow the student to combine the three subskills into one activity. The whole program could be described as a combination tutorial/drill and practice. Tutorial is classified as a mode in which information is presented. The student must then either demonstrate that he understands the information presented or seek further explanation. In the tutorial mode the computer is highly responsive while at the same time it advances the instructional goals of the program. In drill and practice a problem is printed, the student responds with an answer which is immediately scored, and the computer then gives either positive reinforcement or hints for correction. A fixed number of problems may be given, or the process may continue until mastery is achieved. Often drill and practice programs are used as tests by suppressing feedback (Doyle, 1982, p. 99). The instruction pages for each of the subskills is the tutorial part of the program. The student is asked to perform specific, tasks when skimming, scanning and guessing. Then he is tested on his knowledge of the steps involved in the process of skimming, scanning, or guessing. The rest of the program is in a drill and practice format although the guessing questions may be viewed as tutorial because the student is asked to look at the function and the context of the word before he is allowed to guess its meaning. The advantage of this particular program is that the student can begin in any of the skill areas and choose a time suitable for his reading needs. Timing a reading task to suit individual reading capabilities in a classroom situation is difficult for a teacher. However, the program has the ability to set time limits to allow for individual differences and then if the individual finds he can or cannot accomplish the reading task, the time can easily be readjusted. Another strength of CARI is the fact that it is "user friendly." Well designed courseware is "user friendly," making it easy for the student to work through the program without being inhibited by the mechanics of computer operation (Lathrop and Goodson, 1983, p. 61). Clear and explicit step by step instructions help the student concentrate on the subskill task. Without these simple instructions, a student can spend a lot of time trying to decipher the directions. Implementation and Evaluation Having developed a reading CAI program, the next step is to evaluate and implement the program to make sure it is educationally sound. Many of the criteria used to evaluate instructional materials such as textbooks, workbooks, films, filmstrips and other audiovisual media can be applied when evaluating courseware. One method of determining the educational value of a particular program is to field test it. Field testing the program with participants in Level C should tell the writers whether the program does, in fact, do what it purports. An evaluation form will be given to each student who uses CARI (see Figure 1). The questions will try to elicit information about the clarity of instructions, the reading difficulty level, the subject matter itself, and the importance of the program as it relates to the regular curriculum. From this information the writers can begin revising the program.

In addition to the student evaluation form, the teachers of Level C will be asked to review the program and fill out an evaluation form. The evaluation form chosen is an adaptation of the MicroSIFT, Computer Technology Program, Northwest Regional Educational Laboratory courseware evaluation form. On this form there are three main areas: 1. content, 2. instructional quality, and 3. technical quality. Teachers involved in the evaluation process will make judgements based on their teaching experience in Level C. The writers themselves will observe student use of the package. This detailed observation of student use may be the most important element of the evaluation process. Sitting with the student and answering his questions will allow the writers to make the necessary revisions. Good editing and testing facilities are important in the developmental phase and later on for the maintenance and improvement of courseware. During the validation tests with students, the storing and analysis of data to measure course effectiveness and student performance are absolutely essential if the courseware is to become educationally effective (Hallworth and Brebner, 1980, P. 201). Data will be collected and recorded on a one-to-one student-observer basis. An interesting feature of CARI is its capability to easily adapt and permit other ESP materials to be inserted into the program. A framework has been provided allowing one text and its accompanying questions to be substituted for another. Considering the high price reading software usually commands, it is not cost effective to purchase a program that exhausts its content after one or two activities. A good program will have the capability of adding or deleting items. When modification options exist, the program becomes much more
38

Student Evaluation of CARI Check the Appropriate Box 1. I completed a skimming activity. I completed a scanning activity. I completed an integrative activity 2. The directions were clear and easy to understand. 3. The texts were interesting. 4. The texts were at my reading level. 5. The texts and exercises were similar to the ones I do in my reading class. 6. The texts were job related or ARAMCO related. 7. The length of the texts was appropriate for the exercises. 8. I found reviewing the texts helpful in answering the questions. 9. I have a better understanding of the reading skills as a result of doing the exercises in CARI 10. There was a wide selection of texts to choose from. 11. I pushed myself to read as quickly as possible 12. The time I preferred was * 10 sec. * 20 sec. * 30 sec. * 40 sec. What I did not like about CARI was______________________________________ What I liked about CARI was____________________________________________ Figure 1: CARI Student Evaluation Form. YES * * * * * * * * * * * * * NO * * * * * * * * * * * * *

useful for the teachers and students involved in the reading curriculum (Geoffrion and Geoffrion, 1983, p. 160). A computer program like CARI that aims at the process of reading rather than the product, should benefit all who use it. If this process becomes a dialog between the student and the authors of the material rather than a process of passive information osmosis, then there should be educational value to the program. The writers are hopeful that CARI will fulfill the purposes and goals it set out to accomplish. Conclusion The area of courseware development is an exciting one. The writers feel that if effective courseware is to be developed, then educators and CAI program writers need to work together. There is a need for a coordinated effort to build effective CAI curricula with many groups contributing and exchanging materials. In order to avoid duplicating efforts, educators and program writers need to join forces. These efforts are necessary because developing courseware is a time consuming activity. For inexperienced programmers like these writers, CARI took over 250 hours to program one diskette's worth of material. This paper has been an attempt to share with the readers an experience two educators had in developing courseware to fit their students' particular needs. The frustrations encountered and the knowledge gained in the process of producing CARI made the writers aware of the importance of effective educational courseware. Teachers who are not interested in the actual programming of materials are advised to look very carefully at the courseware they are thinking of purchasing. Effective CAI courseware can make the difference! REFERENCES Bork, Alfred. 1980. Interactive Learning. In The Computer In the School: Tutor, Took, Tutee, Robert P. Taylor (Ed.), 56-66. New York: Teachers College Press. Clarke, Mark A. and Silbertstein, Sandra. 1979. Toward a Realization of Psycholinguistic Principles in the ESL Reading Class. In Reading In a Second Language, Ronald Mackay, Bruce Barkman and R.R. Jordan (Eds.), 48-65. Rowley, Massachusetts: Addison-Wesley Publishing Company. Doyle, Danny. 1982. PETCAL: A System for Development and Delivery of Computer-Assisted Instruction. InMicrocomputers in Education, Christopher Smith (Ed.), 99-107. New York: Halsted Press. Geoffrion, Leo D. and Geoffrion, Olga P. 1983. Computers and Reading Instruction. Reading, Massachusetts: Addison-Wesley Publishing Company. Hallworth, H.J. and Brebner, Ann. 1980. Computer Assisted Instruction in Schools: Achievements, Present Developments, and Projections for the Future. (Report No. IR 009 102). Washington D.C.: Educational Resources Information Center, June, 1980. Jenks, Frederick L. 1981. Learners' Needs and the Selection of Compatible Materials. In The Second Language Classroom, James E. Alatis, Howard B. Altman, and Penelope M. Alatis (Eds.), 211-225. Lathrop, Ann and Goodson, Bobby. 1983. Courseware in the Classroom. Reading, Massachusetts: Addison-Wesley Publishing Company.

Linowes, David F. Tasks for Teachers, Parents and Industry. 1983. Paper presented at the Computers and Human Learning International Symposium. Washington, D.C., December, 1983. Mackay, Ronald and Mountfourd, Alan. 1979. Reading for Information. In Reading In a Second Language, Ronald Mackay, Bruce Barkham, and R.R. Jordan (Eds.), 106-141. Rowley, Massachusetts: Newbury House Publishers. MicroSIFT, Computer Technology Program, Northwest Regional Educational Laboratory. 1982. Evaluation's Guide for Microcomputer-Based Instructional Packages. Eugene, Oregon: International Council for Computers in Education. Poirot, James L. 1980. Computers and Education. Austin, Texas: Sterling Swift Publishing Company. Thomas, J.L. 1981. Microcomputers in the Schools. Phoenix, Arizona: Oryx Press. Author's Address ARAMCO Box 10617 Dhahran, Saudi Arabia