2.3 The Behavioral Roots of Instructional Technology

2.3.1 Methodological Behaviorism

Stimulus-response behaviorism, that is, behaviorism which emphasizes the antecedent as the cause of the behavior is generally referred to as methodological behaviorism (see e.g., Day, 1983, Skinner, 1974). As such it is in line with much of experimental psychology; antecedents are the independent variables and the behaviors are the dependent variables. This transformational paradigm (Vargas, 1993) differs dramatically from the radical behaviorism of Skinner (e.g., 1945, 1974) which emphasizes the role of reinforcement of behaviors in the presence of certain antecedents(see 7.3.3), in other words, the selectionist position. Most of the earlier work in instructional technology followed the methodological behaviorist tradition. In fact, from a radical behaviorist position cognitive psychology is an extension of methodological behaviorism (Skinner, 1974). Although we have recast and reinterpreted where possible, the differences, particularly in the Film and TV research(see 11.5 to 11.7)are apparent. Nevertheless, the research is part of the research record in instructional technology and has therefore necessary, and moreover, useful from an S-R perspective.

One of the distinctive aspects of the methodological behavioral approach is the demand for "experimental" data (manipulation)(see 11.2.4.1)to justify any interpretation of behavior as causal. Natural observation, personal experience and judgment(see 40.2)fall short of the rules of evidence to support any psychological explanation (Kendler, 1971). This formula means that an learner must make the "correct response when the appropriate stimulus occurs" and when the necessary conditions are present (Kendler, 1971).

Usually there is no great problem in providing the appropriate stimulus, for audiovisual techniques have tremendous advantages over other educational procedures in their ability to present to the learner the stimuli in the most effective manner possible (p. 36).

In addition, Gropper (1963) suggests that visual materials should help students acquire, retain, and transfer responses because visuals have the capacity to cue and reinforce specified responses.

A problem arises as to when to develop techniques (in which appropriate responses to specific stimuli can be practiced and reinforced). The developer of an instructional medium must know exactly what response is desired from the students, otherwise it is impossible to design and evaluate instruction. Once the response is specified, the problem becomes getting the student to make this appropriate response. This response must be practiced and the learner must be reinforced to make the correct response to this stimulus (Skinner, 1953b). Under the S-R paradigm much of the research on the instructional media was based upon the medium itself (i.e., the specific technology). The medium became the independent variable and media comparison studies became the norm until the middle 1970's (Smith & Smith, 1966). In terms of the methodological behavior model, much of the media (programmed instruction, film, television, etc.) functioned primarily within the stimulus component. From this position, Carpenter (1962) reasoned that any medium (e.g., film, television) "imprints" some of its own characteristics on the message itself. Therefore, the content and medium have more impact than the medium itself. The "way" the stimulus material (again film, television, etc.) interacts with the learner instigates motivated responses. Carpenter (1962) developed several hypotheses based upon his interpretations of the research on media and learning and include the following possibilities:

1. The most effective learning will take place when there is similarity between the stimulus material (presented via a medium) and the criterion or learned performance.
2. Repetition of stimulus materials and the learning response is a major condition for most kinds of learning.
3. Stimulus materials which are accurate, correct, and subject to validation can increase the opportunity for learning to take place.
4. An important condition is the relationship between a behavior and its consequences. Learning will occur when the behavior is "reinforced" (Skinner, 1968). This reinforcement by definition, should be immediately after the response.
5. Carefully sequenced combinations of knowledge and skills presented in logical and limited steps will be the most effective for most types of learning.
6. "...established principles of learning derived from studies where the learning situation involved from direct instruction by teachers are equally applicable in the use of instructional materials" (Carpenter, 1962, p. 305).

Practical aspects of these theoretical suggestions go back to the mid-1920's with the development by Pressey of a self-scoring testing device. Pressey (1926, 1932) discussed the extension of this testing device into a self-instruction machine. Versions of these devices later (after World War II) evolved into several, reasonably sophisticated, teaching machines for the U.S. Air Force which were variations of an automatic self-checking technique. They included a punched card, a chemically treated card, a punch board, and the Drum Tutor. The Drum Tutor used informational material with multiple-choice questions, but could not advance to the next question until the correct answer was chosen. All devices essentially allowed students to get immediate information concerning accuracy of response.

2.3.2 Teaching Machines

Peterson (1931) conducted early research on Pressey self-scoring testing devices. His experimental groups were given the chemically treated scoring cards used for self-checking while studying a reading assignment. The control group had no knowledge of their results. Peterson found the experimental groups had significantly higher scores than the group without knowledge of results. Little (1934), also using Pressey's automatic scoring device, had the experimental group as a test-machine group, the second group using his testing teaching machine as a drill-machine and the third group as a control group in a paired controlled experiment. Both experimental groups scored significant higher mean scores than the control group. The drill-machine group scored higher than the test-machine group. After World War II additional experiments using Pressey's devices were conducted. Angell and Troyer (1948) and Jones and Sawyer (1949) found that giving immediate feedback significantly enhanced learning in both citizenship and chemistry courses. Briggs (1947) and Jensen (1949) found that self-instruction by superior students using Pressey's punch boards enabled them to accelerate their course work. Pressey (1950) also reported efficacy of immediate feedback in English, Russian vocabulary, and psychology courses. Students given feedback via the punch boards received higher scores than those students who were not given immediate feedback. Stephens (1960), using Pressey's Drum Tudor, found students using the device scored better than students who did not. This was true even though the students using the Drum Tudor lacked overall academic ability. Stephens "confirmed Pressey's findings that errors were eliminated more rapidly with meaningful material and found that students learned more efficiently when they could correct errors immediately" (Smith & Smith, 1966, p. 249). Severin (1960) compared the scores of students given the correct answers with no overt responses in a practice test with those of students using the punch board practice test and found no significant differences. Apparently pointing out correct answers was enough and an overt response was not required. Pressey (1950) concluded that the use of his punch board created a single method of testing, scoring, informing students of their errors, and finding the correct solution all in one step (called telescoping). This telescoping procedure, in fact, allowed test-taking become a form of systematically directed self-instruction. His investigations indicated that when self-instructional tests were used a the college level, gains were substantial and helped improve understanding. However, Pressey (1960) indicated his devices may not have been sufficient to stand by themselves, but were useful adjuncts to other teaching techniques.

Additional studies on similar self-instruction devices were conducted for military training research. Many of these studies used the automatic knowledge of accuracy devices such as The Tab Item and the Trainer-Tester (Smith & Smith, 1966). Glaser, Damrin, and Gardner (1954) and Cantor and Brown (1956) all found that scores for a troubleshooting task were higher for individuals using these devices than those using a mock-up for training. Dowell (1955) confirmed this, but also found that even higher scores were obtained when learners used the Trainer-Tester and the actual equipment. Briggs (1958) further developed a device called the Subject-Matter trainer which could be programmed into five teaching and testing modes. Irion and Briggs (1957) and Briggs (1958) found that prompting a student to give the correct response was more effective than just confirming correct responses.

Smith and Smith (1966) point out that while Pressey's devices were being developed and researched, they actually only attracted attention in somewhat limited circles. Popularity and attention were not generated until (Skinner (1953a, b; 1954) used these types of machines. "The fact that teaching machines were developed in more than one content would not be particularly significant were it not true that the two sources represent different approaches to educational design..." (Smith & Smith, 1966, p. 245). Skinner developed his machines to test and develop his operant conditioning principles developed from animal research. Skinner's ideas attracted attention, and as a result, the teaching-machine and programmed instruction movement become a primary research emphasis during the 1960's. In fact, from 1960-1970 research on teaching-machines and programming was the dominant type of media research in terms of numbers in the prestigious journal, AV Communication Review (AVCR), (Torkelson, 1977). From 1960-1969 AVCR had a special section dedicated to teaching machines and programming concepts. Despite the fact of favorable research results from Pressey and his associates and the work done by the military, the technique was not popularized until Skinner (1954) recast self-instruction and self-testing. Skinner (1954) believed that any response could be reinforced. A desirable behavior could be taught by reinforcing a response which was close in time to this behavior. By reinforcing "successive" approximations, behavior will eventual approximate the desired pattern (Homme, 1957). Obviously, this reinforcement needed a great deal of supervision. Skinner believed that, in schools, reinforcement may happen hours, days, etc. after the initial event and the effects would be greatly reduced. In addition, he felt that it was difficult to individually reinforce a response of an individual student in a large group. He also believed that school used negative reinforcers -- to punish, not necessarily as reinforcement (Skinner, 1954). To solve these problems, Skinner also turned to the teaching-machine concept. Skinner's (1958) machines in many respects were similar to Pressey's earlier teaching testing devices. Both employed immediate knowledge of results immediately after the response. The students were kept active by their participation and both types of devices could be used in a self-instruction manner with students moving at their own rate. Differences in the types of responses in Pressey's and Skinner's machines should be noted. Skinner required students to "overtly" compose responses (e.g., writing words, terms, etc.). Pressey presented potential answers in a multiple-choice format, requiring students to "select" the correct answer. In addition, Skinner (1958) believed that answers could not be easy, but that steps would need to be small in order that there would be no chance for "wrong" responses. Skinner was uncomfortable with multiple-choice responses found in Pressey's devices because of the chance for mistakes (Homme, 1957; Porter, 1957; Skinner & Holland, 1960).

2.3.3 Films

The role and important of military research(see 1.10, 11.2.3) during World War II and immediately afterward cannot be underestimated either in terms of amount or results. Research studies on learning, training materials, and instruments took on a vital aspect when it became necessary to train millions of individuals in skills useful for military purposes. People had to be selected and trained for complex and complicated machine systems (i.e., radio detection, submarine control, communication, etc.). As a result, most of the focus of the research by the military during and after the war was on the devices for training, assessment, and troubleshooting complex equipment and instruments. Much of the film research noted earlier stressed the stimulus, response, and reinforcement characteristics of the audiovisual device. "These [research studies] bear particularly on questions on the role of active response, size of demonstration and practice steps in procedural learning, and the use of prompts or response cues" (Lumsdaine & Glaser, 1960, p. 257). The major research programs during World War II were conducted on the use of films by the U.S. Army. These studies were conducted to study achievement of specific learning outcomes and the feasibility of utilizing film for psychological testings (Gibson, 1947; Hoban, 1946). After World War II, two major film research projects were sponsored by the United States Army and Navy at the Pennsylvania State University from 1947 to 1955 (Carpenter & Greenhill, 1955, 1958). A companion program on film research was sponsored by the United States Air Force from 1950-1957. The project at the Pennsylvania State University -- the Instructional Film Research Program under the direction of C.R. Carpenter -- was probably the "most extensive single program of experimentation dealing with instructional films ever conducted." (Saettler, 1968, p. 332). In 1954, this film research project was reorganized to include instructional films and instructional television because of the similarities of the two media. The Air Force Film Research Program (1950-1957) was conducted under the leadership of A.A. Lumsdaine (1961). The Air Force study involved the manipulation of techniques for "eliciting and guiding overt responses during a course of instruction" (Saettler, 1968, p. 335). Both the Army and Air Force studies developed research which had major implications on the use and design of audiovisual materials (e.g., film). Although these studies developed a large body of knowledge, little use of the results was actually implemented in the production of instructional developed by the military. Kanner (1960) suggested that the reason for the lack of use of the results of these studies was because they created resentment among film makers, and much of the research was completed in isolation.

Much of the research on television was generated after 1950 and was conducted by the military because television's potential for mass instruction. Some of the research replicated or tested concepts (variables) used in the earlier film research, but the bulk of the research compared television instruction to "conventional" instruction and most results showed no significant differences between the two forms. Most of the studies were applied rather than using a theoretical framework; i.e., behavior principles (Kumata, 1961).

However, Gropper (1965a,b), Gropper and Lumsdaine (1961), and others used the television medium to test behavioral principles developed from the studies on programmed instruction. Klaus (1965) states that programming techniques tended to be either stimulus centered on response centered. Stimulus centered techniques stressed meaning, structure, and organization of stimulus materials, while response-centered techniques dealt with the design of materials which insure adequate response practice. For example, Gropper (1965, 1966) adopted and extended concepts developed in programmed instruction (particularly the response centered model) to televised presentations. These studies dealt primarily with "techniques for bringing specific responses under the control of specific visual stimuli and...the use of visual stimuli processing such control within the framework of an instructional design" (1966, p. 41). Gropper, Lumsdaine, and Shipman (1961), Gropper and Lumsdaine (1961), and Gropper (1961a, b, c, d) all reported the value of pretesting and revising televised instruction and requiring students to make active responses. Gropper (1967) suggested that in television presentations it is desirable to identify which behavioral principles and techniques that underlie programmed instruction are appropriate to television. Gropper and Lumsdaine (1961) reported that merely requiring students to actively respond to nonprogrammed stimulus materials (i.e., segments which are not well-delineated or sequenced in systematic ways) did not lead to more effective learning (an early attempt at formative evaluation). However, Gropper (1967) reported that the success of using programmed instructional techniques with television depends upon the effective design of the stimulus materials as well as the design of the appropriate response practice.

Gropper (1963, 1965a, 1966, 1967) emphasized the importance of using visual materials to help students acquire, retain, and transfer responses based on the ability of such materials to cue and reinforce specified responses, and to serve as examples. He further suggests that students should make explicit (active) responses to visual materials (i.e., television) for effective learning. Later, Gropper (1968) concluded that, in programmed televised materials, actual practice is superior to recognition practice in most cases and that the longer the delay in measuring retention, the more the active response was beneficial. The behavioral features which were original with programmed instruction and later used with television and film were attempts to minimize and later correct the defects in the effectiveness of instruction on the basis of what was known about the learning process (Klaus, 1966). The use of student responses were used in many studies (e.g., Gropper, 1963; 1966) as the basis for revisions of instructional design and content. Indepth reviews of the audiovisual research carried on by the military and civilian researchers are contained in the classic summaries of this primarily behaviorist approach of Hoban and Van Ormer (1950), May and Lumsdaine (1958), Cook (1960), Hoben (1960), Carpenter & Greenhill (1955, 1958), Schramm (1962), and Chu and Schramm (1968).

The following is a sample of some of the research results on the behavioral tenants of stimulus, response and reinforcement gleaned from the World War II research and soon after based upon the study of audiovisual devices (particularly film).

2.3.3.1. Research on Stimuli. Attempts to improve learning by manipulating the stimulus condition(see 11.3)can be divided into several categories. One category, that of the use of introductory materials to introduce content in film or audiovisual research has shown mixed results (Cook, 1960). Film studies by Wittich and Folkes (1946), Weiss and Fine (1955), Wulff, Sheffield, and Kraeling (1954) reported that introductory materials presented prior to the showing of a film increased learning. But, Lathrop (1949), Norford (1949), Peterman and Bouscaren (1954) and Jaspen (1948) found inconclusive or negative results by using introductory materials. Another category of stimuli, those that direct attention, uses the behavioral principle that learning is assisted by the association of the responses to stimuli (Cook, 1960). Film studies by Kimble and Wulff (1953), Lumsdaine and Sulzer (1951), Ryan and Hochberg (1954), Gibson (1947), Roshal (1949), and McGuire (1953a) found that a version of the film which incorporated cues to guide the audience into making the correct responses produced increased learning. As might be expected, extraneous stimuli not focusing on relevant cues were not effective (Jaspen, 1950; Neu, 1950; Weiss, 1954). However, Miller and Levine (1952) and Miller, Levine and Steinberger (1952a) reported the use of subtitles to associate content to be ineffective. Cook (1960) reported that many studies were conducted on the use of color where it would provide an essential cue to understanding with mixed results and concluded it was impossible to say color facilitated learning results (i.e., Long, 1946; May & Lumsdaine, 1958). Note, the use of color in instruction is still a highly debated research issue.

2.3.3.2. Research on Response. Cook (1960), stated the general belief that, unless the learner makes some form of response that is relevant to the learning task, no learning will occur. Responses (practice) in audiovisual presentations may range from overt oral, written or motor responses to an implicit response (not overt). Cook (1960), in an extensive review of practice in audiovisual presentations, reported the effectiveness of students calling out answers to questions in an audiovisual presentation to be effective (i.e., Kanner & Sulzer, 1955; Kendler, Cook, & Kendler, 1953; Kendler, Kendler, & Cook, 1954; McGuire, 1954). Most studies which utilized overt written responses with training film and television were also found to be effective (i.e., Michael, 1951; Michael & Maccoby, 1954; Yale Motion Picture Research Project, 1947).

A variety of film studies on implicit practice found this type of practice to be effective (some as effective as overt practice) (i.e., Kanner & Sulzer, 1955; Kendler, Kendler, & Cook, 1954; Michael, 1951; McGuire, 1954; Miller & Klier, 1953a, b). Cook (1960, p. 98) notes that the above noted studies all reported that the effect of actual practice is "specific to the items practiced" and there appeared to be no carryover to other items. The role of feedback in film studies was positively supported (Gibson, 1947; Michael, 1951; Michael & Maccoby, 1954).

The use of practice, given the above results, appears to be an effective component of using audiovisual (film and television) materials. A series of studies were conducted to determine the amount of practice needed. Cook (1960) concludes that students will profit from a larger number of repetitions (practice). Film studies which used a larger number of examples or required the viewing the film more than once found students fairing greater than those with fewer examples or viewing opportunities (Brenner, Walter & Kurtz, 1949; Kendler, Cook, & Kendler, 1953; Kimble & Wulff, 1954; Sulzer & Lumsdaine, 1952). A number of studies were conducted which tested when practice should occur. Was it better to practice concepts as a whole (massed) at the end of a film presentation or practice it immediately after it was demonstrated (distributed) during the film? Most studies reported results that there was no difference in the time spacing of practice (e.g., McGuire, 1953b; Miller & Klier, 1953a, b, 1954; Miller & Levine, & Steinberger, 1952a or b). Miller and Levine (1952), however, found results in favor of a massed practice at the end of the treatment period.

2.3.4 Programmed Instruction

Closely akin, and developed, from Skinner's (1954) technology of teaching machine concepts were the teaching texts or programmed books(see 18.4.1). These programmed books essentially had the same characteristics as the machines,; logical presentations of content, requirement of overt responses and presentation of immediate knowledge of correctness (a correct answer would equal positive reinforcement (Porter, 1958, Smith & Smith, 1966). These programmed books were immediately popular for obvious reasons, they were easier to produce, portable, and did not require a complex, burdensome, costly device, i.e., a machine. As noted earlier during the decade of the 60's, research on programmed instruction, as the use of these types of books and machines, became known, was immense (Campeau, 1974). Literally thousands of research studies were conducted. (See Campeau, 1974; Glaser, 1965; Lumsdaine & Glaser, 1960; Smith & Smith, 1966) among others for extensive summaries of research in this area.) The term programming is taken here to mean what Skinner called "the construction of carefully arranged sequences of contingencies leading to the terminal performances which are the object of education" (Skinner, 1953b, p. 169).

2.3.4.1. Linear Programming. Linear programming involves a series of learning frames presented in a set sequence. As in most of the educational research of the time, research on linear programmed instruction dealt with devices and/or machines and not on process nor the learner. Most of the studies, therefore generally compared programmed instruction to "conventional" or "traditional" instructional methods, (see e.g., teaching Machines and Programmed Instruction Department in AV Communication Review, 1962-1969). These types of studies were, of course, difficult to generalize from and often resulted in conflicting results (Holland, 1965). "The restrictions on interpretation of such a comparison arises from the lack of specificity of the instruction with which the instrument in questions is paired" (Lumsdaine, 1962a or b, p. 251). Like other research of the time, many of the comparative studies had problems in design, poor criterion measures, scores prone to a ceiling effect, and ineffective and poor experimental procedures (Holland, 1965). Holland (1961), Lumsdaine (1965), and Rothkopf (1962) all suggested other ways of evaluating the success of programmed instruction. Glaser (1962) indicated that most programmed instruction was difficult to construct, time consuming, and had few rules or procedures. Many comparative studies and reviews of comparative studies found no significance in the results of programmed instruction (e.g., Alexander, 1970; Barnes, 1970; Giese & Stockdale, 1966; Frase, 1970; McKeachie, 1967; Unwin, 1966; and Wilds & Zachert, 1966). However, Marsh and Pierce-Jones (1968), Hamilton and Heinkel (1967), and Daniel and Murdoch (1968) all reported positive and statistically significant findings in favor of programmed instruction. The examples noted above were based upon gross comparisons. A large segment of the research on programmed instruction was devoted to "isolating or manipulating program or learner characteristics" (Campeau, 1974, p. 17). Specific areas of research on these characteristics included studies on repetition and dropout e.g., Rothkopf (1960), Skinner & Holland (1960). Skinner and Holland (1960) suggested that various kinds of cueing techniques could be employed which would reduce the possibility of error but generally will cause the presentation to become linear in nature (Skinner, 1961, Smith, 1959). Karis, Kent, and Gilbert (1970) found that overt responding such writing a name in a (linear) programmed sequence was significantly better than for subjects who learned under covert response conditions. However, Valverde and Morgan (1970) concluded that eliminating redundancy in linear programs significantly increased achievement. Carr (1959) stated that merely confirming the correctness of a student's response as in a linear program is not enough. The learner must otherwise be motivated to perform (Smith & Smith, 1966). However, Coulson and Silberman (1960) and Evans, Glaser, and Homme (1962) found significant differences in favor of small (redundant) step programs over programs which had redundant and transitional materials removed. In the traditional linear program, after a learner has written his response (overt) the answer is confirmed by the presentation of the correct answer. Research on the confirmation (feedback) of results has shown conflicting results. Studies, for example, by Holland (1960), Hough and Revsin (1963), McDonald and Allen (1962), and Moore and Smith (1961, 1962) found no difference in mean scores due the added feedback. However, Meyer (1960), Kaess and Zeaman (1960), and Suppes and Ginsburg (1962) reported in their research, positive advantages for feedback on post-test scores. Homme and Glaser (1960) reported that when correct answers were omitted from linear programs, the learner felt it made no difference. Resnick (1963) felt that linear programs failed to make allowance for individual differences of the learners and he was concerned about the "voice of authority" and the "right or wrong" nature of the material to be taught. Smith and Smith (1966) believed that a "linear program is deliberately limiting the media of communication, the experiences of the student and thus the range of understanding that he achieves" (p. 293).

Holland (1965) summarized his extensive review of literature on general principles of programming and generally found that a contingent relationship between the answer and the content is important. A low error rate of responses received support, as did the idea that examples are necessary for comprehension. For long programs, overt responses are necessary. Results are equivocal concerning multiple-choice versus overt responses, however, many erroneous alternatives (e.g., multiple-choice foils) may interfere with later learning. Many of the studies, however, concerning the effects of the linear presentation of content introduced the "pall effect" (boredom) because many small steps and always being correct (Beck, 1959; Galanter, 1959; Rigney & Fry, 1961).

2.3.4.2. Intrinsic (Branching) Programming. Crowder (1961) used a similar approach to that developed by Pressey (1963) which suggested that a learner be exposed to a "substantial" and organized unit of instruction (e.g., a book chapter) and following this presentation a series of multiple-choice questions would be asked "to enhance the clarity and stability of cognitive structure by correcting misconceptions and deferring the instruction of new matter until there had been such clarification and education" (Pressey, 1963, p. 3). Crowder (1959, 1960) and his associates were not as concerned about error rate or the limited step-by-step process of linear programs. Crowder tried to reproduce, in a self-instructional program, the function of a private tutor; to present new information to the learner and to have the learner use this information (to answer questions). Then taking "appropriate" action based upon learner's responses, such as, going on to new information or going back and reviewing the older information if responses were incorrect. Crowder's intrinsic programming was designed to meet problems concerning complex problem solving but not necessarily based upon a learning theory (Klaus, 1965). Crowder (1962) "assumes that the basic learning takes place during the exposure to the new material. The multiple-choice question is asked to find out whether the student has learned; it is not necessarily regarded as playing an active part in the primary learning process" (p. 3). Crowder (1961), however, felt that the intrinsic (also known as branching) programs were essentially "naturalistic" and keeps students working at the "maximum practical" rate.

Several studies have compared, and found no difference between, the type of constructed responses (overt versus the multiple-choice response in verbal programs) (Evans, Homme, & Glaser, 1962; Hough, 1962; Roe, Massey, Weltman, & Leeds, 1960; Williams, 1963). Holland (1965) felt that these studies showed, however, "the nature of the learning task determines the preferred response form. When the criterion performance includes a precise response...constructed responses seems to be the better form; whereas if mere recognition is desired the response form in the program is probably unimportant (p. 104).

Although the advantages for the intrinsic (branching) program appears to be self-evident for learners with extreme individual differences, most studies, however, found no advantages for the intrinsic programs over branching programs, but generally found time saving for students who used branching format (Beane, 1962; Campbell, 1961, 1962; Glaser, Reynolds, & Harakas, 1962; Roe, Massey, Weltman, & Leeds, 1962; Silberman, Melaragno, Coulson, & Estavan, 1961).

2.3.5 Instructional Design

Behaviorism is prominent in the roots of the systems approach to the design of instruction. Many of the tenets, terminology, and concepts can be traced to behaviorial theories. Edward Thorndike in the early 1900's, for instance, had an interest in learning theory and testing. This interest greatly influenced the concept of instructional planning and the empirical approaches to the design of instruction. World War II researchers on training and training materials based much of their work on instructional principles derived from research on human behavior and theories of instruction and learning (Reiser, 1987). Heinich (1970) believed that concepts from the development of programmed learning influenced the development of the instructional design concept.

By analyzing and breaking down content into specific behavioral objectives, devising the necessary steps to achieve the objectives, setting up procedures to try out and revise the steps, and by validating the program against attainment of the objectives, programmed instruction succeeded in creating a small but effective self-instructional system - a technology of instruction. (Heinich, 1970, p. 123)

Instructional design is generally considered to be a systematic process which uses tenants of learning theories to plan and present instruction or instructional sequences. The obvious purpose of instructional design is to promote learning. As early as 1900 Dewey called for a "linking science" which connected learning theory and instruction (1900). As the adoption of analytic and systematic techniques influenced programmed instruction and other "programmed" presentation modes, early instructional design also used learning principles from behavioral psychology. For example, discriminations, generalizations, associations, etc. were used to analyze content and job tasks. Teaching and training concepts such as shaping and fading were early attempts to match conditions and treatments and all had behavioral roots (Gropper & Ross, 1987). Many of the current instructional design models use major components of methodological behaviorism such as specification of objectives (behavioral), concentration on behavioral changes in students, the emphasis on the stimulus (environment), (Gilbert, 1962; Reigeluth, 1983). It is this association between the stimulus and the student response which characterizes the influence of behavioral theory on instructional design (Smith & Ragan, 1993). Many of the proponents of behavioral theory as a base for instructional design feel that there is an "inevitable conclusion that the quality of an educational system must be defined primarily in terms of change in student behaviors" (Tosti & Ball, 1969, p. 6). Instruction, thus, must be evaluated by its ability to change the behavior of the individual student. The influence of the behavioral theory on instructional design can be traced from writings by Dewey, Thorndike and, of course, B.F. Skinner. In addition, during World War II, military trainers (and psychologists) stated learning outcomes in terms of "performance" and found the need to identify specific "tasks" for a specific job (Gropper, 1983). Based upon training in the military during the second world war a commitment to achieve practice and reinforcement became major components to the behaviorist developed instructional design model (as well as other non-behavioristic models). Gropper (1983) indicates that an instructional design model should identify a unit of behavior to be analyzed, the conditions that can produce a change and the resulting nature of that change. The unit of analysis is the stimulus-response association. When the appropriate response is made and referenced after a (repeated) presentation of the stimulus, the response comes under the control of that stimulus.

Whatever the nature of the stimulus, the response or the reinforcement, establishing stable stimulus control depends on the same two learning conditions: practice of an appropriate response in the presence of a stimulus that is to control it and delivery of reinforcement following its practice. (Gropper, 1983, p. 106)

Gropper (1983) stated that this need for control over the response by the stimulus contained several components; practice (to develop stimulus construction) and suitability for teaching the skills.

Gagné, Briggs, and Wager (1988) have identified several learning principles which apply centrally to the behaviorial instructional design process. Among these are contiguity, repetition, and reinforcement in one form or another Likewise, Gustafson and Tillman (1991) identify several major values which underline instructional design. One, goals and objectives of the instruction need to be identified and stated; two, all instructional outcomes need to be measurable and meet standards of reliability and validity. Thirdly, the instructional design concept centers on changes in behavior of the student (the learner).

Corey (1971) identified a model which would include the above components. These components include:

1. Determination of Objectives - This includes a description of behaviors to be expected as a result of the instruction and a description of the stimulus to which these behaviors are considered to be appropriate responses.
2. Analysis of Instructional Objectives - This includes analyzing "behaviors under the learner's control" prior to the instruction sequence, behaviors that are to result from the instruction.
3. Identifying the Characteristics of the Students - This would be the behavior that is already under the control of the learner prior to the instructional sequence.
4. Evidence of the Achievement of Instruction - This would include tests or other measures which would demonstrate whether or not the behaviors which the instruction "was designed to bring under his control actually were brought under his control" (p. 13).
5. Constructing the Instructional Environment - This involves developing an environment that will assist the student to perform the desired behaviors as response to the designed stimuli or situation.
6. Continuing Instruction (Feedback) - This involves reviewing if additional or revised instruction is needed to maintain the stimulus control over the learners behavior.

Glaser (1965) also, described similar behavioral tenets of an instructional design system. He has identified the following tasks to teach subject matter knowledge. First, the behavior desired must be analyzed and standards of performance specified. The stimulus and desired response will determine what and how it is to be taught. Secondly, the characteristics of the students are identified prior to instruction. Thirdly, the student must be guided from one state of development to another using predetermined procedures and materials. Lastly, a provision for assessing the competence of the learner in relation to the predetermined performance criteria (objectives) must be developed.

Cook (1994) recently addressed the area of instructional effectiveness as it pertains to behavioral approaches to instruction. He notes that a number of behavioral instructional packages incorporate common underlying principles that promote teaching and student learning and examined a number of these packages concerning their inclusion of 12 components he considers critical to instructional effectiveness:

1. Task analysis and the specification of the objectives of the instructional system.
2. Identification of the entering skills of the target population, and a placement system which addresses the individual differences amongst members of the target population.
3. An instructional strategy in which a sequence of instructional steps reflects principles of behavior in the formation of discriminations, the construction of chains, the elaboration of these two elements into concepts and procedures, and their integration and formalization by means of appropriate verbal behavior such as rule statements.
4. Requests and opportunities for active student responding at intervals appropriate to the sequence of steps in #3.
5. Supplementary prompts to support early responding.
6. The transfer of the new skill to the full context of application (the facing of supporting prompts as the full context takes control; this may include the fading of verbal behavior which has acted as part of the supporting prompt system).
7. Provision of feedback on responses and cumulative progress reports, both at intervals appropriate to the learner and the stage in the program.
8. The detection and correction of errors.
9. A mastery requirement for each well-defined unit including the attainment of fluency in the unit skills as measured by the speed at which they can be performed.
10. Internalization of behavior that no longer needs to be performed publicly: this may include verbal behavior which remains needed but not in overt form.
11. Sufficient self-pacing to accommodate individual differences in rates of achieving mastery.
12. Modification of instructional programs on the basis of objective data on effectiveness with samples of individuals from the target population.

2.3.6 Task Analysis and Behavioral Objectives

As we have discussed, one of the major components derived from behavioral theory in instructional design is the use of behavioral objectives. The methods associated with task analysis and programmed instruction stress the importance of the "identification and specification of observable behaviors to be performed by the learner" (Reiser, 1987, p. 23). Objectives have been used by educators as far back as the early 1900's. Although these objectives may have identified content that might be tested, usually they did not specify exact behaviors learners were to demonstrate based upon exposure to the content (Reiser, 1987). Popularization and refinement of stating objectives in measurable or observable terms within an instructional design approach was credited by Reiser (1987) and Kibler, Cegala, Miles and Barker (1974) to the efforts of Tyler (1934), Bloom, Engelhart, Furst, Hill & Krathwohl (1956), Mager (1962), Gagné (1965), Glaser (1962) and Popham and Baker (1970). Kibler and colleagues (1974) point out there are many rational bases for using behavioral objectives. Some of which are not learning theory based, such as teacher accountability. They list, however, some of the tenets that are based upon behavioral learning theories. These include 1) assisting in evaluating learners' performance, 2) designing and arranging sequences of instruction, 3) communicating requirements and expectations and providing and communicating levels of performance prior to instruction. In the Kibler et al. (1974) comprehensive review of the empirical bases for using objectives, they found only about 50 studies that dealt with the effectiveness of objectives. These researchers reported that results were inconsistent and provided little conclusive evidence of the effect of behavioral objectives on learning. They (1974) classified the research on objectives into four categories. These were:

1. Effects of student knowledge of behavioral objectives on learning. Of 33 studies, only 11 reported student possession of objectives improved learning significantly (e.g., Doty, 1968; Lawrence, 1970; Webb, 1971; Olsen, 1972). The rest of the studies found no differences between student possession of objectives or not (e.g., Baker, 1969; Brown, 1970; Weinberg, 1970; Patton, 1972; Zimmerman, 1972).
2. Effects of specific versus general objectives on learning.Only two studies (Dalis, 1970; Janeczko, 1971) found that students receiving specific objectives performed higher than those receiving general objectives. Other studies (e.g., Lovett, 1971; Stedman, 1970; Weinberg, 1970) found no significant differences between the form of objectives.
3. Effects on student learning of teacher possession and use of objectives. Five of eight students reviewed found no significant differences of teacher possession of objectives and those without (e.g., Baker, 1969; Crooks, 1971; Kalish, 1972). Three studies reported significant positive effects of teacher possession (McNeil, 1967; Piatt, 1969; Wittrock, 1962).
4. Effects of student possession of behavioral objectives on efficiency (time).Two of seven studies (Allen & McDonald, 1963; Mager & McCann, 1961) found use of objectives reducing student time on learning. The rest found no differences concerning efficiency (e.g., Loh, 1972; Smith, 1970).

Kibler, et al. (1974) found less than half of the research studies reviewed supported the use of objectives. However, they felt that many of the studies had methodological problems. These were: lack of standardization of operationalizing behavior objectives, unfamiliarity of the use of objectives by students, and few researchers provided teachers with training in the use of objectives. Although they reported no conclusive results in their reviews of behavioral objectives, Kibler et al. (1974) felt that there were still logical reasons (noted earlier) for their continued use.