Computer Assisted Instruction
This article presents an overview of Computer-Assisted Instruction (CAI), with an emphasis on CAI theory and practice in public schools. CAI involves the use of computers to supplement or assist classroom instruction. Even though their capabilities were limited, the first public school CAI systems, operated on teletype machines to drill elementary school students in arithmetic, were shown effective in promoting skill development. As computer technologies advanced, the types of assistance that CAI systems could provide grew. CAI systems provide instructional assistance across all pre-kindergarten through college levels and across virtually all subject areas. One crucial issue in CAI theory and practice is comparative evaluation: how effective is CAI in comparison to other approaches it replaces during the class day? Various cumulative studies suggest neutral or mild to very significant positive benefits to CAI. Computer-Assisted Instruction's original theoretical roots can be traced to variations of Instructional System Design (ISD). Since the early 1990s, because of advances in technology and in theories of learning and cognition, CAI has taken more varied forms, and the kinds of learning that it has been used to promote have expanded significantly.
Keywords Computer-Assisted Instruction (CAI); Cognitive Science; Constructivism; Educational Psychology; Instructional Systems Design; Learning Models; Mastery Learning; Objectives
Technology in Education: Computer-Assisted Instruction
Computer-Assisted Instruction (CAI) involves the use of computers to supplement or assist classroom instruction. In its traditional form, CAI relies on software that presents information and guides a learner through a series of subject matter objectives, quizzing the student periodically and assessing progress to a mastery level. CAI does not supplant or fully replace the teacher in a classroom environment. The term CAI overlaps with the related term of Computer-Based Instruction (CBI). Though these terms have sometimes been used synonymously, CBI also refers to instruction that is organized more fully around the computer system as the primary source of instructional delivery. A systems that combine either full or partial curriculum but also the related testing and records management for large groups of students (for example, a school) is referred to as an Integrated Learning System, or ILS. Some of the most prominent ILS's include the Plato System, Jostens Learning, and Pearson Education.
A primary advantage that CAI offers to the instructional process is that it permits students to proceed through curriculum objectives at a pace that they find comfortable. Classroom settings that move in lock-step at a pace determined by a teacher can either bore students by moving too slowly for them, or lose students by moving on to new material before mastery has been acquired on that which has already been covered. CAI software, however, permits students to dwell on material or return to it until they have reached a mastery level. It allows the student to do so in a more private context than a classroom, where students who learn more slowly may be subjected to embarrassment or ridicule from classmates. It also permits students who learn more rapidly and who have mastered a set of objectives to progress without waiting to new skills and concepts that they have not yet acquired. Thus, CAI permits customization that is usually not possible to attain for all students in a classroom. This individualization is crucial to understanding why CAI is considered a powerful educational strategy.
CAI has several other advantages. In general, as a supplement rather than a replacement for teacher-based instruction, it provides a blended approach to learning and thus a more diverse educational experience for students. Virtually all large scale studies that examine student attitudes find that CAI can motivate students and improve attitudes towards learning. Some studies have shown CAI improves school attendance. CAI can relieve a teacher of routine tasks that are associated with student practice, exercise, and drill, and do so at significantly lower cost than the teacher. Finally, one of the most important advantages involves what educational psychologists refer to as "locus of control," which is when students have a greater sense of control over their learning experience and how it is paced; they have less ambiguity about their performance because CAI systems furnish frequent feedback.
Early Frameworks of CAI
CAI has roots in several conceptual frameworks that were developed primarily in the 1950s through the 1980s. Three of the most important of these are Instructional System Design (ISD), pioneered most prominently by Glaser (1962), Mager (1975), and Dick and Carey (1990); hierarchical learning objectives and conditions for learning advanced by Gagne and colleagues at Florida State University (1977; 1992), and Mastery Learning, pioneered by Benjamin Bloom of the University of Chicago (Bloom, 1981). These frameworks collectively emphasize training and instruction that follow carefully structured sequences of subject material or tasks. This requires the instructional designer to break down the subject matter into a hierarchy of objectives. The student masters each objective in the sequence before moving on to the next objective. Frequent assessment of progress in each step towards mastery allows an evaluation both of the achievements of the student and the effectiveness of the instructional system. It also permits continuous feedback to the learner and positive reinforcement for successful completion of instructional tasks and for mastery of objectives.
This type of methodical, objective-based, and hierarchical approach to learning does not inherently require the use of any computer assistance at all, but it proved to be a good fit for early efforts to build software systems that could assist instruction.
reinforcement for important knowledge and skills. Computers were actually teletype terminals connected to mainframe computers. The first published evaluations of CAI's effectiveness, in 1969, by Stanford researchers Patrick Suppes and Mona Morningstar, provided evidence of learning gains by grade school students in mathematics and in college level Russian (1969a; 1969b). Usage of CAI remained limited, however, through the 1970s in part because of expense and the physical difficulty of providing access to students to appropriate equipment. These and other factors limited the kinds of CAI software that was developed, which in turn limited the appeal of CAI. By the end of the 1970s and early 1980s, though, some studies had taken place to measure the effectiveness of CAI in areas such as physics instruction (Tennyson, 1981), in working with learning-disabled students (Isenberg, 1985; Schmidt, Weinstein, Niemic, & Walberg, 1985), and in comparing the effects of CAI in low socioeconomic status (SES) versus more affluent environments (Mevarech & Rich, 1985). These studies with early systems produced mixed results. The most promising results involved the finding that CAI was effective in structuring learning experiences for students from disadvantaged and low-SES environments.
Evaluations of Effectiveness
One of the most crucial issues in education evaluation has been whether CAI is a cost-effective component of classroom learning. Broadly speaking, three large issues are at stake for public school policy-makers in evaluating the use of CAI. These include the cost of hardware and software, the cost in training teachers, and the time that students devote to CAI that they would otherwise spend in other forms of instruction. The potential advantages of CAI fall into two principle categories: do students attain or exceed the learning objectives that would be available to them from other forms of instruction, and do students find more satisfaction in using CAI than in alternatives? Some researchers, such as Larry Cuban of Stanford University, have frequently challenged the use of computers and CAI as overly faddish (e.g., Cuban, 1993). These critics also complain that CAI can divert resources from more important learning experiences that schools should provide.
Metastudies of their effectiveness were undertaken. Some of the most important were by Kulik and colleagues, who repeatedly (1991; 1983; 1987; 1985) found statistically significant improvement trends in learning achievement for students in CAI. One later metastudy concluded, though, that the actual benefit attributable to CAI may more plausibly be attributed to the fact that CAI content may be higher quality materials than traditional alternatives (Fletcher-Flinn & Gravatt, 1995). Some metastudies focused on CAI with particular groups of students or subject areas. One early study suggested that learning disabled students did not benefit from CAI, but subsequent studies published since 2000 all point to significant gains for learning disabled students who participated in CAI in areas such as mathematics, reading comprehension, vocabulary, and spelling (Fuchs et al., 2006; Hall, Hughes, & Filbert, 2000; Jitendra, Edwards, Sacks, & Jacobson, 2004).
Metastudies continue to suggest that students from disadvantaged environments especially benefit from CAI. That is, CAI may contribute a structure that is especially effective with schools in different types of disadvantaged or low-SES settings (Mevarech & Rich, 1985; Swan & Guerrero, 1990). None of the prominent metastudies undertaken since 1980 have indicated that CAI has a significant negative impact on learning achievement, though several suggest that the impact was vaguely specified, minimal, or else attributable to other factors (Cousins & Ross, 1993). Additionally, some have suggested that CAI has a hidden "de-skilling" effect by focusing attention on basic skills that should be spent on more complex problem solving (Martin, 1999).
Several metastudies suggest statistically significant improvements attributable to the use of CAI (Cotton, 1991). Additionally, students often found CAI materials to be more intrinsically engaging and interesting than other instructional approaches (Anand & Ross, 1987; Cotton, 1991; Del Marie Rysavy & Sales, 1991). The US Department of Education supports research laboratories around the country to provide a...
(The entire section is 4554 words.)