|Australasian Journal of Educational Technology
2005, 21(2), 173-191.
Computer assisted learning (CAL) can involve a computerised intelligent learning environment, defined as an environment capable of automatically, dynamically and continuously adapting to the learning context. One aspect of this adaptive capability involves automatic adjustment of instructional procedures in response to each learner's performance, to facilitate the ease of learning and to minimise errors during learning. This process of dynamically varying the help provided to the learner by the instructor has been termed scaffolding. A bonus from using scaffolding is that the programming algorithms by which scaffolding is achieved allow integrated assessment of the learner's performance. This paper outlines the nature and origins of scaffolding concepts and illustrates their application as instructional design strategies in an experimental intelligent learning environment designed to teach basic reading skills to children with learning disabilities. The paper also illustrates the role of integrated assessment as an essential component of scaffolding and as a means of monitoring and recording the learning process.
When used by human tutors rather than computers, this process of dynamically varying the help provided to the learner by the instructor has been associated with the metaphor of 'scaffolding'. This idea is often attributed to Vygotsky, although the actual metaphor was first proposed by Wood and colleagues (Wood, Bruner & Ross, 1976) and more recently has been refined by others (Lepper, Drake & O'Donnell-Johnson, 1997). Scaffolding has sometimes been confused with the related concept of 'support', but is different in that it implies the removal and reinstatement of support according to the learner's need.
Scaffolding has become widely used in education as a metaphor for the use of additional elements that are added to a basic instructional format to facilitate learning and are later removed when the learner no longer requires them. Simple examples of scaffolding components are: a dot temporarily placed on the left of a printed sentence to cue a beginning reader where to start; hyphens used temporarily to separate syllables in polysyllabic words to assist the reader to sound out the words; and, providing progressively delayed prompts to assist a learner to supply an answer to a question. Scaffolding may be seen as beneficial for learning if it has one or more of the following effects: reduces the amount of instruction needed to reach a learning goal; reduces the number of errors made during learning; or, reduces frustration and other negative emotional responses to learning difficulties.
Although scaffolding has been widely recommended (and debated) for use by human teachers (Butler, 1998; Lepper et al., 1997; Scruggs & Mastropieri, 1998; Wong, 1998), there is little published information about scaffolding strategies for intelligent learning environments used in CAL (Denning & Smith, 1998; Dube, Iennaco, Rocco, Klederas, & McIlvane, 1992). The purpose of this paper is to briefly review behavioural research on scaffolding principles and to illustrate the application of these principles in an intelligent learning environment developed to teach basic reading skills to children with learning disabilities. The paper also demonstrates how integrated assessment of the learner's performance is an essential component of scaffolding algorithms and can be used to provide assessment information to both the learner and the teacher on an ongoing basis.
Unfortunately , educational software designers have done little as yet to apply evidence based scaffolding principles in their products, even though teachers apparently would like access to software that is s responsive to variation in learners' ability levels (Judge, 2001). Because most analyses of scaffolding and handbooks describing applications have been focused on the strategies used by human tutors (Lepper et al., 1997; Wolery, Ault, & Doyle, 1992), it would be useful to have analyses and applications that are focused on the design of digital learning environments.
Summarised below are some of the learning concepts from the behavioural research literature that are relevant to the design of scaffolding components of intelligent learning environments. Although these concepts arise from research in the behaviourist/instructivist paradigm, potentially they can play a useful part within all learning environments, including those based on a constructivist model, provided only that there are clear behavioural goals that can be set for the learner to achieve. Especially in the case where software applications are used to help children with learning difficulties learn specific skills that are not negotiable in standard school curricula, for example letter recognition, instructional approaches to learning will continue to have an important place (Kameenui, Simmons, Chard & Dickson, 1997).
First, several relevant learning concepts are briefly described to place them in theoretical context, then their application is illustrated in components of a specific educational software tool.
A familiar concept seen in digital interactive games is the availability of a tier of performance levels from which players can choose a level appropriate to their self assessment of their skill levels. Interactive adjustment in scaffolding goes beyond this concept, in that the current level of expected performance is based on the system's ongoing assessment of the 'player's' performance and is adjusted dynamically on the basis of each response the 'player' makes (Touchette & Howard, 1984). For a child learning to read a given word, a tutor's prompt (supply the word name) can be given after a delay that is increased or decreased according to a rule.
The following sections present examples of how scaffolding concepts have been integrated into an intelligent learning environment designed to teach basic reading skills to young children with learning disabilities. The software application , which has yet to be evaluated in clinical trials, represents an experiment on the integration of scaffolding strategies into basic skills educational software. The individual scaffolding strategies used in the software have been subject to prior testing with human tutors and, in some instances, also with computerised tutoring (Beale, 2000). Although the curriculum content and target population are quite specific, these examples might serve as useful models for others involved in the design of learning software. In particular, it is possible to follow the process of translating a scaffolding concept into a software design strategy. These examples are not being offered here as models of best practice; there may well be better alternatives found in the future. Rather, they are offered as an illustration of a working model based of the application of learning principles into software to create an intelligent learning environment.
This outline does not provide a full picture of the use of scaffolding concepts and strategies in the software, because it describes only one of a set of integrated activities. Some of the strategies are incorporated across activities, rather than just within activities. The individual activities are located at sites in a 3-D virtual city, which the learner (player) can navigate freely or with guidance from a virtual robot.
The activity described here is called 'Balloon Targets'. Its objective, shared partly with some other activities, is to teach the young player to identify, when they are named, the lower case letters b, d, p, & q. The main interface for this activity is shown in Figure 1. In this activity a full screen, upper level room is displayed. The player looks at a wall with a window out to a cityscape. Balloons traverse this window, left to right. Each balloon may have a letter (b, d, p, or q) displayed on it. As a balloon passes the window it can be hit by a target moved by the player, but only if the player has first clicked on same letter as displayed on the balloon, or named by the voice. As each balloon appears at the left of the window, the name of a letter is spoken is spoken. When a balloon is hit, it flashes a color (red, white or blue) corresponding to the accuracy of the shot. At first the player is guided (prompted) to choose the correct letter. This guidance is gradually removed as the player makes correct choices, but it is gradually restored if the player makes mistakes. As well as guidance, correction is provided whenever the player clicks on an incorrect letter. The parameters chosen for scaffolding strategies, including placement of visual prompts, prompt delay intervals and increments of change, are based as far as possible on prior research (Beale, 2000).
Figure 1: Balloon Targets: Main interface
About 20 percent of children continue to confuse the letters b, d, p and q in reading, spelling and writing after three or more years of schooling, and this impedes the development of more complex skills such as word analysis (Corballis & Beale, 1983). Mastery of letter identification is widely regarded as prerequisite for more advanced reading skills. Although the 'balloons' activity does not explicitly require the learner to speak the letter name when shown the letter, it has been shown that the training given in this activity typically will have this result, an aspect of a phenomenon known as stimulus equivalence (Sidman, 1971).
The following sections list the scaffolding concepts that are used in the software application, and describe how they are represented as software design strategies. These specific concepts are derived from the more general concepts listed in a previous section of this article.
Reading words and sentences, not just letter discrimination, requires an ability to scan consistently from left to right, and this in turn requires the development in the learner of a left-right somatosensory gradient (a difference in feeling between the left and right sides of the body). The software application includes two compulsory activities that facilitate development of the left-right gradient. The activity 'Controlled Intersections' teaches left-right response differentiation, and 'Road Repairs' teaches left-right and up-down stimulus discrimination. These skills supplement those taught in 'Balloon Targets', and both include their own sets of scaffolding strategies.
This activity (Figure 2) teaches left/right response differentiation. The child learns to make left/right responses to instructions that contain no visual prompts. This skill is taught at intersections controlled by traffic lights. The buggy stops automatically for the red light. A spoken instruction is given to go straight ahead, turn left, or turn right. The child must learn to respond by clicking the correct direction control (straight ahead, left, or right). At first the child is automatically guided to make the correct turn. This guidance is gradually removed as the child makes correct choices, but it is gradually restored if the child makes mistakes. As well as guidance, correction is provided whenever the child attempts to make an incorrect choice. When the child clicks the correct turn control, the instruction (e.g., 'turn left') is repeated, and the control flashes green as the computer moves the buggy through the intersection.
Figure 2: Controlled intersection
The child is taught awareness of the left/right and up/down dimensions, and to transpose differences on these dimensions onto objects in the immediate environment (Figure 3). This skill is taught in any street. The buggy stops automatically when it encounters an unpaved hole in a street. The computer displays a full screen view from above of the hole and the pavers (blocks) that must be placed to repair it. When the pavers are correctly placed, the screen returns to the 'buggy view' and the buggy can proceed to the next intersection. The activity teaches the child to distinguish between different orientations of the same basic shape. The pavers are all 'L-shaped', but in 4 different orientations, corresponding to faint paver outlines (templates) in the hole to be paved. The child must drag and drop each paver to the correct shaped hole. At first the child is guided (prompted) to make a correct drop. This guidance is gradually removed as the child makes correct choices, but it is gradually restored if the child makes mistakes.
Figure 3: Road repairs
Correct actions by the learner are immediately rewarded with positive feedback. For example, clicking on the correct letter immediately stops the balloon and produces the target. This is marked by auditory and visual feedback. Correct actions are immediately rewarded by points accumulated on an on screen display. These points may be exchanged later for tangible rewards displayed in a virtual shop. Incorrect actions are immediately corrected and the correct action is prompted.
Additional extrinsic reward is available to the learner in the form of charts showing improvement in skills over time. These may be printed out and stuck on the wall. Motivation is maintained by minimising errors during learning. It has been shown that error minimisation and maximisation of success are important for maintaining motivation for specific learning tasks and for learning generally (Lepper & Malone, 1987). The strategies used for this are described in the following section.
Progressively, the task changes from simultaneous matching to sample, through delayed matching to sample, to a simple identification task in which the learner is able to respond to hearing a letter name spoken, by choosing the correct letter from the display of four alternative letters (b, d, p, q). To further assist the learner with this task, the learner may at first be shown which is the correct 'target' letter by 'flashing' part of the correct letter. The letter stem is flashed, because the position of the letter stem (left, right, up, down) is what distinguishes the four letters from one another. This prompting procedure, called highlighting a distinctive feature, helps the learner make a correct selection (Schreibman, 1975). Later in learning, the highlighting is progressively delayed as described previously, to assist the learner to choose correctly without the aid of the highlighting.
Highlighting, as used here, provides an example of scaffolding using an intra-stimulus prompt. These prompts have been shown to be the most efficient method of drawing the learner's attention to the important features of the task (Schreibman, 1975).
Problems with sustaining attention to task may be minimised by maximising motivation to participate and learn. This is addressed by the interactive features of the activity, such as immediate feedback on actions, including positive feedback for correct actions, immediate correction of errors, repeating of instructions where required, and demonstration (modeling) of required actions. Instructions and modeling are scaffolding features because they are provided when required at the beginning of training, but are subsequently available only on request by the learner.
Problems with selective attention are addressed by requiring the learner to manipulate the letters and other important objects to be learned (Beale, 2000). In the 'Balloon Targets' activity, this strategy is used to direct the learner's attention to the letters to be learned and their distinctive features.
In the current example, an attempt has been made to minimise the impact of memory problems by using tasks with a very low initial working memory requirement and then gradually increasing this as learning progresses. In the 'Balloon Targets' activity, the initial requirement to remember the sample letter is almost zero, since the visual sample is displayed immediately and remains available until the correct action has been performed. The sample and the target array of letters are separated only by a small visual angle that allows simultaneous scanning. When, during progressive delayed prompting, a delay is introduced between the letters and then increased, the process is gradual and should therefore support a gradual strengthening of the relevant process in working memory.
Figure 4: Adult level graph
The graphical record above shows the child's learning of two skills taught in the activity, letter matching (green) and letter naming (orange). Letter matching is the easier skill, choosing the letter beneath the window that matches the letter displayed on the balloon. Letter naming is more difficult, choosing the letter beneath the window that corresponds to the letter spoken, before it is displayed on the balloon.
The strategy used in this activity is to first teach letter matching, then progressively introduce letter naming. As the child learns letter naming, opportunities for letter matching will gradually disappear and will only reappear if mistakes are made on attempts at letter naming.
In the example shown above, the child has used the activity ten times. On the first session, 10% of letter matching attempts were successful (green, 10%) and letter naming was either not tried or not successful (orange, 0%). By the last (10th) session, the child was able to correctly name the letter every time (100%) and there were no opportunities for letter matching (0%). Over the ten sessions on this activity, the child has first learned to match the letters (identify similarities and differences in shape), then has learned to associate each of the four letters with its spoken name.
Figure 5: Child level graph
The child graph allows a young child to see more easily how performance has changed from the first use of the activity to now. The taller the balloon, the better the performance. The balloon and the icon at the top right corner will remind the child which activity the graph refers to.
All that is required to include scaffolding and integrated assessment into educational software is close collaboration between a learning expert with a good understanding of scaffolding and a software design team able to translate the scaffolding requirements into learning algorithms. Setting the parameters for algorithms, such as initial prompt delays and delay increments, is inevitably a 'seat of the pants' operation requiring personal experience in tutoring as well as familiarity with the relevant empirical literature on scaffolding design. Because the best solution is likely to depend on the content being taught and the specifics of the intelligent environment being used, it is advisable to design the software so that parameters can be varied during a testing program. These parameters should not be locked down before the software has been thoroughly tested and adjusted on the basis of responses from representative learners.
Beale, I. L. (1998). Learning disabilities. In A. Bellack & M. Hersen (Eds), Comprehensive Clinical Psychology (Vol. 9, pp. 37-55). London: Pergamon.
Beale, I. L. (2000). Integrated neural learning. Arrowtown, NZ: Brainwaves Limited.
Becker, W. C., Engelmann, S. & Thomas, D. R. (1975). Teaching 2: Cognitive learning and instruction. Chicago: Science Research Associates.
Butler, D. L. (1998). In search of the architect of learning: A commentary on scaffolding as a metaphor for instructional interactions. Journal of Learning Disabilities, 31(4), 374-385.
Corballis, M. C. & Beale, I. L. (1983). The ambivalent mind: The neuropsychology of laterality. Chicago: Nelson Hall.
Denning, R. & Smith, P. J. (1998). A case study in the development of an interactive learning environment to teach problem-solving skills. Journal of Interactive Learning Research, 9(1), 3-36.
Dube, W. V., Iennaco, F. M., Rocco, F. J., Klederas, J. B. & McIlvane, W. J. (1992). Microcomputer-based programmed instruction in identity matching to sample for persons with severe disabilities. Journal of Behavioral Education, 2(1), 29-51.
Etzel, B. C. & LeBlanc, J. (1979). The simplest treatment alternative: appropriate instructional control and errorless learning procedures for the difficult-to-teach child. Journal of Autism and Developmental Disorders, 9, 361-382.
Hitchcock, C. (2001). Balanced instructional support and challenge in universally designed learning environments. Journal of Special Education Technology, 16(4). [verified 26 Apr 2005] http://jset.unlv.edu/16.4/hitchcock/first.html
Judge, S. L. (2001). Computer applications in programs for young children with disabilities: Current status and future directions. Journal of Special Education Technology, 16(1). [verified 26 Apr 2005] http://jset.unlv.edu/16.1/Judge/first.html
Kalyuga, S. (2000). When using sound with a text or picture is not beneficial for learning. Australian Journal of Educational Technology, 16(2), 161-172. http://www.ascilite.org.au/ajet/ajet16/kalyuga.html
Kameenui, E. J., Simmons, D. C., Chard, D. & Dickson, S. (1997). Direct-instruction reading. In S. A. Stahl & D. Hayes (Eds), Instructional models in reading. (pp. 59-84). Mahwah, NJ: Erlbaum.
Lepper, M. R., Drake, M. F. & O'Donnell-Johnson, T. (1997). Scaffolding techniques of expert human tutors. In K. Hogan & M. Pressley (Eds), Scaffolding student learning: Instructional approaches and issues (pp. 108-144). Cambridge, MA: Brookline Books.
Lepper, M. R. & Malone, T. W. (1987). Intrinsic motivation and instructional effectiveness in computer-based education. In R. E. Snow & M. J. Farr (Eds), Aptitude, learning, and instruction III. Conative and affective process analysis. Hillsdale, NJ: Erlbaum.
Moreno, R. & Mayer, R. E. (2000). A learner-centered approach to multimedia explanations: Deriving instructional design principles from cognitive theory. Interactive Multimedia Electronic Journal, 2(2). http://imej.wfu.edu/articles/2000/2/05/index.asp
Mosk, M. D. & Bucher, B. (1984). Prompting and stimulus shaping procedures for teaching visual-motor skills to retarded children. Journal of Applied Behavior Analysis, 17, 23-34.
Nelson, R. O. & Hayes, S. C. (1981). Nature of behavioral assessment. In M. Hersen & A. S. Bellack (Eds), Behavioral assessment: A practical handbook (2 ed., pp. 3-37). New York: Pergamon.
Peterson, N. (1982). Feedback is not a new principle of behavior. The Behavior Analyst, 5, 101-102.
Reid, D. K. (1998). Scaffolding: A broader view. Journal of Learning Disabilities, 31(4), 386-396.
Sampson, D., Karagiannidis, C. & Kinshuk (2002). Personalised learning: Educational, technological and standardisation perspective. Interactive Educational Multimedia, 4, 24-39. http://www.ub.es/multimedia/iem/down/c4/Personalised_Learning.pdf
Schreibman, L. (1975). Effects of within-stimulus and extra-stimulus prompting on discrimination learning in autistic children. Journal of Applied Behavior Analysis, 8, 91-112.
Schwartz, B. & Reisberg, D. (1991). Learning and memory. New York: W. W. Norton.
Scruggs, T. E. & Mastropieri, M. A. (1998). What happens during instruction: Is any metaphor necessary? Journal of Learning Disabilities, 31(4), 404-408.
Sidman, M. (1971). Reading and auditory-visual equivalences. Journal of Speech and Hearing Research, 14, 5-13.
Solman, R. T., Singh, N. N. & Kehoe, E. J. (1992). Pictures block the learning of sightwords. Educational Psychology, 12, 143-153.
Stokes, T. F. & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10, 349-367.
Touchette, P. E. & Howard, J. S. (1984). Errorless learning: Reinforcement contingencies and stimulus control transfer in delayed prompting. Journal of Applied Behavior Analysis, 17, 175-188.
Wolery, M., Ault, M. J. & Doyle, P. M. (1992). Teaching students with moderate to severe disabilities. White Plains, New York: Longman.
Wong, B. Y. (1998). Analyses of intrinsic and extrinsic problems in the use of the scaffolding metaphor in learning disabilities intervention research: An introduction. Journal of Learning Disabilities, 31(4), 340-343.
Wood, D. J., Bruner, J. S. & Ross, G. (1976). The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17, 89-100.
|Author: Ivan L. Beale is Visiting Associate Professor, School of Psychology, University of New South Wales, Sydney. Tel: +61 040 191 8395 Email: email@example.com, firstname.lastname@example.org
Please cite as: Beale, I. L. (2005). Scaffolding and integrated assessment in computer assisted learning (CAL) for children with learning disabilities. Australasian Journal of Educational Technology, 21(2), 173-191. http://www.ascilite.org.au/ajet/ajet21/beale.html