Computer Assisted Instruction in Reading for Students with Learning Disabilities: A Research Synthesis.
The essential skill of reading, including decoding and comprehension, has not been learned by all. The number of children identified with learning disabilities continues to increase in the United States. Of the identified children, the majority are identified in the area of reading. Educators continue to search for interventions to improve students' reading skills. One format that has provided promise for students with Learning Disabilities (LD) is computer assisted instruction (CAI). To evaluate the extent to which this promise has been realized, this literature review was conducted. A methodical search of the literature on CIA in reading interventions for students with learning disabilities yielded 17 studies. The studies were evaluated by type of computer instruction (drill and practice, strategy, and simulation) and type of reading intervention (prereading, word recognition, vocabulary/language, and comprehension/higher order thinking skills). Results indicate that most CAI programs in reading for this po pulation employ drill and practice procedures, followed by strategy instruction, then simulation. The area of reading intervention focus was evenly split between word recognition and reading comprehension, followed by language/vocabulary, then prereading skills instruction. In many studies CAI was found to be a medium in which children improved reading skills. Those studies demonstrating significant differences favoring a CAI reading intervention, employed effective teaching practices. Several characteristics of effective practices using CAI are highlighted here. Implications for future research employing CAI for students with disabilities in reading are presented.
In many respects, the outlook for children who experience learning difficulties in schools is bleak. The problem for schools and society is serious. About 23 million adults have basic skills at a fourth grade level or below, and are classified as functionally illiterate. An additional 35 million adults are semiliterate (skills below eighth grade ability). Illiterate and semiliterate adults account for 75% of the unemployed (Orton Dyslexia Society, 1986). Approximately one-fourth of the students currently enrolled in kindergarten through grade 12 will fail to complete high school (Cegelka, 1995), and many of these students are eligible for special education services. More than 5.1 million children with disabilities from (birth to 21) have been identified with a disability (Eighteenth Annual Report to Congress, 1996). This number represents 12% of the school population. Currently, 52% of children with disabilities have been identified as Learning Disabled (LD), a percentage of school children that has grown dra matically (from 23%) since child counts began in 1976-77 (Heward, 1996). Clearly, the essential skill of reading (decoding and comprehension) is not successfully taught or learned by all.
Concerns about literacy skills have become profound in America to the extent that our current political agenda now addresses reading education. President Clinton has established a goal for all school children to read by grade 3. To emphasize the importance of correcting and decreasing illiteracy rates, the White House is engaged with universities around the nation to institute tutoring programs to support school instruction and improve children's reading skills (The America Reads Challenge, 1997).
Specialized efforts have been developed in an effort to remediate academic skills problems, including reading. Title 1 (formerly chapter 1) and Special Education are two prevalent service providers in public schools specifically designed to remediate and educate students with reading difficulties and disabilities. Seventy-five percent of students identified for Title 1 services, require assistance in reading (followed by math, and language). Reading is a significant problem for students with learning disabilities also. Bender (1995) reported that 85-90% of students with learning disabilities are eligible for services specifically in reading. Reading is the foundation of curriculum pursuits; students unable to read with success will experience difficulty in most curriculum areas (Zoref, Glang, & Hall, 1993). Poor readers frequently lose out on content learning because of inability to acquire knowledge from reading texts (Montali & Lewandowski, 1996). Obtaining knowledge from the printed word (i.e., comprehens ion) is the fundamental purpose of reading. To do so, the reader must adequately translate the text (letters and symbols) into the language they represent (i.e., decode) (Grossen & Carnine, 1991).
A pivotal problem for students with reading difficulties is accurate, fluent word identification skills (Morrison, 1987; Stanovich, 1988). An inability to fluently decode words further limits the ability to read independently and significantly interferes with comprehension (Adams, 1990; Stanovich, 1986). Reading is an aggregate of highly interactive processes for word decoding and comprehension (Larsen, 1995). In a synthesis of beginning reading research completed by Chard, Simmons, and Kameenui (1995) four areas of reading skills for children with diverse learning abilities were identified as essential. Simply stated, these areas include (a) phonological recoding (translating a word into its parts), (b) alphabetic understanding for word recognition, (c) language understanding based on word recognition, and (d) strong word recognition which functions as a key for reading comprehension and higher order thinking skills. Reading is a complex process, thus it follows that successful acquisition and application o f reading requires systematic and planful teaching (Carnine, Silbert, & Kameenui, 1997; Chard, Simmons, & Kameenui, 1994).
An unfortunate paradox, given what we know about the incidence of reading difficulties and remediation practices, is that students with learning disabilities actually receive less reading time (instruction and practice) than their peers (Allington & McGill-Frazen, 1989; Allington & Walmsley, 1995). This phenomenon has been coined the "Matthew Effect" (Stanovich, 1986; Walberg & Tsai, 1983) - The rich get richer (those who read well, read more) and the poor get poorer (those who don't read well, read less). Students with LD in reading do not practice the tasks of reading enough.
Teachers of students with learning disabilities in reading continually look for additional strategies and procedures to obtain success in decoding and comprehension skills. One logical solution is to provide more instructional time and practice for those who most desperately need to learn reading skills. This goal is not easily attained. For example, teacher shortages, school financial limitations, and time factors all impact the ability to provide optimal services to each LD student in reading. One potential solution to provide additional instruction and needed practice comes to us through computer technology. Research and program development in the last two decades includes specific instruction and practice of reading skills using computers. In particular, many educators and researchers enlist the technology of computer assisted instruction (CAI) (Rieth & Semmel, 1991). With the advent of this technology and greater school access to computers, well designed CAT with a reading emphasis may help needed instru ction and practice for students with learning disabilities.
The computer, in tandem with well developed software, has become a potential supplement for classroom teachers when providing instruction concurrent with practice in specific reading skills and strategies. A number of computer software programs have been developed specifically for students with reading difficulties. Additionally, a number of studies have been conducted in the area of reading disabilities and CAI. In this article we present a summary of 17 experimental studies spanning 17 years, specifically addressing CAI in reading for students with learning disabilities.
The appeal of CAI for improving and increasing instruction is based on the belief that teacher-based instruction could be transferred to computer applications with additional advantages. Practices known to be effective for instructional design and delivery for teachers could be applied to computer programs. Specifically, implementation of effective teaching practices such as, explicit, strategic and scaffold instruction, engaged time, success rate, corrective feedback, (Ellis & Worthington, 1994). CAI can (a) instruct the difficult-to-teach at an individual pace, (b) provide immediate feedback, (c) provide instructive and consistent corrections, (d) allow for extensive rehearsal or needed repetition, and (e) be highly motivating (Rieth & Semmel, 1991; Woodward et al., 1984).
When computers were first introduced into schools, programs were available for two basic purposes; drill and practice of basic skills (e.g., math facts, letter identification, and building reading rate), or games. Students' time on the computer was often used by teachers as a reinforcement activity. The use of drill and practice computer activities was particularly extensive for students with disabilities. More recently, applications including computer simulations, videodisk technology, and computer assisted instruction have been created to teach students with disabilities.
Current trends in computer assisted instruction. The use of CAI technology in classrooms has been envisioned to fit needs beyond additional skills practice. The presumption of using CAI in classrooms was that teachers would be able to reallocate teaching time for students (i.e., both the teacher and computer could actually engage students in instruction). Students needing more instruction could obtain that instruction with the addition of CAI into the classroom. For example, Carnine, (1989) noted that the use of technology could preserve or extend teacher instruction and reduce the time and effort required
by teachers to implement interventions, thus leading to more instructional interactions between special education teachers and their students. Other technological interventions have been designed to reduce implementation demands of the intervention rather than target basic instructional skills (e.g., Fuchs, Fuchs, Hamlett, & Hasselbring, 1987). CAI programs have been designed and implemented in many curricu lum areas to meet such instructional needs. Researchers are investigating the effectiveness of CAI under classroom conditions to determine what impact it may have on achievement for students with skill deficits.
Educators and researchers in special education predicted a positive impact by using technology for students with disabilities. Torgensen (1986) stated that "computers have the capacity to deliver motivating, carefully monitored, individualized, and speed-oriented practice in concentrations far beyond those available in traditional instructional format" (p. 159). The computer was seen as a device that could provide the needed task practice of skills for students needing multiple trials to obtain mastery (Larsen, 1995). In spite of the increase of computers in classrooms and the apparent value of CAI for students with LD, empirical evidence of the efficacy of CAI is somewhat ambiguous.
The focus of this article applies specifically to software applications designed for Computer Assisted Instruction in reading for students with learning disabilities. In compiling this literature, it has become apparent that the term CAI is not universally defined. Researchers have embraced varying definitions for computer programs used in education. For our purposes we have adopted the definition of CAT from Posgrow (1990) who defines CAI as, the software designed to teach toward a curricular goal, to provide instruction and practice toward the achievement of an immediate learning objective.
In 1986, Harper and Ewing noted that in the area of special education there existed little evidence regarding the effectiveness of CAI for students with learning disabilities. During the 1980s and 1990s a number of articles appeared regarding the impact of CAI in the special education classroom. Few of those articles were research-based. Our intent in this review was to summarize empirical research conducted using CAI reading interventions with students identified as having learning disabilities.
The authors established a set of parameters for articles on CAI to be included in this literature review. The following sections describe the selection criterion and procedures used for our examination of the literature.
Selection of Studies
The search for CAI research studies was conducted using the following criteria for selection: First, each article had to be empirically based research in which the authors employed an experimental or quasi-experimental design to evaluate the effectiveness of CAI in reading. Experimental design was determined by the description of research features as recommended by Gall, Borg, and Gall (1996). Studies which included assessment development or study skills for reading were eliminated (e.g., Fuchs, Hamlet, Fuchs, Stecher, & Ferguson, 1988). Second, articles were published in refereed journals with a publication date between January 1980 and November 1997. The date criterion was established because the first studies of CAI in educational settings began appearing in the early 1980s. Third, CAI for reading was used in each study for students with LD. In some instances, labels such as learning handicapped or mildly handicapped were used in studies. In such cases, studies were included for review if there were indica tions that these students would have been identified as LD in another district or state (e.g., reported significant ability-achievement discrepancies).
Research articles included in this review were obtained using several sources and procedures. Initially a computer search using the Educational Resources Information Center and Psychological Information abstracts was conducted using the following descriptors: computers, computer assisted instruction, technology, learning disabilities, learning problems, and reading. Additionally, the authors conducted an ancestral search using reference lists from the articles found in the computer search. Next, several textbooks were examined for studies related to the topic. All articles identified with these procedures were analyzed to see if they met the selection criteria.
As a second step, the above reference lists were examined in order to develop a list of journals in which articles of this type might be located. Of the 22 journals identified in this manner, 3 are specific to learning disabilities, 5 relate to special education, 7 are specific to technology and computers, and 7 are general education journals. A final search was conducted by reviewing the table of contents for each journal published within the last 17 years. Relevant articles were examined in relation to the selection criteria.
Analysis Procedures and Findings
Our search of the literature yielded 17 studies which met the selection criteria. Our research review included 13 primary sources (Bahr, Kinser & Rieth, 1991; Collins, Carnine, & Gersten, 1987; Harper, & Ewing, 1986; Hebert & Murdoc, 1994; Helsel-Dewert & Van Den Meiracker, 1987; Jones, Torgesen, & Sexton, 1987; Keene & Davey, 1987; Lin, Podell, & Rein, 1991; Manis, 1985; McGregor, Drossner, & Axeirod, 1990; Thorkildsen, Waters, Cohen, & Torgensen, 1988; van Daal & van der Leij, 1992). These were identified as articles in which one study was conducted and discussed by the authors. Additionally, there were 4 secondary source studies (Woodward, Carnine, & Coffins, 1988; Woodward et al., 1986). The secondary sources were articles in which multiple CAI studies completed by one or more of the authors were presented. The articles reviewed included 1 quasi-experimental design (as described by the authors, no random assignment), 2 single-subject alternating treatment studies, and 14 experimental studies (usually, tre atment/control, pre! posttest design).
Each of the reading CAI studies was reviewed to (a) identify the number and grade level of subjects participating in each study, (b) describe the number and length of CAI sessions, (c) describe the dependent measure(s), and (d) summarize the research outcomes. Additionally we categorized the studies on the following two components (a) type of CAI programming - drill and practice, strategy instruction, or simulation; and (b) type of reading intervention - phonological, word recognition for fluency, vocabulary-word meaning, or reading comprehension and higher-order thinking skills. Detailed information about each source is provided in Tables 1 and 2.
Initial analysis of the articles was conducted by the third author based on definitions and parameters provided by the second author (see Table 1). Reliability of the analysis was obtained by having the first and second authors conduct an identical analysis on 15 of the 17 studies. Initial agreement between the authors was 89.6% and disagreements were discussed to consensus. The following sections contain our analysis of compiled results with additional explanations of the studies reviewed.
Population sample sizes. A total of 569 students participated in the studies evaluated for this article. The number of students in each study ranged from 2 to 93. The mean number of students per study was 32 (SD = 22.1). Studies with the smallest number of subjects (Hebert, & Murdock, 1994; McGregor et al., 1990) used a single subject multiple-baseline design (n = 2). Several studies (n = 12) with a larger number of subjects, were group design studies with varying experimental conditions (two groups). Three studies used a three-group design, which included two treatment groups and a control group.
Subject characteristics. All participants in the research reviewed were school-aged students ranging from kindergarten to grade 12. One study included kindergarten students, 5 studies were conducted at the primary level (grades 1-3), 3 studies investigated intermediate (grades 4-6) students, 3 studies included middle school aged students (grades 6, 7, and 8), and 5 studied high school students (9-12). All studies focused on intervention impact for students with disabilities. In the majority of studies (14), all subjects were identified as LD with instructional deficits in reading. The three studies not categorized in the above grouping included students with LD and control groups of students without disabilities, some considered "at-risk" for reading failure, and/or general education students who served as control group or comparison subjects. In all studies, the primary focus of research was improving learners' reading skills.
Session duration. When analyzing each study in relation to the number and length of sessions, there was much variability (e.g., 1-50 sessions, 10-40 minutes). The mean number of sessions using CAI was 17 (SD = 15). Most (71%) CAI interventions lasted between 5 and 20 sessions. Only 3 of the 17 studies (17%) had greater than 30 sessions. The remaining studies did not report number of sessions. Duration of each session was also analyzed in minutes per session. Most studies (7) reporting time in session had students involved with the CAI for between 15 and 20 minutes per session. Only one study had a 10-minute session, and 3 studies had sessions from between 25 and 40 minutes in length. Six studies did not report the amount of time students were engaged in the CAI sessions.
Intervention procedure. While the research has generally supported the use of instructional technology in the schools, researchers have only begun to investigate specific implementations which include (a) strategy instruction, (b) drill and practice, (c) simulations, (d) tutorial (e) writing, and (f) problem solving (Rieth & Semmel, 1991). We analyzed the 17 CAI research articles for students with LD in reading by intervention type. Ten studies reviewed were drill and practice in nature, while six programs used strategy instruction (see Table 2). Additionally, simulation was used in one study. None of the CAI reading programs for students with LD analyzed for this review used tutoring, writing, or problem solving. The three application types used are described below.
Drill and practice, is the most frequently seen and used form of CAI across applications (Larsen, 1995; Rieth & Semmel, 1991) and in this review. The drill and practice procedure is designed to provide extensive rehearsal and consolidation of skills already in the student's repertoire. Well designed drill and practice CAI also includes corrective feedback and reinforcement with the major focus on repetition of a skill (Larsen, 1995). Drill and practice typically serves as a supplement to other forms of instruction. Findings from the studies reviewed here are presented in the discussion section.
Strategy instruction involves teaching a specific method, approach, or series of steps to learn by way of the computer (Woodward et al., 1988). Students in reading for example, may learn specific steps to understand and recall information presented in a passage.
Simulations are identified as application procedures; students are employing their skills in problem-solving activities. In reading comprehension for example, students may read important information that must be later used as a situation is described on the screen. Students must recall and apply that information to move forward in the simulation (Larsen, 1995; Rieth & Semmel, 1991; Woodward et al., 1986).
Targeted reading skill. Each study was also analyzed in relation to essential reading skills needed for children with diverse learning abilities, as classified by Chard et al. (1995): phonological recoding, word recognition (word reading), learner understanding of word meaning (vocabulary), and reading comprehension and higher order thinking skills (HOTS). Three of these four areas were frequently emphasized by CAI studies reviewed, facilitation of word recognition, reading comprehension, and HOTS (see Table 2). Four studies taught students word meaning or vocabulary building using computerized instruction. One study, designated for primary (K - 2) level, taught phonological awareness skills for word recognition. Alphabetic understanding for word recognition was the focus of six studies, as was reading comprehension and HOTS.
Research design. Six studies from the collected research used CAI as a treatment condition and traditional teacher-lead instruction or textbook work as a control condition. In these 6 studies, 4 yielded significant differences favoring computer instruction. The two studies in which no significant treatment to control difference was found had results, which favored traditional text or teaching procedures.
A number of studies (n = 8) used varying conditions using CAI to evaluate the effectiveness of components of the computer programming on student learning. These variables consisted of (a) elaborated correction procedures, (e.g., teaching corrections with built in repeated practice, versus telling the answer and moving on in the program); (b) varying the size of the learning set (e.g., number of items to learn per lesson or unit); (c) computer output variations (e.g., synthesized speech output versus taped speech output); and (d) training to use CAI specifically with daily reading instruction versus use of CAI as a separate instructional episode). In each of the elaborate versus basic correction CAI studies (n = 5), students in the elaborated correction group outperformed peers in the basic correction group. The remaining two-treatment studies with variations on CAI favored the more elaborate intervention, with the exception of training teachers to use CAT as an integrated part of reading instruction. Here, th ere were not significant differences between groups.
Finally, three studies employed two treatment conditions, and a third group as a no-treatment control. These studies were evaluating alphabetic understanding and reading vocabulary. The alphabetic understanding research varied treatment conditions in (a) computer output mechanisms (computer feedback auditory or written), and (b) student output structure (verbal or typed response). Interestingly, these studies yielded no significant differences between the three groups on the dependent measures. The third (two-treatment and control) study varied speech output mechanisms by the CAI program (computer "voice") versus the control which had no speech for feedback. The speech outcome groups had significant impact on student outcomes over the no speech control group.
In the studies reviewed here many students demonstrated improvement when using CAI (13 of 17 studies). In other words, students with LD receiving CAI in reading increased performance in reading decoding or reading comprehension. There does not appear to be a pattern to the type of CAI, or area of reading instruction for the 4 studies with no between-group differences. These studies dealt with alphabetic understanding (n = 2), and reading comprehension (n = 2). Additionally, 2 studies used drill and practice and the other 2 used strategy instruction. In each of these categories, there were more studies with successful results (see Table 2).
Although successful acquisition of the complex processes of reading appears to be incidental for some children, research literature has demonstrated that learning to read (decoding and comprehension) requires more explicit instruction integrating many complex processes (Chard et al., 1995). Many students, especially those with learning disabilities appear to require systematic and planful instruction. Schools are replete with examples of children experiencing serious difficulty learning to read successfully.
Our analysis of the literature yielded 17 studies specifically designed to evaluate CAI in reading instruction for students with learning disabilities. Each study incorporated empirical research practices to evaluate outcomes of student performance when using CAI. Overall research outcomes for CAI in reading for students with LD are encouraging and suggest a variety of systems and areas of instruction may be addressed with applications of CAI. From this review of the research, we have drawn three conclusions for reading instruction for students with LD in reading: (a) we may be able to help overcome the "Matthew Effect" by having CAI technology available in classrooms, (b) the systematic instructional procedures found to be so effective for reading instruction appear to be available with carefully designed CAI software, and (c) we are again reinforced by the research to apply systematic, elaborate corrections for most efficient and effective learning. Each of these conclusions will be discussed below.
Addressing the "Matthew Effect". Several benefits to CAI, reinforced by the research presented here, include supplementing the instructional episode provided by the teacher with CAI to provide structured practice. None of the researchers recommended replacing teacher directed reading instruction with CAI. Rather, the authors made recommendations to have educators use CAI reading applications for students with LD in coordination with teacher directed instruction. As stated earlier, most students with disabilities require substantial practice to obtain and master skills. The application of CAI as supplemental rehearsal (also known as drill and practice) is frequent. Ten of the 17 CAI studies employed a drill and practice technique (7 of the 10 studies had significant outcomes favoring CAI over other interventions such as traditional teaching, workbooks).
Students with problems in reading, be it decoding or comprehension, need a substantial amount of practice to obtain and then maintain skills. More practice, in fact, than their "normally-reading" peers. Unfortunately, a trend observed in classrooms is that students with reading difficulties actually read less. They have less instructional time, and less practice time (Allington et al., 1989; Allington et al., 1995). Teachers may help to provide more instruction for students with reading difficulties and improve students reading. With the implementation of CAI programs which practice phonological recoding, word recognition, word meaning, or higher order thinking skills, students may be assigned additional instructional time to interact on the computer using a program specifically designed to provide teaching and practice in that skill area. Thus, providing more optimal services for students with LD in reading.
Systematic instructional procedures in CAI. Education literature has substantial research supporting the use of computers in classrooms (e.g., Reith & Semmel, 1991; Woodward et al., 1988). The literature reviewed for this article has demonstrated that there exist programs that employ effective teaching strategies to teach and practice reading skills for students with LD. Several (n = 13) of the studies reviewed here specifically employed practices in the software cited in the effective teaching research. Not surprisingly, each of these studies found significant differences favoring CAI. We strongly advocate the systematic application of effective teaching principles in the design of software for CAI in reading.
Elaborate feedback. In line with the research supporting effective teaching principles, several studies reviewed here also evaluated systematic or elaborate correction procedures in CAI applications. The effective teaching research has found significant positive outcomes in teacher directed instruction favoring elaborate feedback (Ellis & Worthington, 1994; Stevens & Rosenshine, 1981). In other words, corrections that instruct and cycle students back through items to again practice an item in err. Specifically, detailed strategic corrective feedback is more effective than merely notifying students if a response was correct or incorrect (Englert, 1984; Lysakowski & Walberg, 1982). Both strategy errors and simple facts (e.g., sound of letters) are more rapidly and effectively corrected for students when an elaborate, strategic correction procedure is implemented.
Seven of the 17 studies reviewed here included specific information regarding elaborate corrections within their CAI programs. Each of these studies specified how following an error, the program provided specific strategic corrective feedback. Additionally, each of these seven programs also had students complete previously answered tasks and returned to the error before moving on to new tasks. Interestingly, each of these studies had results with significant differences favoring the use of CAI. Several studies mentioned providing the user with the correct answer following an error. However, these were not specific about a strategic correction procedure or additional rehearsal of the correct response following an error. We highly recommend that specific, elaborate, strategic, teaching corrections including a review cycle be a component of any instructional episode, including CAI. Teachers should review CAI programs in reading for the inclusion of specific correction procedures.
Summary. Teachers and school districts are intensely interested in the promise technology may provide with computer assisted instruction for students with disabilities. This appeal builds from the belief that many aspects of effective teacher-based instruction can be readily transferred to CAI applications (Reith & Semmel, 1991). Additionally, schools have increasing access to the technology needed for applications like CAI (Larsen, 1995). Computer assisted instruction has numerous applications in special education. However, having the access to hardware and software, does not result in automatic success for students with disabilities. Empirical evidence is needed which demonstrates the worthiness of CAI technology for students with learning disabilities.
Limitations. This analysis of empirical research on CAI programs for students with learning disabilities in reading suggests that most aspects of research quality have been adhered to well. As noted in the results and discussion, several studies did not provide detailed information on correction procedures; some did not report study length, duration of sessions, or generalization effects. These factors did not impair the ability to interpret the research for the most part. However, studies of effective teaching in general, and each of the studies reviewed here specifically addressing correction procedures found that elaborated corrections resulted in favorable learner outcomes (e.g., higher scores on measures, more rapid acquisition of skills). Additionally, few of the studies addressed maintenance or generalization of skills acquisition in reading. Therefore, the long term impact of the CIA is not known for this population
It is unfortunate that more studies in CIA for reading instruction are not available. We strongly recommend additional research be conducted in this field. Overall, the design of the 17 studies reviewed here was found to be within reasonable research guidelines. Certainly, with the need for instruction in reading at an all-time high for students with and without Learning Disabilities, more high quality research of CIA in reading is recommended.
Reading instruction for students with learning disabilities is an area of continual concern. It appears that interventions incorporating well-designed CAI may become a tool for teachers to not only provide adequate reading practice for students, but also additional instruction. Those CAI programs providing systematic instruction with effective correction procedures contributed most to increasing reading skills.
A surprisingly low number of empirical studies have been conducted and published in the past 17 years. The lack of empirical research on CAI in reading for students with LD has consequences for researchers and practitioners. The greatest problems arise from development of software based on theory, ideas, concepts, popular beliefs, and arguments that are not supported by the research and may not result in improved educational outcomes for learners with disabilities in reading.
CAI programs in reading that require the learner to read and use reading skills (e.g., vocabulary, comprehension) complete the practice, simulation, or strategy instruction, help students to overcome serious reading deficits. Future research and software development of CAI in reading should: (a) contain systematic inclusion of effective teaching practices; (b) specifically address the area of reading for instruction (prereading, word recognition, vocabulary development, comprehension); (c) address maintenance and generalization of reading skills beyond the CAI application; and (d) provide elaborate correction procedures which cycle the learner back through examples relevant to the instruction.
We support the continued empirical research of well-developed CAI in reading. Students with learning disabilities in reading require systematic instruction and continued practice in the all facets of reading. CAI when well designed and routinely applied in classrooms has the potential to reinforce teacher instruction and provide additional teaching to increase practice time and ultimate success in reading. Successful intervention programs, usually delivered by teachers, share several common features: small size, inclusion by choice, flexibility, view of school as a community and involvement of the community outside of school (Ellis & Worthington, 1995). These variables may be made possible in more classrooms through the use of well designed CAI.
Adams, M. J. (1990). Beginning to read: Thinking and learning about print. Cambridge, MA: MIT Press.
Allington, R. L., & McGill-Frazen, A. (1989). School response to reading failure: Instruction for chapter 1 and special education students in grades 2, 4, and 8. Elementary School Journal, 89, 529-543.
Allington, R. L., & Walmsley, S. A. (1995). Redefining and reforming instructional support programs for at-risk students. In R. S. Allington & S. A. Walmsley (Eds.), No quick fix: Rethinking literacy in America's elementary schools (pp. 19-41). New York Teachers College Press and Newark, DE: International Reading Association.
Bahr, C., Kinzer, C. K, & Rieth, H. (1991). An analysis of the effects of teacher training and student grouping on reading comprehension skills among mildly handicapped high school students using computer-assisted instruction. Journal of Special Education Technology, 11(3), 136-154.
Bender, W. N. (1995). Learning disabilities: Characteristics, identification, and teaching strategies (2nd ed.). Needham Heights, MA: Allyn and Bacon.
Carnine, D. (1989). Phonics versus whole-word correction procedures following phonic instruction. Education And Treatment of Children, 3, 323-330.
Carnine, D. W., Silbert, J., & Kameenui, E. J. (1997). Direct instruction reading (3rd ed.). Columbus, OH: Merrill.
Cegelka, P. T. (1995). An overview of effective education for students with learning problems. In P. T. Cegelka & W. H. Berdine (Eds.), Effective instruction for students with learning difficulties (pp. 1-18). Needham Heights, MA: Allyn and Bacon.
Chard, D. J., Simmons, D. C., & Kameenui, E. J. (1994). Understanding the primary role of ward recognition in the reading process: Synthesis of research on beginning reading (Tech. Rep. No. 15). Eugene, OR: University of Oregon, National Center to Improve the Tools of Educators.
Chard, D. J., Simmons, D. C., & Kameenui, E. J. (1995). Word recognition: Curricular and instructional implications for diverse learners (Tech. Rep. No. 16). Eugene, OR: University of Oregon, National Center to Improve the Tools of Educators.
Collins, M., Carnine, D., & Gersten, R. (1987). Elaborated corrective feedback and the acquisition of reasoning skills: A study of computer-assisted instruction. Exceptional Children, 54, 254-262.
Ellis, E. S., & Worthington, L. A. (1994). Research synthesis on effective teaching principles and the design of quality tools for educators (Tech. Rep. No. 5). Eugene, OR: University of Oregon, National Center to Improve the Tools of Educators.
Englert, C. S. (1984). Effective direct instruction practices in special education settings. Remedial and Special Education, 5 (2), 38-47.
Fuchs, L. S., Fuchs, D., Hamlett, C., & Hasselbring, T. S. (1987). Using computers with curriculum-based progress monitoring Effects on teacher efficiency and satisfaction. Journal of Special Education Technology, 8 (4), 1427.
Fuchs, L. S., Hamlet, C. L., Fuchs, D., Stecher, P. M., & Ferguson, C. (1988). Conducting curriculum-based measurement with computerized data collection: Effects on efficiency and teacher satisfaction. Journal of Special Education, 9, 73-86.
Gall, M. D. , Borg, W. R., & Gall, M. (1996). Educational Research: An Introduction (6th Ed.). White Plains, N.Y.: Longman.
Grossen, G., & Carnine, D. (1991). Strategies for maximizing reading success in the regular classroom. In G. Stoner, M. R. Shinn, & H. M. Walker (Eds.), Interventions for achievement and behavior problems (pp. 333-355). Silver Springs MD: The National Association of School Psychologists.
Harper, J. A., & Ewing, N. J. (1986). A comparison of the effectiveness of microcomputer and workbook instruction on reading comprehension performance of high incidence handicapped children. Educational Technology, 25, 40-45.
Hebert, B. M., & Murdock, J. Y. (1994). Comparing three computer-aided instruction output modes to teach vocabulary words to students with learning disabilities. Learning Disabilities: Research & Practice, 9 (3), 136-141.
Helsel-Dewert, M., & Van Den Meiracker, M. (1987). The intelligibility of synthetic speech to learning handicapped children. Journal of Special Education Technology, 9 (1), 38-44.
Heward, W. L. (1996). Exceptional children: An introduction to special education (5th ed.). Engelwood Cliffs, NJ: Merrill.
Jones, K. M., Torgesen, J. K., & Sexton, M. A. (1987). Using computer guided practice to increase decoding fluency in learning disabled children: A study using the hint and hunt 1 program. Journal of Learning Disabilities, 20 (2), 122 128.
Keene, S., & Davey, B. (1987). Effects of computer-presented text on LD adolescents' reading behaviors. Learning Disabilities Quarterly, 10, 283-290.
Larsen, S. (1995). What is "quality" in the use of technology for children with learning disabilities? Learning Disabilities Quarterly, 18, 118-130.
Lin, A., Podell, D. M., & Rein, N. (1991). The effects of CAI on word recognition in mildly mentally handicapped and nonhandicapped learners. Journal of Special Education Technology, 11 (1), 16-25.
Lysakowski, R., & Walberg, H. (1982). Instructional effects of cues, participation, and corrective feedback: A quantitative syntheses. American Educational Research Journal, 19 (4), 559-578.
Manis, F. R. (1985). Acquisition of word identification skills in normal and disabled readers. Journal of Educational Psychology, 77 (1), 78-90.
McGregor, G., Drossner, D., & Axelrod, S. (1990). Increasing instructional efficiency: A comparison of Voice plus text vs. text alone on the error rate of students with mild disabilities during CAI. Journal of Special Education Technology, 10 (4), 192-197.
Montali, J., & Lewandowski, L. (1996). Bimodal reading: Benefits of a talking computer for average and less skilled readers. Journal of Learning Disabilities, 29 (3), 271-297.
Morrison, F. J. (1987). The nature of reading disability: Toward an integrative framework. In S. Ceci (Ed.), Handbook of cognitive, social, and neuropsychological aspects of learning disabilities (pp. 33-63). Hillsdale, NJ: Erlbaum.
Orton Dyslexia Society. (1986). Some facts about illiteracy in America. Perspectives on Dyslexia, 13, 1.
Posgrow, S. (1990). A socratic approach to using computers with at-risk students. Educational Leadership, 47 (5), 61-66.
Richek, M. A., Caldwell, J. S., Jennings, J. H., & Learner, J. W. (1996). Reading problems: Assessment and teaching strategies (3rd ed.). Boston: Allyn and Bacon.
Rieth, H. J., & Semmel, M. I. (1991). Use of computer-assisted instruction in the regular classroom. In G. Stoner, M. R. Shinn, & H. M. Walker (Eds.), Interventions for achievement and behavior problems (pp. 215-239). Silver Spring, MD: National Association of School Psychologists.
Stanovich, K. E. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21, 360-406.
Stanovich, K. E. (1988). Explaining the difference between the dyslexic and garden-variety poor reader: The phonological-core variable-difference model. Journal of Learning Disabilities, 21, 590-604.
Stevens R., & Rosenshine, B. (1981). Advances in research on teaching. Exceptional Education Quarterly, 2, 1-9.
Thorkildsen, R. J., & Friedman, S. G. (1986). Interactive videodisc: Instructional design of a beginning reading program. Learning Disability Quarterly, 9, 111-117.
Thorkildsen, J. K., Waters, M. D., Cohen, A. L., & Torgensen, J. L. (1988). Improving sightword recognition skills in LD children: An evaluation of three computer program variations. Learning Disability Quarterly, 11, 125-132.
Torgensen, J. K. (1986). Computers and cognition in reading: A focus on decoding fluency. Exceptional Children, 53.157-162.
U.S. Department of Education. (1996). Eighteenth annual report to Congress on the implementation of IDEA. Washington, DC: U.S. Department of Education.
U.S. Department of Education. (1997). The America reads challenge. Washington, DC: U.S. Department of Education.
van Daal, V. H. P., & van der Leij, A. (1992). Computer-based reading and spelling practice for children with learning disabilities. Journal of Learning Disabilities, 25(3), 186-195.
Walberg, H.J., & Tsai, 5. (1983). Matthew effects in education. American Educational Research Journal, 20, 359-373.
Woodward, J., Carnine, D., & Collins, M. (1988). Closing the performance gap: CAI and secondary education for the mildly handicapped. Journal of Educational Computing Research, 4 (3), 265-286
Woodward J., Carnine, D., Gersten R., Gleason, M., Johnson, G., & Collins, M. (1986). Applying instructional design principles to CAI for mildly handicapped students: Four recently conducted studies. Journal of Special Education Technology, 8 (4), 13-26.
Zoref, L., Glang, A., & Hall, T. E. (1993). Developing a volunteer reading program in your school: Strategies, successes, and challenges. In J. Marr, G. Sugai, & G. Tindal (Eds.), The Oregon Conference Monograph (pp. 15-21). Eugene, OR: University of Oregon, College of Education.
Table 1 Summary of 17 CAI Studies in Reading Number and length of sessions, Subjects: length of Authors Age/grade intervention Drill and practice 7M/2F Phase 1: Harper, J.A., & (LD/MR) 20-min sessions Ewing, N.J. (1986) ages 11.5-13.5 Phase 2: (junior high) 20-min sessions (16 sessions) Phase 3: 8 weeks Phase 4: Follow-up phase (4 sessions) Phase 5: Games 20 minutes Helsel-Dewert, M., & 15M/9F(LH) Sessions: 20 Van Den Meiracker, min/day M. ages 71-117 10 weeks (1987) months (50 sessions) (elementary school) Hebert, B.M., & 3M/0F 6 days of Murdock, J.Y. (MH & LLD) treatment (1994) 9 sessions ages 11-13 Independent Dependent Authors variable variable Drill and practice microcomputer (a) percentage of Harper, J.A., & instruction (mc) correct responses Ewing, N.J. to reading (1986) comprehension workbook questions instruction (wb) (b) time on task (c) productivity Helsel-Dewert, M., & speech output (a) percentage of Van Den Meiracker, (taped)(st) correctly M. identified words (1987) speech output (synthesized) (ss) Hebert, B.M., & digitized speech (a) percentage of Murdock, J.Y. (ds) vocabulary words (1994) mastered synthesized speech (ss) no speech (ns) Authors Results Drill and practice (mc) & (wb) very Harper, J.A., & little difference Ewing, N.J. with duration of (1986) on-task behavior (mc) & (wb) low scores on both conditions Helsel-Dewert, M., & (ss)[greater than](st) in Van Den Meiracker, number of M. discrimination (1987) errors Hebert, B.M., & (ds) & (ss)[greater than](ns)in Murdock, J.Y. achievement gains (1994) Authors Reliability Drill and practice interobserver Harper, J.A., & agreement for Ewing, N.J. microcomputer = (1986) 98% workbook = 95% Helsel-Dewert, M., & none provided Van Den Meiracker, M. (1987) Hebert, B.M., & interrater Murdock, J.Y. agreement = 100% (1994) Jones, K.M., 22M/8F Sessions 15 Hint & Hunt Torgeson, J.K., & (LD & NH) min/day program word Sexton, MA. 5 days/week identification (1987) grades 3,4,5 10 weeks (H,H) Elementary Volume 2 (control) Lin, A., 45M/48F 10 sessions CAI Podell, D. m., & Mean CA. 15 min/day paper-pencil (pp) Rein, N. (1991) MH = 7.81 mildly Mean CA, handicapped (mh) NH = 8.86 nonhandicapped (nh) grade 2 measured by response time Manis, F. R. 17M/3F 4 sessions pronunciation (1985) (LD & NH) 25 min each task (pt) delayed grades 5 & 6 pronunciation task (dpt) McGregor, G., 1M age 7 29 days synthesized Drossner, D., & 1F age 6 Number of speech plus text Axelrod. S. (1990) lessons/session (st) (LD/MR) varied text alone (t) 20 min sessions daily Torgesen. J. K., 17 subjects Sessions: auditory-visual Waters, M.D., (LD) 15 min daily (av) Cohen.A.L., & 8 weeks auditory only (a) Torgesen. J. L. grades 1,2,3 visual only (v) (1988) Jones, K.M., (a) fluency (H,H) substantial Torgeson, J.K., & gains in speed, Sexton, MA. (b) accuracy accuracy and (1987) response time to generalized words Lin, A., (a) word (CAI)[greater than](pp) Podell, D. m., & recognition on posttest accuracy Rein, N. (1991) (nh)[greater than](mh) on (b) response time response time & to posttest accuracy questions Manis, F. R. (a) pronunciation (pt) & (dpt) (1985) of words performance (b) word naming strongly related to latency regularity & (c) accuracy complexity of (d) meaning of words for LD words population McGregor, G., (a) increase in (st) significantly Drossner, D., & word recognition more effective Axelrod. S. (1990) skills than (t) (b) measured by percentage accuracy Torgesen. J. K., (a) fluency and (av)=(a)=(v) wese Waters, M.D., accuracy of word equally successful Cohen.A.L., & recognition Torgesen. J. L. (1988) Jones, K.M., not provided Torgeson, J.K., & Sexton, MA. (1987) Lin, A., not provided Podell, D. m., & Rein, N. (1991) Manis, F. R. not provided (1985) McGregor, G., computer Drossner, D., & managed Axelrod. S. (1990) Torgesen. J. K., not provided Waters, M.D., Cohen.A.L., & Torgesen. J. L. (1988) Van Daal, V.H.P., & 17M/11F Sessions: Van der Leij, A. (LD) 10 min daily (1992) 15 days Mean CA = 9.7 Strategy Instruction 32M/16F 15 min sessions Bahr, C., Kinzer, C. (LD/EMR) once/week for 6 K., & Rieth, H months (1991) grades 9-12 approximately 24 sessions Collins, M., 28 subjects 5 lessons Carnine, D., & LN (N = 13) & Gersten, R. (1987) Remedial (N = 15) junior high Van Daal, V.H.P., & reading from (a) accuracy and Van der Leij, A. computer screen fluency of spelling (1992) (rs) and reading copying from screen (ss) writing from memory (wm) Strategy Instruction trained w/out (a) reading Bahr, C., Kinzer, C. strategy for comprehension K., & Reith, H integrating (1991) software into (b) attitudes daily reading toward computers instruction (t) training strategy (c) general for integrating classroom software into ecology daily reading instruction (ts) homogenous grouping (hg) heterogeneous grouping (het) Collins, M., elaborate (a) acquisition of Carnine, D., & correction (ec) reasoning skills Gersten, R. (1987) brief correction (b) transfering of (bc) skills (c) student attitudes Van Daal, V.H.P., & (rs)=(ss)=(wm) on not provided Van der Leij, A. reading outcomes (1992) Strategy Instruction (t) & (ts) interobserver Bahr, C., Kinzer, C. differences were agreement = 94%, K., & Reith, H not significant 96%, & 97% (1991) (hg) greater achievement on posttest scores than (het) Collins, M., (ec) significantly internal Carnine, D., & higher scores on consistency for Gersten, R. (1987) transfer test for form A = 90% & elaboration group form B = 91% (ec) & (bc) significant improvement on reasoning skills pre/post and maintenance (ec) more confident Keene, S., & 36 M/15F Sessions: Davey, B. (1987) (LD) training 1 sessions of grades 9-12 testing 1 week later- posttest Thorkildsen. R, & 17 M/15F average number of Friedman, S. (1986) (LD/BD) sessions = 5.68 grades: K,N = 20 l,N = 10 2,N = 2 Woodward J., Study 1: Study 1: Carnine, D., 28 MH 5 lessons Gersten R., junior high no time specified Gleason, M., Johnson, G., & Collins, M. (1986) Study 2: Study 2: Study 1: 30 MH 12 sessions reasoning skIlls high school 40 min each Study 2: health vocabulary Keene, S., & text on computer (a) reading Davey, B. (1987) (tc) comprehension text on printed page (tp) (b) learning strategies (c) attitudes toward task Thorkildsen. R, & high remediation (a) accuracy Friedman, S. (1986) (hr) (b) total time on low remediation system (lr) (c) following directions Woodward J., Study 1: Study 1: Carnine, D., elaborate (a) test of formal Gersten R., correction/versus logic pre/post Gleason, M., basic on reasoning transfer measures Johnson, G., & skills Collins, M. (1986) Study 2: Study 2: Study 1: concept (a) comprehension reasoning skIlls instruction on measures post and content maintenance Study 2: health information vocabulary Keene, S., & (tc) did not effect internal Davey, B. (1987) comprehension consistency = 73% (tp)[greater than](tc) Interrater student interest agreement = 90% - (tC) increase in 100% looking re-inspection strategy Thorkildsen. R, & (hr)[greater than](lr) in test-retest overall Friedman, S. (1986) effectiveness agreement = 90% Woodward J., Study 1: Carnine, D., significant Gersten R., difference Gleason, M., favoring elaborate Johnson, G., & correction Collins, M. (1986) Study 2: significant Study 1: difference reasoning skIlls favoring CAI instruction Study 2: health vocabulary Simulation Woodward, J., Carnine, D., & Collins, M. (1988) Study 1: vocabulary Study 1: Study 1: 24 subjects 11 sessions (MH) maximum some grades 9-12 fewer if mastery reached Study 2: reasoning Study 2: Study 2: 34 subjects 5 lessons (remedial/MH) high school Simulation Woodward, J., Carnine, D., & Collins, M. (1988) Study 1: vocabulary Study 1: Study 1: large learning set (a) mastery of (11) word meanings smaller learning set (sl) Study 2: reasoning Study 2: Study 2: basic correction (a) percentage (bc) correct elaborate correction (ec) Simulation Woodward, J., Carnine, D., & Collins, M. (1988) Study 1: vocabulary Study 1: Study 1: (11)[greater than](sl) not provided number of sessions to reach mastery Study 2: reasoning Study 2: Study 2: (ec)[greater than](bc) parallel reliability more = 84% effctive Table 2 CAI for LD - Matrix of Intervention Type and Focus Prereading - Alphabetic understanding for phonological recoding word recognition Strategy Thorkildsen & Friedman Instruction (1986) Drill and Helsel-Dewert & Jones, Torges, & Practice Van Den Meiracker Sexton, (1987) (1987) McGregor, Drossner, & Axelrod, (1990) Torgensen, Waters, Cohen, & Torgensen (1988) Van Daal & Van der Leij (1992) Simulation Language understanding - vocabulary Strategy Woodward, Carnine, & Instruction Collins, (1988) Study 1 Drill and Herbert, & Murdock, Practice (1994) Lin, Podell, & Rein, (1991) Mains (1985) Simulation Reading comprehension and HOTS Strategy Bhar, Kinzer, & Rieth (1991) Instruction Collins, Camine, & Gersten, (1987) Keene, & Davey (1987) Woodward, Carnine, & Collins, (1988) Study 2 Drill and Harper & Ewing (1986) Practice Simulation Woodward, Carnine, Gersten, Gleason, Johnson, & Collins (1986) Study 1 & 2
|Printer friendly Cite/link Email Feedback|
|Author:||Hall, Tracey E.; Hughes, Charles A.; Filbert, Melinda|
|Publication:||Education & Treatment of Children|
|Date:||May 1, 2000|
|Previous Article:||Transdisciplinary Consultation: A Center-based Team Functioning Model.|
|Next Article:||Measuring Behavioral Health Outcomes: A Practical Guide.|