Diseño basado en la ciencia del aprendizaje y requisitos de desarrollo: Una actualización de “Diseño para diseñadores” de Hendrix y Tienmann

T. V. Joe Layng

Change Partner

Descargar / Download PDF

Abstract

As new “technologies of tools” emerge for the delivery of instruction, it is of critical importance that we apply “technologies of process,” derived from the basic and applied learning sciences, to the design and development of instruction (Layng & Twyman, 2013). Those who would design effective instruction must master a range of skills, competencies and objectives that have been derived from over forty years of research and application. Mastery is not easy and requires significant training. This article describes in detail these requirements. They are divided into six separate domains: analysis, design, media selection, testing (formative evaluation), system management, and research evaluation. The requirements listed here provide a guide for anyone looking to become an effective instructional designer capable of making use of current and emerging software and hardware technologies.

Keywords: learning science, matrix, instructional design, instruction

Resumen

Conforme emergen nuevas “tecnologías de las herramientas” para llevar a cabo instrucción, es de importancia crítica que apliquemos “la tecnología del proceso”, derivada de la ciencia del aprendizaje tanto básica como aplicada, para el diseño y desarrollo de la instrucción (Layng & Twyman, 2013). Aquellos que quieran diseñar instrucción efectiva deben dominar un rango de habilidades, competencias y objetivos que se han generado después de cuarenta años de investigación y aplicación. El dominio no es fácil y requiere entrenamiento significativo. Este artículo describe en detalle estos requisitos. Se dividen en seis dominios separados: análisis, diseño, selección de medios, pruebas (evaluación formativa), manejo de sistemas y evaluación de la investigación. Los requisitos enlistados proveen una guía para cualquiera que busque convertirse en un diseñador instruccional efectivo y capaz de hacer uso de tecnologías de software y hardware que están emergiendo actualmente.

Palabras clave: ciencia del aprendizaje, matriz, diseño instruccional, instrucción

 In 1971 Hendrix and Tiemann wrote, “A new professional is needed to assist in the complex process of developing quality instruction. … [this individual] will be ‘a versatile and highly trained instructional designer.” To meet this goal a group from the Upper Midwest Regional (educational) Laboratory (USA) painstakingly set out what instructional developers needed to do if high quality instruction were to be developed. That effort was based on the extensive work done in the learning laboratory and in the application of learning theory to educational problems. Hendrix and Tiemann produced a series of objectives originating in that report which provided a guide for anyone looking to master the competencies required to become an effective instructional developer. Over the years, their objectives have remained remarkably descriptive of what it takes to design high quality instruction. With some additions and modifications, much of what they described is entirely germane to the design of quality learning environments of all types across a vast range of delivery systems and technologies. Accordingly, much of what they specified is included here with some editing, updating, and expanding upon their original work. In addition, some entirely new items have been incorporated into the requirements. It is a testament to the work done by the investigators of that day that their work remains, and will continue to be, relevant now, and into the future. Today, however, instructional design is an eclectic field that finds many of its practitioners working in the corporate environment. As such, the instructional design field, for the most part, has moved away from its learning science and education focus to become more expedient in its applications, thus requiring a somewhat different, though overlapping, set of skills (see Koszalka, Russ-Eft, & Reiser, 2013, for a good overview of the competencies for instructional designers of this type). In contrast, the requirements described here find their roots in the scientific study of learnng and behavior, and are reflective of basic and applied research as well as application. New technologies are continually being developed that are providing new opportunities for practitioners who have met the requirements to make full use of them. The requirements described here are an attempt to make explicit what is required, and serve as a guide for those who wish to contribute to the science and technology of the instructional/learning process and its application (see Layng & Twyman, 2013, for a discussion on the distinction between the technology of tools and the technology of process in teaching and learning). The following learning science design and development requirements are divided into six categories, an addition of one from the original five. Further, more recent work on learning hierarchies and generative instruction have been added. These are terminal requirements—each requirement would have a considerable list of “enabling objectives” that would lead to mastery. The goal is to provide a list of requirements necessary for a new generation of educational entrepreneurs to achieve that will allow them to make full use of the new technology of educational tools. The entire design process has been described elsewhere (see Leon, Layng, & Sota, 2011; Twyman, Layng, Stikeleather, & Hobbins, 2004), but in brief, it begins with the analysis of content and the formulation of objectives, followed by the construction of criterion evaluation, the identification of the entry repertoire, the design of instruction, the use of performance data during developmental testing to revise the design, the creation and use of maintaining consequences to ensure student engagement, and the application of a range of planning and management skills to ensure optimal development. This process is represented in Figure 1. The requirements listed do not follow the design sequence since many of them are used in the various phases of development. In addition, unlike the design process, the competency categories listed need not be acquired in any specific order.

Analysis

The design of instruction or learning environments begins with analysis. Designers must have a clear idea of what the accomplishments of learners will be when they complete the instruction or other learning activities designed to teach. This begins with what is called a content analysis and culminates in objectives based upon that analysis. Often, these two processes interact: sometimes we have a goal or objective we would like to achieve and other times we must understand the subject matter to formulate an objective. Classifications of skills based upon types of learning have evolved over the past few decades to help with that task. The one used as a reference here is a modification by Layng (2007; see also Sota, Leon, & Layng, 2011) of a classification scheme developed by Philip Tiemann and Susan Markle (1991). A more detailed description can be found therein, but a brief description here may be helpful. In this classification

Figure 1. A diagram adapted from Markle and Tiemann (1967) representing the steps of the design process.

scheme, three main types of learning are represented by columns in a matrix: Differentiated Relations (Psychomotor), Discriminative Relations (Simple Cognitive), and Extended Relations (Complex Cognitive). Three types of behavioral units are represented as rows in the matrix: Basic Units, Linked Units, and Combined Units. The intersection of the cells provide a description of a type of learning. A simplified version of the matrix is shown in Figure 2. In the first column, the emphasis is on the topography and fluent performance of the response or response sequence. Discrimination is often important to the successful completion of a sequence or in the production of complex response combinations; however, the emphasis is on the precision of which the response can be performed (e.g., ice skating, gymnastics, etc.). In the second column, the opposite holds. The emphasis here is on the discrimination. The response is typically considered to be in the student’s repertoire, and the focus is on under what conditions the response will be performed. The relations here range from simple occasion–behavior relations (paired associates) to successive conditional discriminations (algorithms) to verbal repertoires1 that are a product of often complex combinations and recombinations of response sequences. One defining characteristic of this column is that the occasion–behavior topographies taught are the same as those that are tested. The third column describes extended relations, which means that the specific occasion–behavior topographies taught are not the ones tested. In essence, the extension to new conditions is required. For intradimensional concepts, the critical features guiding responding are found within a single stimulus or event. This means, for example, that a new example of a single chair not provided in instruction would be correctly identified as a chair while a new nonexample (i.e., a love seat) would be excluded. For interdimensional concepts, the critical attributes that define the relation between at least two events (stimuli) would be learned and applied to new situations not involving the original stimuli (e.g.,things that are opposite, distant, wider, larger, smaller, or that require a coordinate concept to be taught such as me, you, believe, know, etc.). Principles similarly are applied to new situations. Principles describe the relation between concepts and the range of application is limited by the smallest range of extension of a component concept (see Tiemann & Markle, 1991). Principles also involve “when” aspects that set the conditions under which the principle holds. The generative repertoires are those that result in producing behavior that meets requirements (typically encountered in problem solving) not already in the student’s repertoire or that are not a “simple” extension of a previously taught relation. These repertoires typical are categorized as reasoning, inquiry, analytical thinking, and more advanced problem solving (see for example Robbins, 2011). For the designer, analysis of content described by these types of learning are essential to successfully building learning environments and to adequately describing

Figure 2. Matrix depicting a categorization of types of learning

objectives. Objectives comprise four qualities; they describe the context (including tools to be used), the domain (such as the range of examples and nonexamples used in concept testing), the behavior, and the performance standard. Without a thorough content analysis it is nearly impossible to write worthwhile objectives and lack of adequate analyses often results in limited or vague objectives. A popular taxonomy is Benjamin Bloom’s analysis of learning domains (Bloom, Engelhart, Furst, Hill, & Krathwohl, 1956). Bloom’s taxonomy is not an analysis of content, but a representation of general behaviors that comprise a range of skills. Even if objectives are to be stated in accord with Bloom, a content/task analysis is still required. This becomes clearer when we consider where Bloom’s learning types fall within the matrix used here. We find the first level, Knowledge, falling into the simple cognitive column and primarily into the basic units cell and some of the linked units cell. Depending on the precise performance requirement, different design considerations may be required for two different performances both of which might be consider “knowledge.” The same is true for the other levels as well: Comprehension, verbal repertoire and discriminative sequences; Application, categorical relations, and some ordered relations; Analysis, the entire discriminative relations column combined with elements from the basic and linked units in the extended relations category; Synthesis, elements from all three columns; Evaluation, similarly involves all three columns, however, here there is particular emphasis on interdimensional concepts. It is the criteria which are identified as what is being compared varies. One sees that the increasing difficulty found in Bloom’s taxonomy comes from the increasing combination of different components necessary to perform as required. Recent revisions (Krathwohl, 2002) have changed some of the learning level labels, order, and qualities, but all “Bloom” tasks, old or more recent, must be analyzed to find the precise relations required. The section below describes the analysis requirements for effective instructional or learning environment design.

I. Analysis Requirements

1. Given a relatively specific description of what learners are expected to do, perform a deficiency analysis which determines a. whether a performance deficiency exists and, if so, b. whether instructional or behavioral management problems, or both, need to be resolved.
2. When specific tasks must be learned, perform a task analysis and specify terminal objectives for the instruction.
3. When given a general educational goal, or the results from a task analysis, use a learning types hierarchy to analyze the content to be taught/learned in terms of differentiated (psychomotor), discriminative (simple cognitive), and extended (complex cognitive) relations.
   1. Given a differentiated (psychomotor) performance to be learned, perform acontent analysis to determine if response learning, chain or other sequence learning, or kinesthetic repertoires are required.
   2. Given a discriminative (simple cognitive) performance to be learned, perform a content analysis to determine if occasion – behavior relations (paired associates), symmetric occasion – behavior relations (two-way paired associates), multiple occasion – behavior relations (multiple discriminations), conditional occasion – behavior relations, serial criterion sequences, successive criterion sequences, conditional criterion sequences, or verbal repertoires are required.
   3. Given an extended (complex cognitive) performance to be learned, perform a content analysis to determine if intra-dimensional concept learning, interdimensional concept learning, disjunctive concept learning, metonymical occasion – occasion relations, stimulus equivalence relations, ordered relations (joined concepts, principles, rules, conditional equivalences, etc.), or generative repertoires are required.
4. Given a production task (essay writing, robot building, etc.), analyze it into its components, and submit the components to a content analysis.
5. Given a loosely stated educational goal, probe the stater’s behavior to obtain further specificity. (The stater could be oneself but not necessarily.)
   1. Discriminate between objectives stated in specific behavioral form and objectives which are not so stated.
   2. Write objectives that include a statement of context (the tools or environment), domain (the pool from which examples will be drawn), the behavior required, and the performance standard to be met.
   3. Given generally stated objectives, further define these in whatever detail necessary for statement in specific behavioral terms.
6. Given a behaviorally-stated objective, select those parts of it requiring further analysis techniques, e.g., illustrate the range (or generative characteristics)
7. Given a set of terminal objectives, select applicable learning strategies as determined by the content analysis.
8. Given an existing instructional module, state the terminal objectives implicit in the module.
9. Given an existing test for an instructional module, state the terminal objectives implicit in the test covering the module.
10. Given an existing instructional module, specify the minimum prerequisite behaviors required of learners.
    1. Given a specific behavioral objective and associated content for sample prerequisite skills, discriminate between essential and nonessential prerequisite skills (those that are likely to be absent versus those are necessary, but likely to be in the student’s repertoire).
    2. Given a specific behavioral objective and associated content, specify in terms of behavior and content all prerequisite skills required of the learners.
11. Given the elements of a decision system with all dependencies specified, prepare a PERT-type (decision tree, algorithm) analysis.
12. Employ various heuristics for determining the “worth” (value) of a given set of potential course objectives.
13. Employ various other heuristics to determine the attainability (feasibility) of a given set of potential course objectives.
14. Identify the responses, response sequence or kinesthetic repertoires contained in a given course or unit-sized chunk of conceptual subject matter.
15. Identify the occasion–behavior relations, multiple discriminations, discriminative sequences or verbal repertoires contained in a given course or unit-sized chunk of conceptual subject matter.
16. Identify the concepts, intradimentional or interdimensional, contained in a given course or unit-sized chunk of conceptual subject matter.
17. When prospective learners must “understand” a stated concept, perform a concept analysis and specify terminal behavioral objectives, the achievement of which will satisfy subject matter experts.
18. In conjunction with a subject matter expert (or group thereof), determine the critical attributes of each concept identified, the varying attributes of typical examples, and a set of close-in nonexamples of and each concept.
19. On the basis of the analysis of each concept, construct a conceptual hierarchy relating each concept to the others in terms of superordinate, coordinate, and subordinate relationships.
20. Given a set of superordinate and subordinate concepts, describe the superordinate attributes that are inherited by subordinate concepts and distinguish between critical and defining properties of the subordinate concept (polygon: 2-D, closed figure, straight lines; triangle: polygon with 3 sides –inherits all critical attributes of superordinate polygon, with 3 sides the defining subordinate attribute).
21. Identify the behaviors (usually written or spoken) that may be linked in a two-way (symmetric) association with concept examples (seeing an example or wider— saying wider; hearing wider—selecting wider example; saying wider— having someone hand you the wider of two objects).
22. Identify stimuli that share no features, but guide the same response (stopping at red light, stop sign, policeman with raised hand; place on a table, on a book, on a car).
23. Identify and analyze any joined concepts contained in a given course or unitsized chunk of complex subject matter (two or more concepts act together, e.g., small car).
24. Identify the principles contained in a given course or unit-sized chunk of complex subject matter.
25. Determine the critical concepts embedded in the principle, and the relation between the concepts, often stated in terms of “if, then” relations.
26. Determine the generative procedures required to provide students with any inquiry skills required.
    1. Identify components that can be combined into new composite repertoires.
    2. Determine if the application of specific reasoning training (prescribed supplementary verbal stimulation that restricts response alternatives in accord with problem requirements) is appropriate.
    3. Determine if students need to acquire component repertoires “on their own” as a component not provided as part of instruction to be combined with those provided.
27. Identify any repertoires required to interact with the technology requirements of the teaching/learning environment (computers, tablets, simulation equipment, etc.).
28. Given specific behavioral objectives, determine appropriate practice strategies for the objective.
29. Discriminate between repetitive and deliberate practice.
30. Discriminate between spaced versus massed practice.
31. Discriminate between accurate performance and fluent performance.
32. Given a particular performance to be learned, determine the frequency of performance that would reasonably indicate the performance is fluent.
33. Given an instructional sequence, determine the points in the sequence where there will be sufficient practice to ensure fluency.
34. Given an instructional objective and a set of reinforcers, discriminate between the appropriate and inappropriate reinforcers, based on the characteristics of a population of learners.
35. Given an instructional objective, specify effective and/or efficient reinforcers, based on the relevant characteristics of a population of learners.
36. Discriminate between program–extrinsic and program–specific reinforcers, and determine which one or what mix of each may be used in the sequence.

Design

Designing instruction or creating learning environments involves the sequencing of activities that take a student from their entry repertoire to the repertoire described in the objectives. A range of sequences, problems, projects, inquiries, and discoveries can be designed to accomplish this outcome (see Johnson & Layng, 1992; Markle, 1990). The section below describes the design requirements for effective instructional or learning environment design.

II. Design Requirements

1. With access to a set of the typical students, devise measures of the familiarity of points in the content analysis hierarchy, for example, by testing recognition of examples and nonexamples of each concept. (It is not the case that low level concepts are necessarily learned before high level ones, despite the “logic” of such a statement.)
2. Based on the results of the empirical data coordinated with a logical analysis, “logically” sequence instruction on unknown relations (e.g., concepts) by proceeding from familiar to unfamiliar. Some of these sequences will involve training of discriminating subclasses within a known larger class (e.g., will classes involve of “insects”). Other sequences will involve training in extending to a broader class made up of known subclasses (e.g., class “herbivores” made up from cows and deer).
3. For composite repertoires, ensure each component is acquired prior to asking for the composite performance.
4. Based on the logical analysis, construct an appropriate relational (e.g., concept) learning criterion consisting of new examples and nonexamples of each class. (Actually, they are “new” in the sense that they will now be kept out of the instructional sequence and reserved for the testing sequence.)
5. In designing content for teaching sequences, select widely differing examples and juxtapose these to create maximum extension.
6. In designing content for teaching sequences, select the closest nonexamples possible for each critical attribute of the class, and juxtapose these to create fine discriminations.
7. Depending on knowledge of the verbal sophistication of the target of the students, adjust the verbalization of the differences between examples and nonexamples and the similarity between examples and further examples to as simple a level as required.
8. Reserve further examples and nonexamples for use in an expanded program should developmental testing demonstrate that the teaching is insufficient.
9. In developmental testing, probe for student difficulties with varying attributes not located by subject matter experts but proving distracting to students and with distinctions proving too difficult.
10. Check each expansion of the program against items reserved for the test to determine that one is still holding to the basic conceptual objective and is not “teaching the test.”
11. Given a specific problem-solving process to be mastered, construct exercises that require the process to already-known phenomena and then extend to novel problems from the same domain.
12. Given a set of problems to be solved (within or across domains) design sequences that require talk aloud problem solving qualities, beginning with simpler problems and building to more complex.
13. Given a set of objectives and an intended population of learners, determine an appropriate level of student achievement at which developmental testing will cease. (The level reasonably may be raised or lowered based on practical experience and cost requirements during development.)
14. Having determined there is sufficient time to do so, design the leanest possible draft of materials to achieve a given set of objectives and let student difficulties and queries determine the expansion at appropriate points.
    1. Given specific behavioral objectives, discriminate between appropriate and inappropriate content for objectives.
    2. Given specific behavioral objectives, generate rules for discrimination between appropriate and inappropriate content for the instructional model.
    3. Given a specific behavioral objective, describe examples of appropriate content.
    4. Given a terminal objective and intermediate objectives, sequence the intermediate objectives in an effective manner.
    5. Given a terminal objective and all necessary intermediate objectives, specify all the dependencies among the intermediate objectives and indicate potential branching (for adaptive sequences) before determination of student trial.
15. Given a recording or transcript of student feedback, make changes in the instructional materials which retain the original objectives but reduce the likelihood of error by the next student. 
16. Given a set of discriminations to be learned, arrange them according to least potential for interference, if the learning ability of the students makes this possible. 
17. Given a set of component behaviors, determine whether a logical hierarchy is implied in sequencing these for instruction. 
18. Given a logical hierarchy, validate its existence by appropriate tryouts with students.
19. Given a set of objectives with no predetermined sequence, select an appropriate first draft sequence when time does not permit further exploration with students. 
20. If time permits further exploration, determine which sequencing appeals most to students by observing their inquiry behaviors. 
21. Given a principle to be mastered, sequence it in the instruction in such a way that almost all elements in it have been mastered. 
22. Make first-draft adaptations of timing, vocabulary and sentence length, and question difficulty appropriate to medium. 
23. Given a specific behavioral objective and content, construct elicitors, cues, problems, etc. 
24. Given a specific behavioral objective, and content describe appropriate responses. 
25. Given the essential specifications for instructional modules (objectives, content criterion measures, elicitors, cues, problems, etc.), write effective instructional modules.
26. Given summary data from the administration of an instructional module, identify ineffective frames, sequencing, stimuli, etc. 
27. Given a set of student entry repertoires and the amount of instructional testing time available, determine the degree of adaptation to build into the instructional sequence. 
28. Design specific adaptive sequences based on a content analysis in terms of the hierarchy of learning types (for example, in response to a concept’s nonexample being identified as an example, a specific branch to more instruction discriminating nonexamples of that specific type would provided). 
29. Given a set of items to be memorized by students, design an instructional system which will permit individualization of practice according to student need. 
30. Given a program sequence, arrange for the occurrence of program–extrinsic reinforcers that may be required (points, badges, praise; making the subject matter fun) 
31. Given a program sequence, arrange for the occurrence of program–specific reinforcers that may be required (finding the fun in the subject matter), and arrange for their occurrence.