(ED), (131) · The IPISD Model 35 The Dick and Carey Model 38 The See Is and Glasgow Model 41 The Diamond Model 43 SUMMARY 47 REFERENCES 50 ANNOTATED ERIC BIBLIOGRAPHY 53 Periodicals

ED 335 027

AUTHORTITLE

INSTITUTION

SPONS AGENCY

REPORT NOPUB DATECONTRACTNOTEAVAILABLE FROM

PUB TYPE

EDRS PRICEDESCRIPTORS

ABSTRACT

DOCUMENT RESUME

IR 015 277

Gustafson, Kent L.; Powell, Gary C.Survey of Instructional Development Models with anAnnotated ERIC Bibliography. Second Edition.ERIC Clearinghouse on Information Resources,

Syracuse, N.Y.Office of Educational Research and Improvement (ED),

Washington, DC.IR-91; ISBN-0-93597-32-591RI8806200877p.; For the first edition, see ED 211 097.Information Resources Publications, SyracuseUniversity, 030 Huntington Hall, Syracuse, NY13244-2340 ($7.50 plus $2.00 shipping and

handling).Information Analyses - ERIC Clearinghouse Products

(071) -- Reference Materials - Bibliographies (131)

MF01/PC04 Plus Postage.Annotated Bibliographies; Curriculum Development;*Instructional Development; *Instructional Systems;*Models; *Organizational Development; SystemsAnalysis; *Systems Approach; Systems Development

This paper, which updates and expands on the first

edition (1981), begins by presenting a definition of instructional

development (ID) and a discussion of its origins. A taxonomy for

classifying ID models is then suggested, and it is noted that most ID

models have been subjected to only a limited degree of testing. The

12 ID models that are reviewed are divided into the three categories

specified by the taxonomy as they focus on the classroom, the

product, and the systems. Specific models are described and discussed

in earth of these categories: (1) classroom--Gerlacil and Ely; Heinich,

Molenda, and Russell; Dick and Reiser; and Kemp; (2) product--Van

Patten; Leshin, Pollock, and Reigeluth; and Bergman and Moore; and

(3) systems--/DI (Instructional Development Institute); 1SPISD

(Inservices Procedures for Instructional Systems Development); Dick

and Carey; Seels and Glasgow; and Diamond. The taxonomy and 11 of the

12 moJels are depicted in 12 figures. A guide to searching ERIC for

ID models is provided in the Foreword, and the annotated ERIC

bibliography lists 19 journal articles and 14 documents. Instructions

for obtaining copies of ERIC documents and journal articles are

included. (30 references) (DB)

***********************************************************************Reproductions supplied by EDRS are the best that can be made

from the original document.*********************************************A*************************

SURVEY OF INSTRUCTIONAL

DEVELOPMEN

by Kent L Gusta son

U.S. DEPARTMENT OF EDUCATIONOffice of Educational Research and Improvement

EDUCATIONAL RESOURCES INFORMATIONCENTER (ERIC)

0 This document has been reproduced asreceived from the person or organizationoriginating it.

0 Minor changes have been made to improverproduCtion Quality.

Points of view or opinions stated in this docu-ment do not necessarily represent OficialOERI posttion or policy.

'

,

cta

BE

with cm Annotated ERIC Bibliography by

Gary C. Powell2.

,

.

Ilk

SECOND EDITION

Ir 1

L(

E I

MARINGHOMI

OH INF ORMAIION

R1SOURH S

SYRAM1

.1111IVE [NH

1991

e

This publication is available from Information ResourcesPublications, Syracuse University, 030 Huntington Hall,Syracuse, New York 13244-2340 (IR-91; $7.50 plus $2.00shipping and handling).

Kent L. Gustafson is a Professor in the Department ofInstructional Technology, the College of Education,University of Georgia, Athens, Georgia 30602.

Gary C. Powell is a Doctoral Candidate in the Departmentof Instructional Technology, College of Education,University of Georgia.

ISBN: 0-937597-32-5

This publication was prepared with funding from the Office ofEducational Research and Improvement, U.S. Department ofEducation, under contract no. RI88062008. The opinionsexpressed in this report do not necessarily reflect the positionsor policies of OERI or ED.

Table of Contents

Foreword

Finding Models in ERIC

INTRODUCTION 1

PurposeDefinition of Instructional Development 2

Why Models? 3

Early ID Models 4

A Taxonomy of ID Models 6

CLASSROOM ID MODELS 10

Assumptions 10

The Gerlach and Ely Model 11

The Heinich, Molenda, and Russell Model 14

The Dick and Reiser Model 16

The Kemp Model 18

PRODUCT DEVELOPMENT MODELS 22

Assumptions 22

The Van Patten Model 23

The Leshin, Pollock, and Reigeluth Model 25

The Bergman and Moore Model 28

SYSTEMS DEVELOPMENT MODELS 31

Assumptions 31

The IDI Model 32

The IPISD Model 35

The Dick and Carey Model 38

The See Is and Glasgow Model 41

The Diamond Model 43

SUMMARY 47

REFERENCES 50

ANNOTATED ERIC BIBLIOGRAPHY 53

Periodicals 53

ERIC Documents 57

How to Order ERIC Documentsand Journal Articles 67

Table of Figures

Figure L A Matrix of Types of Modelsand Selected Characteristics 8

Figure 2. The Gerlach and Ely Model 12

Figure 3. The Dick and Reiser Model 17

Figure 4. The Kemp Model 19

Figure 5. The Van Patten Model 24

Figure 6. The Leshin, Pollock, andRegieluth Model 26

Figure 7. The Bergman and Moore Model 29

Figure 8. The IDI Model 33

Figure 9. The IPISD Model 36

Figure 10. The Dick and Carey Moael 39

Figure 11. The See Is and Glasgow Model 42

Figure 12. The Diamond Model 44

Foreword

Although much has been written about instructionaldevelopment in recent years, most scholars tend to agreethat instructional development, regardless of tne form amodel may take, is a process of systematically designing,sequencing, implementing, evaluating, and constantlymonitoring instruction with the intent of improving itsquality and effectiveness, and thereby enhancing learning.In the same vein, instructional development lids aibu beenoften referred to as the systems approach. And, not surpris-ingly, one can cite many references in which the systemsapproach has been defined using the same terms employedto define instructional development.Confusion also proliferates in the attempt to define instruc-tional technology and educational technology. One doesnot have to read many journals in the field to find the termsinstructional technology, educational technology, systemsapproach, and instructional development usedinterchangeably.Nevertheless, the conviction of most ISD professionals isthat instructional development is a team-oriented processaimed at product development; and the product is to beused by people. As Gustafson said in 1971: "The most im-Portant element of instructional development is people. ..people are its energy, its insight, its product and its con-sumer and to enge3e in instructional development is tochange people."

Finding Models in ERICIn May of 1990, the ERIC Clearinghouse on InformationResources conducted an ERIC search which included -

publication dates ranging from 1980 to March of 1990 withthe following search terms:

(Instructional Development/major OR InstructionalDesign/major)

AND(Models/major OR Theories/major)

This search resulted in 161 citations. As an aside, they wereoriginally asked to conduct a search using the followingsearch terms, but they resulted in far too many citations:

(Instructional-Design OR Curriculum-Design ORInstructional-Development OR Instructional-ImprovementOR Educational-Technology OR Curriculum-Development

OR Educational-Strategies)AND

(Models- OR Research-Design OR Systems-Approach ORTheories-)

Of the original 161 citations, 33 titles (19 periodicals and 14ERIC documents) were found to be h'.ghly relevant, and soare included in this annotated bibliography.It is interesting to note the publication dates of these 33periodicals and ERIC documents. Four are dated 1980 orearlier, 17 are dated between 1981 and 1985, and only 11 aredated between 1986 and 1990. This would seem to indicatethat the number of publications about ID modelseitherdescribing a new one or augmenting an oldis eitherdecreasing or leveling off.Regardless of age, the following annotated bibliographyrepresents the documents entered in the ERIC databasefrom 1980 to March 1990 that have a significant link toInstructional Development (models).

Gary C. Powellwith

Kent L. GustafsonAthens, GeorgiaApril 1991

INTRODUCTION

The purpose of this ERIC synthesispaper is to update and expand on ear-lier publications by Twelker et al. (1972)and Gustafson (1981) on the topic ofinstructional development (ID) models.Since the first appearance of ID models

in the sixties, there has been an ever increasing numberappearing in the literature of both instructional technologyand genera! education. This publication reviews some ofthe current ID models, presents a taxonomy for classifyingthem, and describes trends in their content and focus. A listof references and an annotated bibliography of selected IDmodels found in the ERIC database are also provided.In preparing this survey it was necessary to select only afew models to describe in detail. This was difficult sincethere are literally hundreds in the literature within and out-side of what might be considered the mainstream in instruc-tional technology. Decision criteria included the historicalsignificance of the model, its unique structure or perspec-tive, and its general distribution. Obviously too, it wasnecessary to select models to match each of the categoriesin the classification taxonomy. Thus, many excellent modelsare not included in this survey. The decision was also madeto exclude models that represent only part of the process,e.g., needs assessment, media selection, or evaluation.Those that were selected are believed to be representativeof the literature and, among them, to contain all of the mainconcepts found in other models.

1

One of the major problems plaguing thefield is inconsistent use of terminology.The term instructional development isno exception. Although an attempt wasmade over a decade ago by a committeeof professionals under the leadership ofDr. Kenneth Silber and the Aswciation

for Educational Communications and Technology todevelop a standard set of definitions, they have not beenwidely adopted. It is their definition, however, that is usedto structure this survey, although, as will be notedthroughout the descriptions of models, they use a variety ofother terms for essentially the same concepts and processes.The AECT definition is as follows:

Instructional development. A systematic approach tothe design, production, evaluation, and utilization ofcomplete systems of instruction, including allappropriate components and a management patternfor using them; instructional development is largerthan instructional product development, which is con-cerned with only isolated products, and is larger thaninstructional design, which is only one phase ofinstructional development. (1977, p. 172)

As can be seen, this definition encompasses a wide array ofactivities, from the initial sensing of a concern that "some-thing" ought to be done, to implementing and monitoringthe instruction that has been developed on a long termbasis. Consistent with the above definition, the term instruc-tional design has a more limited meaning, usuallyassociated with specifying the sequence of instructionaltopics and activities, the nature of interactive lesson ele-ments, and motivational plans. However, in recent yearsthe term instructional design has become increasingly usedas the broader term to represent the er tire process, whilethe term instructional development is used to describe whatsome have called production activities. If this trend con-tinues, the title of any subsequent ERIC review paper mightwell be changed from instructional development to instruc-tional design.Another term that adds to the confusion is "system." Theterm system is used in at least three different ways, one of

2

which is equivalent with how we have chosen to useinstructional development in this survey. When used torepresent the broad array of activities we call instructionaldevelopment, the term instructional systems development(ISD) is commonly employed. In essence, the model be-comes a system for guiding the preparation of instructionto accomplish specific goals and objectives. However, theterm system is also used by some authors to describe theoutcomes or products in the development effort. From theirperspective, the learners, environment, and its relatedmanagement and support components together comprisean instructional system. Still a third, but less common useof the term system, is in the context of general systemstheory (GST). Within this perspective, numerous generalsystem theory concepts (e.g., open and closed systems,entropy, and interdependence) are applied to the instruc-tional development process.In some respects the profession finds itself in an Alice inWonderland setting where any term means whatever theauthor wants it to mean. This situation is one of the reasonsthis author has found it desirable to create a taxonomy forclassifying models. By carefully examining each one youcan determine what activities their creators are describingand the goals and settings in which the activities are tooccur. You are then in a position to understand what theyare talking about even though the terminology isinconsistent.

Instructional dc ielopers and theirmodels have often been compared toLinus and his blanketyou never seeone without the other! While otherprofessionals s'aare this trait,developers have elevated it to newheights. In the physical and natural

sciences, models serve a variety of purposes, includingtheory building and testing, description, prediction, and ex-planation. However, developers seem to have more limitedpurposes in mind. With apologies to the very few theoriatsin our field, ID prectitioners typically use models primarilyas: (1) communication devices with their clients and eachother; (2) planning guides for management activities; or (3)prescriptive algorithms for decision making. While these

3

purposes can overlap, most models tend to focus on asingle purpose.This focus on a single purpose is understandable. If amodel is to communicate with an unskilled client, it mustbe simple and devoid of professional jargon. On the otherhand, to be a useful management tool, it should account forall c' the major tasks to be performed. And, if it is to beprescriptive, it must contain extensive detail describingprecisely how to perform specified tasks. Thus models varywidely in their purposes, the amount of detail provided,and the technical jargon they contain. No single model isuseful for all settings and purposes, a point to be elaboratedlater in this paper. We now turn our attention to some ofthe early ID models and the origins of the ID process.

Of necessity, one must pick an arbitrarydate from which to begin to trace theorigins of the ID model buildingprocess. Otherwise one can make thecase that the creators of the earliestrecorded cave drawings and the scribesthat produced papyrus scrolls representthe pioneers of systematic instruction.

Similarly, many ideas and procedures commonly found inID models (e.g., job analysis, measurable objectives, andperformance testing) predate the period generally acceptedas representirg the beginnings of ID model building.

The specific term instructional development, defined as a sys-tematic process for improving instruction, appears to havehad its origin in a project conducted at Michigan StateUniversity from 1961 to 1965. Entitled Instructional SystemsDevelopment: A Demonstration and Evaluation Project (1967),this project directed by Dr. John Barson produced one ofthe early ID models. The setting for the model and itsrelated project was higher education, with the goal of im-proving college courses. Barson's model was reviewed inthe first ERIC paper by Twelker et a/. (1972). The reader isalso referred to the Barson project final .,-cport (ED 020 673)for more details. The Barson model is notable in that it isone of the few models ever subjected to evaluation. The Bar-son project also produced a set of heuristics for instruction-al developers, many of which continue to serve as generalguides to developers in higher education.

4

Ot.ier early works by a number of authors also produced IDmodels, although they did not use the specific term instruc-tional development. The programmed instruction movementalso used a systematic process, but generally did not recog-nize the major contribution of the tryout and revisionprocess to the successes it recorded. In the 1950s and 1960sone of the most influential model builders was L. C. Silvern(1965). His work with the military and aerospace industryresulted in an extremely complex and detailed model (withvariations) that drew heavily on general systems theory.The model is not widely circulated today, but remains anexcellent source document for those willing to wadethrough Silvern's sometimes obscure writing. Students ofthe ID process will readily see his influence on the contentof contemporary models.The model by Hamreus (1968), developed at the TeachingResearch Division of the Oregon State System of HigherEducation, is another classic. One of his significant contribu-tions was to present the model in a "maxi" and a "mini"version. This "two-size" approach was based on the beliefthat there is a need for a simple model to communicate withclients and a more detailed operational version for thoseworking on thc project. Hamreus' model evolved into theInstructional Development Institute (IDI) model (1971). Thelatter model has received extremely wide distribution andis among the best known in the United States. In fact, theIDI model was recently reproduced and described by See Isand Glasgow (1990) in their book on the ID process. SinceHamreus' model was extensively reviewed h.; the Tweikerpaper, the reader is referred there for details. However, theIDI model, because of its wide circulation and notoriety,will be discussed in a later portion of this monograph.In addition to the Twelker review, at least three other majorreviews of ID models have been conducted and are worthyof study by developers. As early as 1972, Stamas reviewed23 models by determining whether or not each included alist of model components that he felt were desirable.Originally part of a doctoral dissertation at Michigan. StateUniversity, this study (Stamas, 11;72) was also reproducedas an occasional paper by the Division of InstructionalDevelopment of the Association for Educational Com-munications and Technology. In 1980, Andrews and Good-

5

son reviewed 40 models in the Journal of InstructionalDevelopment. Like Stamas they developed a matrix of ID ele-ments and analyzed the models using those elements. Theyalso attempted to trace a logical progression or evolution oflater models from earlier ones, but were unable to detectany pattern. Their findings add weight to the view that theliterature on models is circular rather than cumulative, withlittle of substance being added in the last few years. Morerecently, Salisbury (1990) reviewed a number of ID modelsfrom rnajor textbooks in the field to assess the degree towhich they contained specific references to a range ofgeneral systems theory concepts. He concluded that mostmodels contained few specific references to those generalsystems concepts contained in his matrix.

A scholarly wit once said that only twothings are certain: death and

A taxonomies. This survey is no exceptionon the latter point. The presenttaxonomy is presented as having twobenefits. First, the author has foundthat creating a taxonomy is an excellentmeans of reducing an otherwise unwiel-dy body of ID model literature into a

manageable package. The literature contains dozens of IDmodels, but as will be discussed, most are simply variationson one of three basic structures. The taxonomy provides avehicle foT examining the assumptions and conditionsassociated with any ID model and places it in a scheme foreasy understanding. Second, practicing developers can usethe taxonomy to assist in analyzing the type of project onwhich they are about to embark. Then it is possible to selecta model for adaptation to the specific shuation. Theapproach helps to eliminate the "I have a model, nowwhat's your problem" syndrome by better matching amodel to the situation.The proposed taxonomy divides the world of ID modelsinto three categories: (1) classroom focus; (2) product focus;and (3) systems focus. The reader is cautioned, however,that like most taxonomies in the social sciences, thecategories are a little fuzzy around the edges and notmutually exclusive. In fact, some models can exist as sub-sets of others, so no absolute hierarchy should be inferred.

6

The first category of ID models has a classroom focus. Themodels assume that there is already a teacher, some stu-dents, a curriculum, and a facility. The goal of the teacher isto do a better job of instruction within these constraints.The development situation often occurs when a teacherwants to improve his/her teaching. The teacher is frequent-ly not part of a team and improvement will be limited tohis/her own classroom, and only for as long as he/shechooses to use whatever results. Emphasis is usually placedon selecting and adapting existing materials and instruction-al strategies rather than developing them "from scratch."A product focus is different from a classroom focus in thatits goal is production of one or more specific instructionalproducts. It usually assumes that development of theproduct is a given. Further, the product's objectives may bealready determined. The goal is to prepare an effective andefficient product as quickly as possible. The product isusually expected to produce replicable results with anaudience possessing specified characteristics. Productmodels are common in business settings where decisions onwhether or not development should be done are made bysomeone other than the developers. Often, but not always,this decision is made in the absence of objective data.Proprietary and off-the-shelf courses are also examples ofproducts that may have been developed using a productfocused model.A systems focus is somewhat different from a productfocus, but the latter in some cases may become a subset ofthe former. The systems focus has as its goal the develop-rnent of instructional output, which itself is considered tobe a system. The output of the development effort may in-clude materials, equipment, a management plan, and per-haps an instructor training package. This "system" can thenbe implemented or disseminated to target locations. The sys-tems focus usually demaniis extensive analysis of: (1) theuse environment, (2) char acteristics of the task, and (c)whether or not development should even take place. Itoften employs a problem solving approach requiring datacollection to determine the precise nature of the problem.A matrix comparing the three classes of models on selectedattributes is presented in Figure 1. In reviewing this figure,keep in mind the caveats that the taxonomy is not "pure" in

7

°° Selected Characteristics

TypicalOutputs

AmountResources Team or Emphasis on Front-End Amount

Committed to Individual Dew. or Select Analysis/Needs Tryout Distribution./Dew. Process Dev. Materials Assessment and Revision Dissemination

ClassroomOrientation

Hour ofInstruction

VeryLow

Ind. Select Noneto

Low

Lowto

Med.

None

ProductOrientation

SelfInstructionalPackage

High Ind.or

Team

Dev.or

Select

Lowor

Med.

VeryHigh

High

SystemsOrientation

School, College,or MilitaryCounc(l))

High Team De v. VeryHigh

Med.to

High

Med.to

High

Figure 2. A matrix of types of models and selected characteristics.

a scientific sense, and that many models can be and havebeen used in a variety of instructional settings.Figure 1 portrays a matrix of selected characteristics thatoften differ among the three classes of models. For example,classroom oriented models typically focus on creating oneor a few hours of instruction, whereas product models oftenfocus on producing self-instructional packages, and systemmodels result in entire courses or curricula. The number ofresources available is usually quite different, ranging fromvery low for classroom oriented efforts to very high for sys-tem oriented projects. Teams are more commonly employedfor developing products and systems, while classroomteachers often have to work on their own or with minimalassistance. The degree to which materials are selected ordeveloped varies, with classroom teachers often havinglittle choice but to select, whereas product and systemdevelopers usually create original materials. Similarly,individual teachers usually are not able to conduct an exten-sive front-end analysis, but system developers consider thatstep essential as an early step in the process. The amount ottryout and revision usually varies from none by classroomteachers to extensive for product and system developers.And lastly, classroom teachers rarely distribute theirinstructional lessons whereas products and systems arespecifically designed for this purpose.Keeping these characteristics in mind when reviewingmodels will help you to identify their major assumptions,which are often not described by their creators. Applyingthese characteristics to your analysis will also help you toplace any model in one of the three classes presented in thistaxonomy.

9

CLASSROOM ID MODELS

A,ssumptionsClassroom ID models are primarily ofinterest to professional teachers whoaccept as a given that their role is toleach and that students require someform of instruction. Users include

elementary and secondary school teachers, communitycollege and vocational school instructors, and universityfaculty. Some training programs in business and industryalso assume this classroom orientation, but the systemsfocus is becoming more common in such settings.

As indicated above, there exist a wide variety of classroomsettings. Most teachers assume (with real justification) thatstudents will be assigned to or will enroll in their classes,and that there will be "n" number of class meetings, each of

"t" length. The teacher's role is to decide on appropriatecontent, plan instructional strategies, identify appropriatemedia and strategies, deliver the instruction, and evaluatelearners. Due to the on-going nature of the instruction,which is often accompanied by a heavy teaching load, thereis little time for developing new materials. Also, funds andtime for development are usually limited. Hence their need

is to identify existing resources for adaptation to existingconditions, rather than engaging in original development.Also, since many elementary and secondary teachers teachany topic only once a year, they have less concern for therigorous formative evaluation associated with courses andworkshops that are offered on a repetitive basis.

Teaching personnel usually view any ID model as a generalroad map to follow. Typically only a few functions are out-

lined in the model, and it simply provides a guide to the

10 Li

teacher. It should be noted that although there are a num-ber of classroom oriented ID models, they are not widelyknown to or adopted by teachers. The developer whoworks with teachers within the givens and assumptionsdescribed above would do well to employ any ID modelwith caution since the teachers are unlikely to be familiarwith the concepts or processes of systematic design.Teachers may also view the process as mechanistic andresulting in dehumanized instruction. However the modelsdiscussed below have been found to be acceptable to andreadily understandable by at least some teachers and repre-sent a class of models with which all developers should befamiliar.Four models have been selected to represent the variety ofID models most applicable in the classroom environment.The respective authors are: (1) Gerlach and Ely, (2) Heinich,Molenda, and Russell, (3) Dick and Reiser, and (4) Kemp.

The entry point the Gerlach and Ely(1980) model (see Figure 2) calls foridentifying content and specifyingobjectives as simultaneous, interactiveactivities. While Gerlach and Ely clearlyprefer the approach of specifying objec-tives as a "first task," they recognizethat many teachers first think aboutinstruction from the standpoint of con-tent. Their model is one of only a few

that recognize this content orientation of teachers. Be-havioral objectives are to be written and classified beforemaking several design decisions. Their classificationscheme is based on Gerlach's other scholarly work andpresents a five-part cognitive taxonomy with singlecategories for affective and motor skill objectives.The next step in their model is assessing the entry be-havior of learners, a step common to many classroomoriented models. However, despite the specification ofentry behavior as a major step in the ID process, few con-crete procedures describing how to accomplish this step areprovided. The next step includes five activities to be per-formed simultaneously. These activities are viewed as inter-active, with any decision in one area influencing the rangeof decisions available in the others; e.g., the design process

11

Specificadon

CONTENT

A

Areeeenteittof

ENTERINGBEHAVIORS

Specification

OBJECTWES

ofSTRATEGY

Orpnintionof

atoms

Allocation

of

Tea

Allocation

of

SPACE

Selection

of

Evaluation

of

PERFORMANCE

Analysis

of

FEEDBACK

F;gure 2. A systematic approach to instruction.(Gerlach/Ely/Melnick, Teaching and media: Asystematic approach (2nd ed.) Copyright C)1980,p. 11. (Reprinted by permission of Allyn &Bacon, Needham, MA.)

12

is itself a system. The five activities are: (1) determinestrategy, (2) organize groups, (3) allocate time, (4) allocatespace, and (5) select resources.Under strategies they posit a continuum from exposition(all cues) to discovery (no cues). The teacher/designer'srole is to select one or more strategies along this con-tinuum. Students can be organized into configurations rang-ing from self-study to whole-class activities based onstrategies, space, time, and resources. Time is viewed as aconstant to be divided up among various strategies. Spaceis not a constant, since it is pointed out that teachers canand should extend learning experiences beyond the class-room. Also, the classroom itself can usually be rearrangedfor different grouping patterns.Selection of resources focuses on the teacher's need tolocate, obtain, and adapt or supplement existing instruction-al materials. Emphasis is placed on where and how to findsuch resources and the importance of previewing and plan-ning for their use as a part of the overall instructionalstrategy. This emphasis on selecting rather then developinginstructional materials is a common feature of classroomoriented ID models.Following these five simultaneous decisions is evaluationof student performance. This step directs the teacher/designer's attention to measuring student achievement andtheir attitude toward the content and instruction. Evalua-tion is seen as closely linked to the learner objectives statedearlier, with attention also directed to evaluating the "sys-tem" itself. The last step in their model is feedback to theteacher regarding the effectiveness of the instruction. Feed-back focuses on reviewing all earlier steps in the modelwith special emphasis on reexamining decisions regardingthe objectives and strategies selected.The Gerlach and Ely model is a mix of linear and simul-taneous design and development steps. Several steps areseen as simultaneous, but the model is generally linear inits orientation. Its main strength is that practicing class-room teachers can readily identify with the process itdescribes. Its objectives classification taxonomy is simpleand non-threatening to teachers. Also, the authors relatetheir taxonomy to specific instructional strategies. Its mainweakness is that it may unintentionally reinforce teachers

13

and administrators in maintaining existing organizationsand staffing patterns rather than re-examining the entirebasis of how schools shbuld operate.

Heinich, Molenda, and Russell (1989)present their classroom oriented instruc-tional development modelASSUREin what is currently the most widelydistributed college text on instructionalmedia and technology. While somemight argue it is not a complete orformal instructional developmentmodel, teachers can readily identifywith the planning process it describes,and its wide circulation alone would

warrant its inclusion in this review. Unlike most ID models,ASSURE is not portrayed in graphic or pictorial form.ASSURE is simply an acronym for:

A - analyze learnersS - state objectivesS - select media and materialsU - utilize materialsR - require learner participationE evaluation/review

The A for Analyze Learners acknowledges the importanceof determining the entry characteristics of learners.Heinich, Molenda, and Russell caution teachers that it isnot feasible to analyze all learner attributes. They suggestonly selected "general characteristics" (e.g., grade level,job/position, and cultural and economic factors) andselected "specific entry competencies" (e.g., knowledge,technical vocabulary, attitudes, and misconceptions) beexamined. They also suggest that "learning style" (anxiety,aptitude, visual/auditory preference, etc.) be considered,but acknowledge problems in defining and measuring thesechcracteristics.Their second step, S for State Objectives, emphasizes theneed to state the desired outcomes of instruction in specificand measurable terms. A rationale for stating measurableobjectives is presented, including their role in strategy and

14

media selection, assasment of learning, and comi.unicat-ing the intent of the instruction to learners.The second S in thir model, Select Media and Materials,recognizes that most teachers have little time for designingand developing their own materials. However, they do dis-cuss the option of modifying existing materials and indicatethat sometimes original development may be possible. Theprocedures and criteria they prevent for selecting mediaand materials provide useful guidelines to teachers and tothose assisting teachers in that task.The U, or Utilize, step in their model describes howteachers need to plan for utilizing the selected media andmaterials in the classroom. The practical advice they offerrecognizes the realities of most American classrooms andthe fact that teachers play the central role in deliveringmost instruction.The R, Require Learner Participation, step in the ASSUREmodel emphasizes the importance of keeping learnersactively involved. The roles of feedback and practice arealso described. While one might question why learner par-ticipation is singled out over other design considerationsfor elevation to being a step in the ASSURE model, Heinich,Molenda, and Russell obviously think it to be of primaryimportance.The last step in their model, E for Evaluation/Review, is inreality two steps, evaluation and review. They discuss theimportance of evaluating the "total picture" to assure bothlearner achievement of the objectives and the feasibility ofthe instructional process itself. Review is then plannedbased on discrepancies between the intended and actualoutcomes.Although Heinich, Molenda, and Russell's model focuseson media and materials selection and utilization in contrastto a wider view of the ID process, their model has much tooffer classroom teachers. The obvious relationship of itssteps to their "real world" and its practical guidance andstructure make it easy to understand and apply. Further,the well written text and accompanying teacher's manualare an excellent resource for teaching teachers the rudi-ments of the ID process.

15

In their recent book, Dick and Reiser(1989) present both an outline for aninstructional plan and a developmentmodel (they use the term design) forpreparing and implementing the plan(see Figure 3). While their planningstructure and model are clear and welldeveloped, they add to the confusion ofterminology in the field by using

design to describe both the overall process and one of itselements. They write, "A formal definition of instructionaldesign is: a systematic process for designing, developing,implementing, and evaluating instruction. While the worddesign is repeated in the definition, it is used to representthe entire process as well" (1989, p .3).In presenting their model, Dick and Reiser first present theskeleton of the plan they propose that teachers prepare.They then present their development model. Since therelationship of the plan to the model seems somewhatobscure, we will turn our attention directiy to their model.Dick and Reiser's model starts with setting goals, afterwhich specific measurable objectives are to be written. Theimportance of textbooks to classroom instruction is acknow-ledged by the authors in the interactive link they make be-tween writing objectives and selecting or reviewing thetextbooks teachers plan to use. The prominent role oftextbooks in classrooms is further indicated by the arrowfrom textbooks back to the setting of goals. The relationshipof learner characteristics to the content of objectives is indi-cated by the single headed arrow from learner charac-teristics to objectives.In a somewhat unusual next step, they link both text selec-tion/review and objectives to test development. They thenprescribe developing instructional activities followed bychoosing instructional media. It is interesting to note thatHeinich, Molenda, and Russell reverse these two steps, afactor one may wish to consider in deciding which model toemploy or use to communicate with others. After choosingmedia. the next step is to implement the instruction, whichin their case also includes evaluating its results. The laststep is to review the instruction by returning to the point at

4, J16

Select ReviewTestis

WriteObjectives

Analyze StudentCharacteristics

DevelopTests

Developinstructional

Activities110

Chooseinstrudional

Media

Revise Instruction 1.4

ImplementInatruction

Figure 3. Model for developing effective instruction. (From Planning effective instruction by Walter Dickand Robert A. Reiser. Copyright ©1989. Used with permission of Allyn & Bacon.

4.

which you wrote the objectives as you plan for the nextoffering of the instruction.Dick and Reiser's model and accompanying narrative arestraightforward and easy to comprehend. The process theydescribe is influenced by their experience and expertise ineducational psychology, measurement, and evaluation.While the amolint of detail concerning how to perform eachstep in the process is somewhat limited, the text does pro-vide a sound introduction to the concepts of instructionaldevelopment for classroom teachers.

Jerrold Kemp's model (1985) (see Fig-ure 4) is similar in a number of ways toGerlach and Ely's. He states that thereare four essential elements of instruc-tional technology: (I) students; (2) whatmust be learned (objectives); (3) whatprocedures and resources will workbest to reach desired learning levels

(methods); and (4) how we will know when the requiredlearning has taken place (evaluation). Kemp's model com-municates his belief that ID is a continuous cycle withrevision as an on-going activity associated with all of theother elements. He feels that the teacher/designer can startanywhere and proceed in any order. This is essentially ageneral system view of development wherein all elementsare interdependent and may be performed simultaneouslyif appropriate.Although Kemp's model indicates that the developer canstart anywhere, its accompanying narrative is presented ina conventional framework starting with TopicsJob Tasksand Purposes. The classroom orientation of the model isapparent t'.rough Kemp's choice of words, topics, and sub-ject content for determining what will be taught. Thesewords can be readily accepted by classroom teachers.

Kemp's second element is to enumerate important LearnerCharacteristics. These include such academic factors asnumber of students, GPA, IQ, and reading level, and suchsocial factors as age, maturity, and attention span.

The model's third element is to conduct a Subject ContentTask Analysis.

18 27

.15

Figure 4. (From The instructional design process by Jerrold E. Kemp. Copyright ©1985 by Harper & Row,Publishers, Inc. Reprinted with permission of Harper Collins Publishers.)

28

The fourth element is the specification of Learning Objec-tives. He suggests using Bloom's taxonomy for categorizingobjectives and ensuring that a broad range of objectives isincluded.Elen-.ent five, Teaching/Learning Activities, is the step atwhich decisions are made regarding instructionaistrategies, grouping, media, and oher resources. Kempcombines into this step most of what Gerlach and Ely haveseparated into five elements in their model. For this step hepresents a set of "principles for successful learning" anddescribes a variety of teaching and learning models withaccompanying lists of their advantages and disadvantages.

Element six, Instructional Resources, contains descriptionsof a variety of media and their relative attributes.Element seven, Support Services, deals with the budget,facilities, time, equipment, personnel, and materialsrequired to engage in the development process.Element eight, Learning Evaluation, addresses differentforms of testing and stresses the importance of matchingtest requirements of previously stated objectives.

Element nine, Pre-testing, describes the importance of deter-mining learners' prior knowledge of the content of thecourse under development and their general abilities thatare relevant to successful achievement of the objectives.

The "band" that encircles the model includes Revision (acontinuous process) and Formative and Summativeevaluation, also conceived as on-going in the classroomenvironment.The box in the center of the modelLearning Needs,Goals, Priorities/Constraintsdeals with the general ques-tion of whether instruction is appropriate or needed, and ifso why and what its general goals are.

From a teacher's perspective, the strength of Kemp's modelis the concept of starting "where you are." Also, the em-phasis on subject matter content, goals and purposes, andselection of resources makes it attractive to teachers. Itsmajor weaknesses are the lack of specification in the ele-ments dealing with teaching/learning activities, formativeevaluation, and revision. The version of Kemp's modelreviewed here is different than the one reviewed in the ear-

20

lier edition of this survey. The current version placesgreater emphasis on both formative and summative evalua-tion as on-going processes and places all activities withinthe context of Goals, Priorities and Constraints. Kemp isone of very few authors who has modified his ID modelover time.

30

21

PRODUCT DEVELOPMENT MODELS

Product development models are char-. acterized by four key features: (1) they

usually assume that the instructionalproduct is needed; (2) they assumesomething should be produced rather

than selected or modified from existing materials; (3) theyplace considerable emphasis on tryout and revision; and (4)they assume that the product must be usable by a variety of

"managers" of instruction. The assumption of need shouldnot necessarily be considered a limitation of these models.In some settings a front-end analysis has already been con-ducted and needs determined for a variety of products. The

task then becomes developing the products efficiently andeffectively. Also, in a number of situations, the need is soobvious, that it is unnecessary to ask "should," but only"what" should be done. An example would be the necessityfor developing an operator training package for a newmachine that is about to be marketed.

Extensive tryout and revision often accompany productdevelopment because the client cannot, or will not, toleratelow performance. Also, the performance level may be exter-nally established; e.g., the user must be able to use all of thecapabilities of the word processing software. This is in con-trast to classroom settings where the performance level is

often subject to considerable up or down adjustment basedon the effectiveness of the instruction. Cosmetic appearanceof the product may also be important to clients, thusmaking subjective evaluation an important part of thetryout process. Use of the product by managers as opposedto teachers simply means the product is often required to

22

stand on its own without a content expert available to thelearner. An example would be training for a telephone com-pany lineman on how to install a specialized piece of equip-ment. The demand for free standing products is anotherreason for emphasizing tryout and revision in productdevelopment. As computer-based instruction has becomemore popular, the demand for effective instructionalproducts has increased and is likely to expand even morerapidly in the future. Hence, the demand for efficient andeffective prescriptive development models, unique to avariety of settings and products, will probably accelerate inthe decade of the 90s.Product models often contain elements that might qualifythem as systems models (to be presented in the next sec-tion). Those chosen for this review were selected based onthe belief that they are primarily focused on creating instruc-tional products rather then more comprehensive instruction-al systems. The three models reviewed are by Van Patten(1989), Leshin, Pollock, and Reigeluth (1990), and Bergmanand Moore (1990). For examples of earlier product modelssee the previous edition of this publication.(Gustafson,1981).

Van Patten's model is described in thechapter he wrote for the book Instruc-.tional Design: New Alternatives for Effec-tive Education & Training edited byJohnson & Foa (1989). His introductionto his model is a little confusing be-cause he first talks about the instruc-tional design process as performed by

instructional designers to produce products. Later, how-ever, he talks about the process as Instructional SystemsDesign (ISD).

Van Patten claims that the ISD model which he presentsand describes (see Figure 5) can be used to create paper-based instructional materials. It has nine phases, eachhaving a deliverable, one or more persons responsible forits execution, and one or more persons responsible for itsevaluation. These phases are: Analysis, Design, Develop-ment, Pilot Test, Review, Production, Duplication, Im-plementation, and Maintenance. Analysis includes definingthe problem, identifying the audience, determining

3 2 23

Phan Deliverable Created by Evaluated by

Analysis Analysis Report Designer/Evaluator ClientDesign Specification Designer Client/SMEDevelopment Draft materials Developer/Designer Client/SMEFA4 test Test results Designer/Evaluator Client/SMERevon Rnal materials Developer/Editor aient/SMEProductkm Camera-ready Editor/Graphks Client/SMEDuplication Inventory Graphks/Printer Client/AdministratorImplementation Training begins Instructor/Administrator ClientMaintenance Periodic Evaluations Instructor/Designer Client

Administrator/Evaluator

Figure 5. (From J. Van Patten: What is instructional design? Reprinted with permission of MacmillanPublishing Company from Instructional design: New alternatives for effective education andtraining, K. A. Johnson & L. K. Foa (Eds.) Copyright ©1989 by the American Council andMacmillan Publishing Company, a Division of Macmillan, Inc.).

resources, and specifying the goals of the effort. TheDesign phase involves preparing the "floor plan" and "penand ink" renderings of design specifications. The Develop-ment phase has four subphases; developing definitions ofeach topic, developing examples for each definition,developing practice exercises for the examples, and develop-ing "everything else." Phases four and five, Pilot Test andReview, are described together as an interactive loop that isrepeated until the instruction is judged "good enough."Phase six, Production, is the step at which all materials areput through final production and prepared for duplication.Duplication is essentially the task of building an inventoryof material in preparation for its distribution. Phases eightand nine, Implementation and Maintenance, are describedtogether as an interactive loop that takes place as long asthe product continues to be used.Van Patten's model is similar to other product models inthat it specifies extensive tryout (Pilot Test and Review)before the product is finalized. It also specifies implementa-tion and maintenance, activities not always associated withproduct development. If the product is publicly marketed,no formal implementation or maintenance (in Van Patten'suse of the terms) would likely occur. Van Patten's model isquite serviceable as a general guide, but its lack of opera-tional detail limits its use to those already familiar withspecific procedures for performing the activities hedescribes.

Leshin, Pollock, and Reigeluth (1990)have developed a model (see Figure 6)that they claim redresses the shortcom-ings of other models which "have notincluded any guidance for the selectionand use of instructional strategies andtactics" (p. 1). While the creators ofmany of the other models might dis-agree with their statement, their pointis well taken. Many models place heavyemphasis on the analysis side of

activities with the actual design of instruction receiving con-siderably less attention. Leshin, Pollock, and Reigeluth'smodel is significantly influenced by Reigeluth's and others'earlier work in elaboration theory and recent developments

34 25

Overview of the Instructional Design Process

AnalysingNeeds

Selecting &Sequendng Content

DevelopingLesions

Evaluatingthe instruction

Step IAnalyze Problem

Step 3Analyze and

Sequence Task

Step 5Spedfy Learning

Events & Activities

Step 7Evaluate

instruction

Step 2Domains F.

Step 4Analyze and Sequence

Supporting Content1.11101i

Step 6Perform Interactive

Message Design

Figure 6. Overview of the instructional design process. (From Instructional design: Strategies and tacticsfor improving learning and performance by C. Leshin, J. Pollack, and C. Reigeluth. Copyright ©by Educational Technology Publications, in press. Used with permission of EducationalTechnology Publications.)

35

in cognitive psychology. Although the graphic repre-sentation of their model appears to be linear, they em-phasize the cyclical and non-linear nature of thedevelopment process in the accompanying narrative.Leshin, Pollock, and Reigeluth's model contains only sevensteps, which are clustered under four headings: AnalyzingNeeds, Selecting and Sequencing Content, Developing Les-sons, and Evaluating the Instruction. Step one is to Analyzethe Problem, which can be a performance deficiency in atraining situation or simply a lack of knowledge in aneducational setting. Identifying the audience, clearly statingthe problem, determining possible solutions, and com-municating the results are all part of this first step. Steptwo, Analyze the Domains, contains four subcomponents:identify tasks, identify performance deficiencies, write per-formance objectives, and develop performance measures.Step three, Analyze and Sequence Tasks, contains eightcomponents not elaborated upon in this review, but heavilyinfluenced by elaboration theory. Step four, Analyze andSequence Supporting Content, is also based on Reigeluth'searlier work and provides considerable detail on how toperform these tasks. Step five, Specify Learning Events andActivities, involves classifying each piece of content as totype of learning, planning instructional "strategies and tac-tics," writing practice and test items, and specifying theinstructional management plan. Step six, Perform Interac-tive Message Design, is essentially an examination of fivealternate delivery systems along with a set of general con-siderations for message design. Step seven, Evaluation, con-sists of three components: one-on-one evaluation, pilottesting, and summative evaluation via field testing.In summary, Leshin, Pollock, and Reigeluth have created aseven step model specifically to address their belief thatgreater attention is needed to what some have called thepsychological components of instructional design. Theirmodel really exists on three levels of detail, but they havenot chosen to highlight this feature. Its strength is the exist-ence of numerous "job aids" to guide the developmentprocess. These job aids will be of considerable assistance tonovice instructional developers. Its major limitations arethe lack of attention to project management and toimplementation.

36 27

Bergman and Moore (1990) recently. published a model (see Figure 7)

directed at producing "InteractiveVideo/Multimedia" products. Theirmodel focuses on one of the currentareas of keen interest in education andtraining technology. Although theirmodel includes specific reference tointeractive video (IVD) and multimedia

(MM) products, it is generally applicable for a variety of"high-tech," interactive instructional products. With only afew changes it could readily pass for one of several earlierproduct models that focused on the then current high tech-nology such as Control Data Corporation (see Gustafson,1981).

Bergman and Moore's model contains six major activities:Analysis, Design, Develop, Production, Author, andValidate. For each activity they specify input, deliverables(output), and evaluation. As can be seen, the output of eachactivity provides the input for the subsequent activity. Theyrefer to each horizontal row of their model as a "phase" andremind the reader that although not shown, it may be neces-sary to "review a phase and review selected activities."They also emphasize the importance of evaluating the out-put (deliverable) from each activity before proceeding. Thechecklists they provide for performing these evaluationsare extensive, and would be valuable even if one wereusing a different product development model for IVD orMM development.Bergman and Moore report that a request for proposal(RFP) initiates the development process. They suggest thateven if an external RFT' does not exist, preparing an internalRFP is desirable. The RFP drives Analysis activities includ-ing identification of the audience, tasks, user environments,and content. Design activities include sequencing the majorsegments and defining their treatment, labeled by Bergmanand Moore "High-Level Design." Detailed design then fol-lows and includes specification of motivational elements,media, interaction strategies, and assessment mi!thodology.Development includes preparing all of the documentsnecessary for later production. Examples of what Bergmanand Moore call "Producible Documents" are storyboards,

2837

The Development Model

hist Adbidoo -00 Delbraubbe L'albeedeu

41110

ProdudbleDoDocuments

11111.4111111m=

41/psoleiwnwrew

joductioe

amiable

ILIChidatAudwal"

adding

Figure 7. The development model. (From Managing inter-active videolmultimedia projects by Robert E.Bergman and Thomas V. Moore. Copyright©1990 by Educational Technology Publications.Used with permission of EducationalTechnology Publications.)

:4R 29

audio scripts, shot lists, art and graphics renditions, and adatabase for managing production. Production "transformsthe producible documentation into its correspondingmedium: video sequence, audio, graphic, or text." Author-ing activities integrate the individual media into the com-pleted product. Its three sub activities are coding, testing,and tuning. Validation consists of comparing the finishedproduct with its original objictives. Revision may alsooccur at this time, as can assessment of the product'sachievement of its sponsor's goals.

Development of sophisticated IVD and MM productsalmost always requires a team, a point made repeatedly byBergman and Moore. IVD and MM also require a soundmanagement system, the structure for whkh this modelprovides. This model was selected for review partially be-

cause of its focus on new technology, and partially becauseof the excellent and extensive checklists and other guidescontained in the text. Even without the model the supportmaterials are well worth examining.

3)30

SYSTEMS DEVELOPMENT MODELS

Instructional systems models are charac-s terized by four key features: (1) large

scale team development, (2) a lineardevelopment process, (3) wide distribu-tion of the system, and (4) a problem

solving orientation. The models usually begin with a datacollection phase to determine the feasibility and desirabilityof developing an instructional solution to a "prublem." Anumber of the models require that a problem be specified ina given format beflre proceeding. Thomas Gilbert's (1978),Mager and Pipe's (1984), and Rossett's (1987) work in front-end analysis is highly relevant to the models discussedherein. Their position is that, while a problem may have aninstructional solution, one should first consider lack ofmotivation and environmental factors as alternative areasof action. Systems models, as a class, differ from productdevelopment models in the amount of emphasis placed onanalysis of the larger environment before committing todevelopment. Systems models also typically assume alarger scope of effort than product development models.However, in the design, development, and evaluationphases, the primary difference between systems models andproduct models is one of magnitude rather than type ofspecific tasks to be performed.Five systems oriented models were selected for review. Thefirst two are classics that appeared in the earlier edition ofthis publication, but because of their continued notorietyare included here as well. The other three are of morerecent vintage. The five models selected for review are: (1)Instructional Development Institute ODD; (2) Interservices

31

Procedures for Irstructional Systems Development (IPISD);(3) Dick and Carey; (4) See Is and Glasgow; and (5) Diamond.

The Instructional DevelopmentInstitute (IDI) model (see Figure 8) isone of the most widely publicized IDmodels in existence. it is taught inmany professional preparationprograms, and has been the focus of anational workshop for large numbers of

public school personnel. In the earlier ERIC paper byTwelker et al. (1972), the IDI model provided the frame ofreference for analyzing other models. The model is a jointeffort of the University Consortium for InstructionalDevelopment and Technology (UCIDT), which was original-ly known as the National Special Media Institute. Createdas a tool for public school personnel who desired to tacklelarge-scale instructional problems, the IDI model is prob-lem oriented, specifies team development, and assumesdistribution or dissemination of the results of the effort. Itis similar ill a number of its steps to an earlier modelcreated by Dale Hamreus, and some developers consider itas simply a variation on his model.The IDI model is essentially linear in its approach. Theclaim is briefly ma,ae that ID can be non-linear, but the pro-cedures accompanying the graphic model provide noevidence of how this can be accomplished. The model hasthree stages and nine steps, with each step further sub-divided for a total of 24 elements. In essence, the model isconceived as being useful at ail three levels of detailstages, steps, or elements.The model is reviewed here at its intermediate level ofdetail since describing it at the 24-element level wouldresult in a lengthy description. The IDI's first step is toIdentify the Problem. This requires conducting a needsassessment, establishing priorities among various and con-flicting needs, and, finally, stating one or more problems tobe addressed. Emphasis is placed on separating symptomsfrom problems and stating problems in measurable terms.This permits later assessment of progress toward alleviat-ing or solving the stated problems. Step two (Analyze theSetting) specifies additional data collection to be performedregarding the previously stated problem. Data are collected

32 41

Instructional Development Model

Identify Problem Analyze Setting Organize Management

Ames Needs Audience Tasks

&tab lbh Notifies Conditions Responsibilities

Stale Problem Relevant Resources Timelines

Identify Obfrdivee Specify Methods Confined Pro loi)pes

Terminal (10) Learning instructionalEnabling (E0) Instruction Materials

Media Evaluation Materials

Test Prototypes Analyze Results Implement/Recycle

Conduct Tryout Obiectives ReviewCollect Evaluation Data Methods Decide

Evaluation Techniques Act

NSW /Int

Figure 8. The Instructional Development Model ODD. (From the University Consortium for Instruc-

w tional Development and Technology; formerly the National Special Media Institute).

44

concerning audience (learner) characteristics, charac-teristics of other affected personnel, conditions underwhich development must occur, constraints on any solu-tion, and what relevant material and human resources areavailable for both developing and delivering the solution.Step three is to Organize Management, including thedevelopment team. This step is somewhat unique to the IDImodel. Its creators made this step highly visible because oftheir belief that poor management often leads to failure ofdevelopment efforts. Organizing management includes stat-ing all major tasks, assigning responsibility for those tasksto team members, and establishing timelines for their com-pletion. Monitoring of progress is also included as part ofthis step. How a team is to perform steps one and twobefore becoming organized is never explained.Step four (Identify Objectives) is similar to other models inrequiring behaviorally stated objectives. The mnemonicABCD provides a helpful reminder ti.at objectives mustinclude an Audience (A), Behavior (B), Condition (C), andDegree of performance (D). Step five (Specify Methods)uses a taxonomy developed by Ed ling and Hamreus (latermodified by Merrill and Goodman, 1972) for classifyingobjectives, and then selecting strategies and media based onthe type of objective. The strategies and media prescriptionmatrix is viewed as a set of suggestions rather than a rigidmatching activity. Designers and developers are encour-aged to use whatever additional knowledge they have tomake final determinations.Step six (Construct a Prototype) prescribes buildingtestable drafts of all the materials. These include instruction-al units, teacher/manager instructions, and evaluationmaterials. The emphasis is on constructing a prototype thatis complete enough to test, but not so expensive that it can-not be changed. The seventh step specifies Testing thePrototype under conditions as similar as possible to itseventual use. This step is often called formative evaluationin other models. Step eight specifies Analyzing the Resultsin terms of learner achievement, effectiveness and prac-ticability of the methods of instruction, and appropriate-ness of the evaluation techniques. The last step in the IDImodel is to Recycle (if the data indicate a deficiency) or toImplement the solution if it is effective. Recycling to any

34 43

previous step should be considered, but it may be necessaryto return to the original problem and re-analyze needs. Itshould be noted that, in recent years, the UCIDT Consor-tium has developed a workshop on dissemination that is anextension of the model to another step, but the originalmodel has not been modified.The basic strength of this model is its three levels of detail.This permits its initial presentation to non-developers in asimple form that can be elaborated as their knowledgeincreases. Its basic limitation is the implication of a linearstep-by-step development process beginning with defini-tion of a problem. This limitation is common to many sys-tems models. Its processes are also becoming dated due to alack of refinement in over ten years.

The Interservices Procedures forInstructional Systems Development(IPISD) model (see Figure 9) is, as thename suggests, a joint effort of the U.S.military services. The Army, Navy,Marines, and Air Force created thismodel in the interest of utilizing a com-mon approach to instructional develop-

ment. The motivation was to facilitate shared developmentefforts and improve communication with contractorsengaged in instructional development across differentbranches of the military. Of course, the underlying concernof each service was to have a rigorous procedure fordeveloping effective instruction. A large number of person-nel contributed to creating the 1P1SD model; however, thename most commonly associated with it is Robert Branson.

The 1PISD model is similar to the IDI model in that it hasseveral levels of detail. At its simplest level it has fivephases: analyze, design, develop, implement, and control.These phases sub-divide into 20 steps which can be furtherdivided into hundreds of sub-steps. In fact, the IPISDmodel is one of the most highly detailed models of the IDprocess generally available. It is published as a four volumeset (Branson, 1975) and can be ordered from the NationalTechnical Information Service (NTIS) or ERIC.Since a detailed review of all the steps in this model isbeyond the scope of this survey, it will be reviewed only atthe phase level. The reader should keep in mind that the

4,4 35

Ma

11Malys*Job

1.2Wad Comma lob AssiblaTeaks/ Moroi= SAWSPowdoes Moms Come

11.1

DewlapObOctivis

112DevelopSalo

113

Bray

11A

DLI

SPdtfLewinBewee/Activities

5PsdaYInsingliceMaappwAriPM It Dewy

11L2M3Rsview/Select :LopSatinsMAW& bee:action

IV.1lotplsomeburridiooldlamegoomett

1V.2Condo*balm:boo

V.1Quota

bell111110111

CaedoetSelma Sporn

Ravin

Figure 9. IPISD model: Detailed breakdown of activitiesto be performed in each phase. (From Interser-vice procedures for instructional systems develop-ment: Executive summary and model. TRADOCPamphlet 350-30, August 1, 1975.)

36

IPISD approach is designed specifically for military train-ing in the skills/job area; most other models have a muchbroader range of intended applications. The narrower focusof IPISD is both a blessing and a bane. Its virtue is the ex-tremely detailed level of specification it contains. However,the price of this specification is its lack of generalizabilityto other environments.Phase one of IPISD (Analyze) requires specification of thetasks personnel perform on the job. Tasks which are alreadyknown or easy to acquire are subtracted, and a list of tasksrequiring instruction is generated. Performance levels andevaluation procedures are specified for the tasks and exist-ing courses are examined to determine if any of the iden-tified tasks are included. A decision is then made either tomodify the existing course to fulfill task requirements or toplan a new course. In the latter case, parts of an existingcourse may be adapted for the new one. The final step inphase one is to determine the most appropriate site forinstruction; i.e., school or non-resident instruction.Phase two (Design) begins with the arrangement of jobtasks into instructional outcomes classified by the learningelements involved, i.e., mental skills, physical skills, infor-mation, and attitudes. Tests are generated and validated ona sample of the population and instructional objectives writ-ten in behavioral form. Next, the entry behavior expected oftypical students is determined, followed by the design ofthe sequence and structure for the course. Design specifica-tions are then forwarded to phase three of the process.The development of prototype materials occurs in phasethree of the model. Development begins by specifying a listof events and activities for inclusion in instruction. Mediaare then selected and a course management plan developed.Existing instructional materials are reviewed for theirrelevance and, if appropriate, adopted or adapted for thecourse. Necessary new materials are then produced and theentire package field tested and revisedluntil satisfactorylearner and system performance are achieved. The develop-ment phase concludes when the entire course package isready for large scale implementation as phase four of themodel.

Phase four (Implement) includes training for coursemanagers in the utilization of the package, content training

AP 37

of subject matter personnel, and distribution of allmaterials to the selected sites. Instruction is then conductedand evaluation data collected on both learner and systemperformance.Phase five (Control) is the last part of the IPISD model.Internal evaluation is performed by "on-line" staff who areexpected to make small-scale changes to improve the sys-tem after each offering. In addition, they forward evalua-tion results to a central location. External evaluation is ateam effort directed toward identifying major deficienciesrequiring immediate correction. External evaluation alsofollows course graduates to the job site to assess real-worldperformance. Changes in practice in the field are alsomonitored to determine necessary revisions to the course.Thus, the emphasis in phase five is on quality control andrelevance over an extended period of time.The major strength of the IPISD model is the extensivespecification of procedures to follow during the ID process.It is an excellent reference for students who are in trainingto become instructional developers or managers of ID con-tracts. To augment its already extensive resources,Berkowitz and O'Neil (1979) prepared an annotated bibliog-raphy of additional relevant resources for tne IPISD model.Its major limitations are its narrow instructional focus andlinear approach to ID. Further, the level of analysis andprescription it specifies could be done only by a heavilystaffed, highly financed organization. Use of this modelrequires a commitment of substantial resources on a long-term basis. This model will find little use outside of themilitary, the government, and a few large corporationshaving major job training programs.

Walter Dick and Lou Carey (1900) haveproduced one of the most widely usedintroductory texts on instructionaldevelopment; hence its selection forinclusion in this publication. The Dickand Carey model (see Figure 10) mightbe considered product oriented ratherthan system oriented depending on thesize and scope of step one activities(Identify Instructional Goals). Many ofthe examples and worksheets seem to

38 47

IdentifyInstructional

Goa Ks)

Y

.1 ConductInstructions

Analysis

-

011 IdentifyEntry Behaviors,Characteristics

iiI.

1

`I

WritePerformanceObfectives

ReviseInstruction

DevelopCriterion-ReferencedTest Items

V. T

I ,--1---Develop and

SelectInsfructiona

Materials

,..-......-.).Develop

InstructionsStrategy

*

...

1

r....----,Design and

Conductill Formative&alumna,

II

II

Y.---resign andConduct

SummativeEvaluation

Figure 10. (From The systematic design of instruction, 3rd edition by Walter Dick and Lou Carey. Copyright

©1990 by Walter Dick and Lou Carey. Reprinted by permission of HarperCollins Publishers).

48

be directed at developing specific instructional products,but parts of the narrative suggest a more encompassingperspective. For our purpose we consider it to be a systemsmodel that is also applicable to projects having a morelimited focus. It should also be noted that they use the terminstructional design for the overall process, which we havedefined as instructional development.Dick and Carey's model begins with Identify InstructionalGoal(s). Here they emphasize ihe importance of decidingwhat you are trying to achieve before preceding. Two stepsare then done in parallel: Conduct Instructional Analysisand Identify Entry Behaviors and Characteristics. Theformer is vintage hierarchical analysis as conceived byGagné. The latter step specifies collecting informationabout prospective learners' knowledge, skills and attitudes.Facilitative prerequisites as defined by Gagne are also ex-amined. The next step is to Write Performance Objectivesin measurable terms. Criterion referenced test items arethen generated for each objective. One of the strengths ofthis model is its emphasis on developing tests having defen-sible validity and reliability. Developing an InstructionalStrategy for each objective comes next and again closelyhews to prescriptions promulgated by Gagné. To achievethe specified strategies the next step is to Develop andSelect Instructional Materials. To their credit Dick andCarey acknowledge the desirability of selecting as well asdeveloping materials, but the degree of emphasis devotedto development suggests they are far more interested in per-forming original development. The next step is to Designand Conduct Formative Evaluation, a process for whichthey give excellent guidance. The last step, that interesting-ly is placed on a lower line then those preceding it, isDesign and Conduct Summative Evaluation to determinethe degree to which the original instructional goals (andperhaps other unintended ones) have been achieved. Asuper-ordinate activity, Review occurs across all but thefirst and last steps in the process, indicating their belief thatthe output should be examined and revised as appropriateat all intermediate steps. However, little guidance isprovided on how to accomplish this systematic revision ateach step.

40 49

In summary, Dick and Carey have published a widelyadopted introductory text and their model is well known bymany professionals in the field. Although it is readilyapplicable to developing finite instructional products, it canalso be applied at the broader level of developing instruc-

ial systems. The text is a valuable resource, especiallyfor novice developers wanting to know more about how toengage in instructional development.

In their rece..it book Exercises in Instruc-tional Design, See Is and Glasgow (1990)review several instructional develop-ment models including some of thosereviewed here. They then present theirown model (see Figure 11), although, asthe title of the text indicates, they usethe term instructional design to definethe comprehensive process defined asinstructional development in the

present publication. The See Is and Glasgow model is inmost ways similar to other systems models. Its single some-what unique element is including project management asan encompassing activity that pervades the entire process.Their inclusion of project management in this manner wasthe basic reason for selecting this model for review.The See Is and Glasgow model specifies ten major activities.And, as noted, "the steps are undertaken within theparameters of the project management plan" (p. 57).Although Seels and Glasgow report that there is some"back and forth" activity among the steps, their visualrepresentation of the model contains no such indication.Problem Analysis (step one) includes deciding whetherthere is an instructional problem by performing a needsassessment and then stating the problem. Task and Instruc-tional Analysis (step two) involves collecting informationon performance standards and determining prerequisites.Objectives and Tests (step three) are written in behavioraland measurable form and assessed via criterion referencedtest items. Instructional Strategy (step four) involves deter-mining the components of instruction, "such as presenta-tion or practice conditions." Media Decisions (step five)include "methods and media" to meet the conditionsspecified in step four. Materials Development (step six) is

50 41

4a See Is And Glasgow Id Model

FPmblemAlislY41

Task andlnatrudional

1

Analysis

andTests

InstrudionalStrategy

4

MediaDedsions

MaterialsDevelopment

61-

1FormativeEvaluation

7I

IlitOJEa MANAGEMENT

dImplementationMaintenance

SummativeEvaluation

DiseeminationDiffusion

10

Figure 11. Instructional design model: Project management. (Reprinted with permission of Merrill, animprint of Macmillan Publishing Company, from Exercises in instructional design by BarbaraSeels and Zita Glasgow. Copyright ©1990 by Merrill Publishing Company. )

the point at which production occurs. Their description oftypical production activities is limited to instructionalmaterials, but elsewhere they indicate managementmaterials are also needed. Formative Evaluation (stepseven) includes both data collection and revision asappropriate. Implementation Maintenance (step eight)recognizes the need to plan for both short term andextended use of the instruction by users other than thedevelopers. Summative Evaluation (step nine) seems toaddress the question of whether the system is solving theoriginally stated problem, but See Is and Glasgow do notelaborate on this point. Similarly Dissemination Diffusion(step ten) is not elaborated by the authors.

In summary, the See Is and Glasgow model is quite similarto many ID models in the literature, a fact they have readilyacknowledged by comparing it to a five step generic model,i.e., Analysis, Design, Development, Implementation, andEvaluation. Despite a brief discussion of its non-linearity,they present it in essentially linear form. Their book is par-ticularly useful to novice developers due to the large num-ber of application exercises it contains. However, thetreatment of the ten steps is uneven, with those addressinganalysis receiving much more attention than later stepssuch as Implementation Maintenance and DisseminationDiffusion. An instructor's manual is also available.

Robert Diamond (1989) developed andrefined over a number of years adevelopment model (see Figure 12) thatis specific to higher education institu-tions. Although Diamond's modelmight be considered classroomoriented, we have placed it in the sys-tems category because of his belief that

development is a team effort and is often directed at totalcurricula in addition to individual courses. Diamond alsoemphasizes the need to be sensitive to political and socialissues existing on the campus and within academic depart-ments. Assuring that the proposed development effort isconsistent with organizational priorities and missions isanother critical concern to Diamond. He believes that it is a

team process with significant input from university person-nel specifically assigned to assist faculty. For these reasons,

52 43

Process For Educational Program Development

Miler SILIICTION AND DESIGN

Peeled Gam Mooand **Wm

Seek Planning ler*

Manta ol linowledieStuds* knowledge,

alludes sad ptiewilleeSocetal maleRoom*

a Educational petwItio

Proled.SvodlIc Paden

Itatennita Gosh

Twoa Reolancee

I Hee1.64mob

I WIN./ /WISloutials

/ Mob.Loafte

The "Wear 7 ISeiectioa Samoa

ESODUCTION. IMPLESSEPRATION, AND EVALUATION FOR EACH wor

rile* evaluation Winton* and preced2jee

Piero kw°blanks

endal Matingntedale

Bye luate+at 401 Pit:duce end

~Weavaileld ton am and

able- 11 LLl evalte4o, aad

mem

Figure 12. Process for educationai program development. (From Designing and improving courses andcurricula in higher education by Robert M. Diamond. Copyright ©1989 by the a. thor. Reprintedwith permission. )

53

his model seems most appropriate for classification as asystems model.Diamond's modei is divided into two phases, with Phase Iincluding Project Selection and Design, and Phase II in-cluding Production, Implementation, and Evaluation.During phase one the feasibility and desirability of launch-ing the project are examined. Instructional issues such asenrollment projections, level of effectiveness of existingcourses, and institutional priorities, in addition to facultyenthusiasm, ai e all weighed before commencing develop-ment. At this point Diamond urges thinking in terms of an"ideal" selection without regard to existing constraints. Hisargument is that by thinking in ideal terms, a team will bemore creative and innovative in outlining powerful solu-tions. Once a decision is made to begin a project, an opera-tional plan is developed that accounts for the goals,timeline, human and other resources, and student needs inthe involved department.During phase two of development, each unit of the courseor curriculum proceeds through a seven step process. Thefirst step is to determine the unit's objectives. This is fol-lowed by the design of evaluation instruments and proce-dures, a step which proceeds concurrently with selectingthe instructional format and examining existing materialsfor their possible inclusion in the system. Once these stepshave been accomplished, any new materials are producedand modifications made to any that exist but requiremodification. Interestingly, Diamond includes field testingas part of the same step as materials production althoughmost model developers make them separate steps. Alsoimplicit to this step is revision of the unit based on field testdata, but Diamond includes revision later in the process.The next to the last step is coordinating logistics forimplementation, followed by full scale implementationincluding evaluation and revision.In summary, Diamond's model is d signed specifically forhigher education environments. He emphasizes matchingthe decision on whether to engage in instructional develop-ment to institutional as well as instructional issues. He alsostresses the need to assure faculty ownershi, of the resultsof the development effort and the need for a formalorganization to support faculty's development efforts.

54 45

Diamond's model is particularly recommended for instruc-tional developers working in higher education institutions.

5546

SUMMARY

This review of representative ID models may leave youunsure of how to react to.such a wide variety of models.The literature is replete with models, each chiming to beunique and deserving of attention. However, while thereare hundreds of models, there are only a few major distinc-

.

tions among them. Many of the models are simply restate-ments of earlier models by other authors using somewhatdifferent terminology. Also, there is a disturbingly smallvolume of literature describing any testing of the models.While no one can be certain, it appears that well over halfof the ID models have never actually been applied, nevermind rigorously evaluated. The typical publication contain-ing an ID model simply describes its major steps or stagesand perhaps how they are to be performed. The author(s)usually assumes the model is worthwhile, but presents noevidence to substantiate that position. In a few instances, acase study of an actual development project is presentedalong with the model, but even this low level of validationis relatively uncommon.It can only be hoped that in the future some ID models will

be subjected to rigorous scientific validation. Such valida-tion would require precise descriptions of the elements of

the model followed by systematic data collection concern-ing the application and impact of those elements. The inves-tigator would also need to be alert to possible discrepantdata not accounted for in the model. Repeated trials undersuch conditions would, if the model had any validity, resultin a precise set of findings regarding the conditions underwhich the model was valid. It is safe to say none of themodels currently available in the literature has been sub-

56 47

jected to such rigorous scrutiny. In fact, most authors com-pletely ignore the issue of what co:Iditions should bepresent if one plans to use their moOel. For a more completediscussion of procedures for validath g a model, the readeris referred to an excellent chapter on models and modelingby Rubenstein (1975) and the more recent work by Richey(1986).

What then, should be the response of the responsible IDprofessional to the plethora of unvalidated ID models? It issuggested that developers acquire a working knowledge ofa few models representing a variety of types of instruction-al development. Then, as new and different models areencountered, they can be compared to those with which oneis familiar. If a client brings a model to a developmentproject, it is probably better to use it (modified if required)rather than force the client to adopt your favorite model.Another suggestion is to have available in your repertoireexamples of models that can be presented with varyinglevels of detail. This will provide an easy introduction foruninformed clients that can later be made more detailed asdevelopment progresses. Also, when facing a range of situa-tions, developers should be in the position of selecting anappr.priate model rather than forcing the situation to fitthe model. As has been noted in other contexts, "If the onlytool you have is a hammer, you tend to treat everything likea nail." Developers should have a number of tools in theirtool bags and use the right tool for the right job.Looking back over trends in ID models for the last ten yearsin an attempt to forecast the future is probably doomed tofailure. As noted earlier, there has been little substantivechange in the corceptual framework of ID models that sug-gests any trend. While some recent models (e.g., Bergman &

Moore, 1990) are focused on new delivery systems, they donot represent any new conception of the ID process. Thus,the safe forecast based on the past would be that littlechange is likely in the new few years. However, this authorfeels that we are on the threshold of major changes inseveral fundamental concepts related to the instructionaldevelopment process. These changes will not result in dis-carding current concepts, but will significantly expand therange of possible approaches to developing instruction. Infact, one of the profound changes will be to move away

48 57

from the concept of education or training occurring in oneenvironment and performance in another. In the increas-ingly complex and everchanging world of the next decade,information will be too abundant and too transitory towarrant formal instruction. Embedded instruction, expertsystems to guide performance, microworlds, and increasedemphasis on learning how to learn and apply knowledgewill call for new design and development procedures verydifferent from those depicted by our current ID models. Torepeat, these new approaches will not replace existing pro-cedures, but will become alternatives to them. For a morecomplete description of these new developments see Briggs,Gustafson, and Tillman (1991) and also prepare yourself foran exciting and challenging decade ahead.

5849

REFERENCES

Andrews, D. & Goodson, L. (1980). A comparative analysisof models of instructional design. Journal of InstructionalDevelopment, 3(4): 2-16

Barson, John. (1967, June). Instructional systems development.A demonstration and evaluation project: Final report. EastLansing, MI: Michigan State University. 125pp. (ED 020673)

Bergman, R. & Moore, T. (1990). Managing interactivevideolmultimedia projects. Englewood Cliffs, NJ:Educational Technology Publications.

Berkowitz, M., & O'Neil, H. (1979). An annotated bibliog-raphy for instructional systems development. (ED 186 023)

Branson, R. (1975). Interservice procedures for instructionalsystems development (4 volumes plus executive summary).Tallahassee, FL: Florida State University, Center forEducational Technology. (National Technical InformationService, 5285 Port Royal Rd., Springfield, VA 22161.Document Nos. AD-A019 486 to AD-A019 490)

Branson, R. K., et al. (1975). Interservice procedures for instruc-tional systems development. Phase I: Analyze (318pp., ED122 018); Phase Li: Design (145pp., ED 122019); Phase III:Develop (380pp.,ED 122 020); Phase IV: Implement [and]Phase V: Control (209pp., ED 122 021); Executive summaryand model (157pp., ED 122 022)

Briggs, L., Gustafson, K., & Tillman, M. (1991). Instructionaldesign. Englewood Cliffs, NJ: Educational TechnologyPublications.

50

Diamond, R. M. (1989). Designing & improving courses andcurricula in higher education: A systematic approach. SanFrancisco, CA: Jossey-Bass.

Dick, W. & Carey, L. (1990). The systematic design of instruc-tion (3rd ed.). Glenview, IL: Scott, Foresman/Little,Brown Higher Education.

Dick, W. & Reiser, R. (1989). Planning effective instruction.Englewood Cliffs, NJ: Prentice-Hall.

Gerlach, V. S. & Ely, D. P. (1980). Teaching and media: Asystematic approach (second edition). Englewood Cliffs,NJ: Prentice-Hall.

Gilbert, T. (1978). Human competence: Engineering worthyperformance. New York: McGraw-Hill.

Gustafson, K. (1981). Survey of instructional developmentmodels. Syracuse, NY: ERIC Clearinghouse on Infor-mation Resources. (ED 211 097.)

Hamreus, D. (1968). The systems approach to instructionaldevelopment. In The contribution of behavioral science toinstructional technology. Monmouth, OR: Oregon StateSystem of Higher Education, Teaching Research Division.(ED 041 448; microfiche only.)

Heinich, R., Molenda, M., and Russell, J. (1989). Instruction-al media and the new technologies of instruction. New York:Macmillan.

Kemp, J. (1985). The instructional design process. New York:Harper & Row.

Leshin, C., Pollock J., & Reigeluth, C. (1990). Instructionaldesign: Strategies & tactics for improving learning and perfor-

mance. Englewood Cliffs, NJ: Educational TechnologyPublications.

Mager, R. & Pipe, P. (1984). Analyzing performance problems:

Or you really oughta wanna. Belmont, CA: Lake PublishingCo.

Merrill, M. D., & Goodman, 1. (1972). Selecting instructionalstrategies and media. E. Lansing, MI: National SpecialMedia Institute.

51

National Special Media Institute. (1971). What is an IDI?East Lansing, MI: Michigan State University.

Richey, R. (1986). The theoretical and conceptual bases ofinstructional design. New York: Kogan Page.

Rossett, Allison. (1987). Training needs assessment.Englewood Cliffs, NJ: Educational TechnologyPublications.

Rubenstein, M. (1975). Patterns of problem solving.Englewood Cliffs, NJ: Prentice-Hall.

Salisbury, D. F. (1990, February). General system theory andinstructional system design. Performance and Instruction,29 (2): 1-11.

Seels, B. & Glasgow, Z. (1990). Exercises in instructionaldesign. Columbus, OH: Merrill Publishing Co.

Silber, K. (Ed.). (1977). Educational technology definition andglossary of terms. Washington, DC: Association for Educa-tional Communications and Technology.

Silvern, L. C. Basic analysis. (1965). Los Angeles, CA:Education and Training Consultants Co.

Stamas, S. (1972). A descriptive study of a synthesized model,reporting its effectiveness, efficiency, and cognitive andaffective influence of the development process on a client.(Doctoral dissertation, Michigan State University). Disser-tation Abstracts International, 1973, 34, (UniversityMicrofilms No. 74-6139)

Twelker, Paul A. et al. (1972). The systematic development ofinstruction: An overview and basic guide to the literature.Stanford, CA: Stanford University, ERIC Clearinghouseon Educational Media and Technology. (ED 059 629)

Van Patten, J. (1989). What is instructional design? In K.Johnson & L. Foa (Eds.). Instructional design: New alterna-tives for effective education and training. New York:Macmillan.

6152

Annotated ERIC Bibliography

PeriodicalsAlbero-Andres, Magdalena. (1984). The

use of instructional development inthe television series on economiceducation: "Give and Take." Instruc-tional Science, 13 (2): 181-191 (EJ 305622).

Describes a case study which used the Ageiiy fuiInstructional Television ID model to develop atelevision series on economics. Findings show how themodel guided the series, its advantages, and itsstrengths and limitations. Also identified are the dis-crepancies between the ideal use of instructionaldevelopment and its actual application.

Andrews, Dee H. & Goodson, Ludwika A. (1980). A com-parative analysis of models of instructional design. Jour-nal of Instructional Development, 3 (4): 2-16. (EJ 228 351).

Examines 40 models of instructional design, identifyingwhich of 14 common tasks in model development eachincludes and categorizing them by origins, theoreticalunderpinnings, purposes and uses, and documentation.General purposes and uses of systematic instructionaldesign models are discussed, and an explanation for thevariety of moc:els is offered.

Bonfadini, John E. (1982). A competency-based instructionalmatrix. ManlSocietylTechnology, 41(8): 14-15. (EJ 262 819).

Presents a model for developing competency-basedinstruction via a matrix containing five conceptualdimensions: area of instruction, learning units, time

62 53

allocations, levels of mastery, and objectiveclassification.

Bonner, Jodi. (1982). Systematic lesson design for adultlearners. Journal of Instructional Development, 6 (1): 34-42.(EJ 274 764).Preseni.s a model for lesson design which accommodatesadult learning based on the Gagne and Briggs (1979)model of instructional design. Four stages of lessonlevel are discussed: (1) definition of performance objec-tives; (2) preparing lesson designs; (3) developing orselecting materials and media; and (4) assessing studentperformance.

Cantor, Jeffrey A. (1986). The strategic weapon system train-ing program. Journal of Educational Technology Systems, 14(3): 229-238. (EJ 334 755).

The U.S. Navy's Strategic Weapon System training pro-gram has five principal components: needs assess-ment/task analysis (Personnel Performance Profile);instructional design (Training Path System); instruction-al development (Curricula); training implementation;and personnel and training evaluation. This trainingmodel illustrates how a program constructed around anISD model can work.

Coldeway, Annabel E. & Coldeway, Dan 0. (1987). Anextension of PSI through the application of instructionalsystems design technology. Canadian Journal of Education-al Communication, 16 (4):279-293. (EJ 362 654).

Presents a model for course development which utilizesa behavioral approach to instruction within the contextof ISD. An overview of the ISD process is given, andan introductory undergraduate psychology courseprepared via modifying Keller's Personalized System ofInstruction (PSI) is described.

Davies, Ivor K. (1982). The CLER model in instructionaldevelopment. Viewpoints in Teaching and Learning, 58 (4):62-69. (EJ 273 607).

H.S. Bhola's CLER (configurations, linkages, environ-ments, resources) Model, when applied to instructionaldevelopment, can anchor instructional development in

54 63

real time and space, assist in focusing the variablesinvolved, help determine the range of choices, andassist with implementation and evaluation. It can helpdevelopers assess alternatives and choose among them.

Flouris, George. (1989). The use of an instructional designmodel for increasing computer effectiveness. EducationalTechnology, 29 (1): 14-21. (Ej 389 240).Presents an instructional design model for the develop-ment of software to increase the effectiveness of com-puter assisted instruction (CAI) in the classroom. Theimportance of basing the design of instructionalmaterials on valid learning theories is discussed.

Gallini, Joan K. & Fisk, Arthur D. (1986). An information-processing approach to instructional systems design.Educational Technology, 26 (4): 24-26. (EJ 336 231).

Proposes an instructional design model which bridgesthe gap between cognitive psychology and the field ofinstructional design.

Harmon, Paul. (1981). The conceptualization of instruction-al design. Performance and Instruction, 20 (5):'20-23. (EJ 248

928).

Presents a model used to help students of instructionaltechnology conceptualize the range of instructionaleesigns they might utilize and which could be used toclassify existing materials for easy reference.

Hooper, Simon, & Hannafin, Michael J. (1988). Learning theROPES of instructional design: Guidelines for emerginginteractive technologies. Educational Technology, 28 (7): 14-

18. (Ej 376 635).Guidelines presented for the design of instruction usinginteractive technologies are based on theory andresearch in learning, instruction, and/or media develop-ment. Also explained is a meta-model, ROPES, whichhighlights retrieval of information, orienting activities,presentation, encoding in the cognitive structure, andthe sequencing of lessons.

64 55

Keller, John M. (1979). Motivation and instructional design:A theoretical perspective. Journal of Instructional Develop-ment, 2 (3): 26-34. (EJ 222 085).

Reviews research and theories in the area of motiva-tion, and develops a model for a systematic approach todesigning motivating instruction.

McCombs, Barbara L. (1986). The instructional systemsdevelopment (ISD) model: A review of those factors criti-cal to its successful implementation. Educational Com-munication and Technology, 34 (2): 67-81. (EJ 349 601).

Reviews factors leading to the successful implementa-tion of the Instructional Systems Development (ISD)Model and derives implications for the design of an ISDusers training program. Military and civilian ISDmodels are discussed, and recommendations for furtherresearch and development are given.

Nelson, Wayne A. et al. (1988). The intellectual content ofinstructional design. Journal of Instructional Devpinpment,11 (1) :29-35. (EJ 380 496).Describes instructional design as a high-level thinkingprocess to provide more information on the way instruc-tional design is learned and actually practiced. Topicsdiscussed include cognitive psychology, design models,planning, schema theory, development of expertise,metacognition, problem solving, and implications fortraining and future research.

Noel, Kent L. & Hewlett, Brent. (1981). Plying your craft:Instructional development and the use of heuristics. Per-formance and Instruction, 20 (7) : 15-18. (EJ 252 524).

Examines an ISD model used by Bell Laboratories as anillustration of how heuristics can be brought to bearupon the design and development of instructionalmaterials.

Romiszowski, Alexander J. (1981). A new look at instruc-tional design. Part I. Learning: Restructuring one's con-cepts. British Journal of Educational Technology, 12 (1):19-48. (EJ 247 474).

Discusses the limitations of existing instructionaldesign models, including Gagnd's, and describes a new,

56 6

more complete model meant to improve upon those nowin use.

Silber, Kenneth H. (ed.). (1980). Symposium on ID models.journal of Instructional Development, 4 (2): 19-37. (El 240864).

Presents papers on four different ID models currently inuse in either a university or a business setting. Allphases of systematic development are covered, includ-ing project selection, production, implementation, per-formance analysis, constraints, and unusual featuresthat distinguish each model.

Taylor, Robin & Doughty, Philip L. (1988). Instructionaldevelopment models: Analysis at the task and subtasklevels. Journal of Instrudional Development, 11 (4): 19-28.(El 385 884).Describes, justifies, and illustrates a process for analyz-ing and comparing ID models at the task and subtasklevels. ID tasks and subtasks are identified, and variousinstructional design and instructional developmentmodels are compared as to both comprehensiveness andoperational level.

Wileman, Ralph E. & Gambil, Thomas G. (1983). Theneglected phase of instructional design. EducationalTechnology, 23 (.1): 25-32. (EJ 292 009).

This article dis(:usses three phases of instructional(1) defilition of instructional parameters; (2)

analysis and synthesis of instructional alternatives; and(3) implememation of the instructional plan. Also in-cludes a 15-step ID model, which at each step providesa descriptive name, a boxed example, and a rationaleand current state of the technology for that step.

Bass, Ronald K. & Dills, Charles R.(Eds.). (1984). Instructional develop-ment: The state of the art, II. Dubuque,

0 IA: Kendall Hunt Publishing Co. (ED298 888).

Designed to serve both as a textboikand as a reference source for instruc-

tional developers who wish to keep up with develop-

57

ments in the field, this book presents a broad view ofthe field of instructional development together within-depth descriptions of important new topics and thelatest developments in the old ones.

Berkowitz, Melissa & O'Neil, Harold F., Jr. (1979). Anannotated bibliography for instructional systems development.Alexandria, VA: Army Research Institute for theBehavioral and Social Sciences, 57pp. (ED 186 023).This annotatcd bibliography lists instructional develop-ment resources relevant to the Interservice Proceduresfor Instructional Systems Development Model (IPISD),a standardized model providing for the assessment oftraining needs; the design, development, and implemen-tation of instruction; and the assessment of instruction-al quality. Following a literature search, relevantdocuments are classified according to the 19-blockIPISD Model, and summaries identify documents onauthoring aids, procedures, or techniques. The purposeof each block in this model is defined, and documentsare listed alphabetically within blocks. A status sectionfor each block indicates the availability of authoringaids sufficient to guide an individual through allactivities specified by the block, as well as the avail-ability of relevant procedures and techniques that couldbe developed into authoring aids. Directions for futureresearch, based on the lack of authoring aids available,are identified.

Davies, Ivor. (1984). Instructional development: Themata,archetypes, paradigms, and models. In Instructionaldevelopment: The state of the art, II. Dubuque, IA: KendallHunt Publishing Co. (ED 298 890).This chapter discusses the foundations of instructionaldevelopment and analyzes the development of variousmodels, paradigms, archetypes, and themata used todescribe instructional development. Two key strands inthe literature of instructional developmentinstruction-al efficiency ("doing the right things") and instruction-al effectiveness ("doing the things right")aredescribed. Objective and subjective paradigms are thenconsidered, as well as three archetypes of instructionaldevelopment: the audiovisual, engineering, and problem

58 67

solving archetypes. In conclusion, it is argued thatinstructional development should not be viewed as aprocess, but as a set of criteria to which the tech-nologies of the field must contribute.

Dixon, Terry. (1984). A suggested model for development ofcomputer assisted instruction for higher education. 19pp. (ED250 982).A mu:lel for the development of computer-assistedinstruction (CAI) is presented for the college classroomteacher. The following common software design modelsthat have been helpful in developing CAI models arebriefly reviewed: composite or structural design model,Jackson model, META.step-wise refinement model, andhigher-order software model. Design principles thatencourage the development of good software are alsoexplained, i.e., modularity, abstraction, localization,and hiding. The CAI design model is based on theMETA step-wise refinement model and consists of fivephases: problem clarification, system design, blueprint-ing, CAI synthesis, and documentation development.The problem clarification phase is composed of threetasks: objective development, content research, andnarrative synthesis. The purpose of system design is tor..repare the instruction for computer coding. Blueprint-ing involves the development of a detailed descriptionof the CAI from frame to frame and function to func-tion. This phase involves two stages of development:frame development and frame design. CA1 synthesisrefers to the actual encoding of the computer, alongwith debugging and evaluation. Finally, documentationinvolves the development of manuals and technical data.

Gerlach, Vernon S., & Cooper, Mary E. (1985). A model forthe development of computer instructional specifications.39pp. (ED 270 097).The development of computer-based instruction will begreatly facilitated by the use of an effective and effi-cient design model. Five components of such a modelare: (1) the objective; (2) the content; (3) the questions;(4) the boundaries; and (5) the entry skills. A goodobjective describes either something observable thatlearners do, the conditions under which they do it, and

59

the standard of an acceptable performance; or an observ-able product that learners produce, the conditions underwhich they produce it, and the standards of an accept-able product. The content may take many farmsdefini-tion, description, generalization, principle, rules, andothers. Learner processes will vary according to the con-tent type, i.e., a rule-using task requires the learner tostate the rule and then to apply it to a previously unen-countered example of the class of problems for whichthe rule is intended. The questitals enable the student tointeract with the content and they test the student'smastery of the objective. Questions may be stated ininterrogative (Which one is green?) or imperative form(Write the sum of three numbers), and thre :-.. kinds offeedback can bc piiiiidcd: knowledge of results (KR),knowledge of correct :esults (KCR), or KCR with advan-cement to next problcm or question. The boundariesdefine the area the objective covers; one part of thatarea is the domain (lirmilus or display), the other is _therange (answers to the questions). Entry behaviors arestated for two reasons: (1) to ensure that nothing isomitted and that there are no overlaps between oldknowledge and new instruction, and (2) to weed outstudents who lack prerequisite skills. Examples areprovided for each of the five model components.

Hymel, Glenn IA. (1981). An instructional design model forguided mastery learning research and development efforts.Paper presented at the Annual Meeting of Mid-SouthEducational Research Association (Lexington, KY,November 11). (ED 222 527).

Proposed is a comprehensive instructional design modelwhich (1) suggests a systems-based approach to prepar-ing, implementing, and evaluating instruction at theprogram syllabus, course syllabus, and instructionalunit levels; (2) subsumes those various dimensions of amastery learning strategy which encompass precondi-tions, operating procedures, and anticipated conse-quence; (3) relates the generic activities ofinstructional design to those essential elements compris-ing a mastery learning strategy; (4) provides a broad-based context for interpreting completed and on-goingresearch and development efforts in mastery learning;

60 Aci

and (5) represents a multidimensional framework forguiding future basic and applied research as well as thesubsequent development efforts in mastery learning.Areas of congruence between generic activities ininstructional design and essential components in amastery learning strategy, as weil as additionalactivities mandated by any comprehensive instructionaldesign effort, have provided the basis for identifyingpossible variables of an independent, moderator, depend-ent, and intervening nature that might be considered infuture mastery learning research and developmentendeavors.

Keller, John M. (1983). Use of the ARCS model of motivation inteacher training. IDD&E working paper no. 10. Syracuse,NY: Syracuse University, School of Education, 11pp. (ED288 520).This paper presents the ARCS model, which is aresearch-based systematic design model that interfaceswith typical instructional design and developmentmodels to improve the motivational appeal of instruc-tional materials, of instructor behavior, and of the wayin which lessons (or modules) and courses are designed.The description of the model, which provides strategiesthat a course designer or teacher can use to makeinstruction responsive to the interests and needs oflearners, includes: (1) four major conditions defined bythe model that have to be met for people to become andremain motivated, i.e., Attention, Relevance, Con-fidence, and Satisfaction (ARCS); and (2) the threephases of using the model in an instructional develop-ment sequence, i.e., define, develop, and evaluate. Thepaper concludes with a discussion of two developmentaltests of the ARCS model, which involved its presenta-tion at teacher training workshops and evaluation of theresults of the participants' subsequent applik;ations ofthe model in their classrooms.

Montague, William E. & Wulfeck, Wallace H., II. (1982).Improving the quality of navy training: The role of R&D insupport of instructional systems design. Final report. SanDiego, CA: Navy Personnel Research and DevelopmentCenter, 25pp. (ED 243 472).

61

This report reviews the background of the ISD model(used to develop training for Naval personnel), iden-tifies problems in the ISD process and in its manage-ment and implementation, and recommends methods ofISD improvement. The ISD model is described as aprocess originally developed to remind instructionaldevelopment experts about steps needed to producequality instruction, but subsequently implemented tohelp content specialists (who are relatively inex-perienced in instructional design and development)build instruction. It is noted that ISD methods as usedby nonexperts are not successful because they lackdetailed procedural guidance. Instructional engineeringand management problems in implementing ISD are out-lined and three alternative solutions to these problemsare considered and rejected. Several recent researchefforts are then summarized, including the instructionalquality inventory (IQI), which provides qualityassurance methods for the ISD; the development ofguidelines for building more relevant criterion refer-enced tests; and the initial development of computerassisted training development. It is recommended thatthe Naval Education and Training Command (NAVED-TRACOM) develop: (1) systematic methods for monitor-ing ISD implementation and the performance of ISDpractitioners and managers; (2) training and profes-sional development programs for these persons; and (3)automated aids for ISD. A 36-item bibliography and areport distribution list are provided.

Reigeluth, Charles M. & Schwartz, Ellen. (1987). An instruc-tional theory for the design of computer-based simulations.IDD&E working paper no. 23. Syracuse, NY: SyracuseUniversity, School of Education, 26pp. (ED 289 470).A simulation is described in terms of its three majoraspects: the scenario, the underlying model, and theinstructional overlay. The major focus of this paper isthe instructional overlay as the component that servesto optimize learning and motivation. Functions ofsimulations are identified as the acquisition of content,the application of the content, and the assessment oflearning. Five simulation features that act as vehiclesfor achieving these functions are then discussed:

62 71

generality, example, practice, feedback, and help, Ageneral model for the design of computer-based simula-tions is presented which offers prescriptions for thedesign of the introduction, acquisition, application, andassessment stages of simulations, and for dealing withthe issue of control (system or learner). Variations onthe general model are then presented which are basedon the nature of the behavior (procedures, process prin-ciples, and causal principles); complexity of the con-tent; form of learner participation; form of changes(physical or non-physical); and motivational require-ments. In conclusion, it is noted that these prescriptionsare only a first step in an attempt to construct avalidated prescriptive theory for the design of computersimulations, and that considerable research and exten-sive field tests are needed to provide the informationnecessary for both confirmation and revision of thevarious aspects of the theory.

Smith, Patricia L. (1985). Supplantation versus generativemodels: Implications for designers of instructional text. Paperpresented at the Annual Convention of the Associationfor Educational Communications and Technology(Anaheim, CA, January 17-23). (ED 256 338).Two instructional design alternatives are described anddiscussed: (1) the supplantation model of Ausburn andAusburn (1978), where learning strategies are built intothe instructional materials, and (2) a generative designmodel, where strategies are "built" into the learner.These contrasting models are proposed as representingthe extremes of a continuum that illustrates the ratio ofthe amount of processing support provided by theinstruction to the amount of cognitive elaborationrequired of the learner. The relative advantages anddisadvantages of each model are examined, and a lineof research is suggested for investigating the conditionsunder which each model, or compromises between thetwo models, might be appropriate. This discussion isreiated to current philosophical questions in the field ofinstructional design. Examples are used from the designof instructional print.

7263

Stumpf, Mark R. (1987). An instructional design model fordeveloping a computer curriculum to increase employeeproductivity in a pharmaceutical company. Ed.D. practicum,Nova University, Florida. (ED 294 565).This report presents an ID model that was developed foruse by the End-Users Computing department of a largepharmaceutical company to develop effective, but notlengthy, microcomputer training seminars to train officeworkers and executives in the proper use of computersand thus increase their productivity. The fourteen stepsof the ID model are described, and appendices includean instructional design model flowchart; sample lessonmodules; instructor, observer, and student evaluationforms; and a course module design checklist.

Sullivan, Robert Francis. (1982). Toward an integrative open-systems model of instructional development in educational andnon-educational organizations. Ph.D. Dissertation, StateUniversity of New York at Buffalo. (ED 248 879).The open-systems model of instructicnal development(ID) proposed for use in both educational and non-educational organizations is based on an extensiveexamination of ID, the systems and communication/con-sulting literature, and the results of a nationwide surveyof 750 currently active ID professionals in both highereducation and business/industry regarding ID processbehaviors. The survey focused on 26 behaviors that maybe practiced in the process of developing instruction.Respondents indicated the percentage of time they cur-rently practice each behavior and the percentage of timethey feel each behavior should be practiced under"ideal circumstances." Analysis of the 411 usable sur-veys received (69%) revealed that, while there are iso-lated differences between education and businessrespondents, both subgroups follow the same basic IDprocess, and respondents as a group believe that all 26ID process behaviors should be practiced significantlymore often than they currently are. Further researchinto the ID process is suggested, including field testingand validation of the proposed ID model. A 22-pagebibliography and reference list, a glossary, the surveyinstrument, survey cover and follow up letters, and cor-relation matrices are included.

64 73

Tessmer, Martin & Jonassen, David H. (1988). A CBI modelfor the design of CAI software by teachers/non-progamers. In Proceedings of selected research paperspresented as the annual meeting of the Association for Educa-tional Communications and Technology (New Orleans,Louisiana, January 14-19). (ED 295 668).This paper describes a design model presented inworkbook form which is intended to facilitate CAIsoftware design by teachers who have no programingexperience. The seven-part instructional program usedto teach the CAI model includes: (1) determination ofthe objective of the lesson; (2) a task analysis of thelesson; (3) classifications of the learning outcomes ofthe lesson; (4) design of the readiness for learningstrategy; (5) design of the CAI instructional component;(6) conversion of workbook program design into screendisplays; and (7) a formative evaluation of the lessonvia screen displays.

Trimby, Madeline J. & Gentry, Caste Ile G. (1984). State of IDsystems approach models. In Instructional development:The state of the art, II. Dubuque, IA: Kendall Hunt Publish-ing Co. (ED 298 896).This chapter presents and analyzes current develop-ments in the evolution of the instructional systemsdevelopment model. First, basic definitions of generalsystems and instructional development model terms areprovided, and the differing uses of the terms "instruc-tional development," "instructional technology," and"educational technology" are considered. The idea of amodel and four different ways in which to use this termare then discussed, as well as current and classicinstructional development models. These instructionaldevelopment models are analyzed and compared, andcriteria for selecting a model for use in a specificproject are given. Finally, a process for developinginstructional development models is outlined, and areasof needed research are detailed.

7465

How to Order ERIC Documents and ArticlePhotocopies

DocumentsIndividual copies of most ERIC documents (ED numbers)are available in either microfiche or paper copy from theERIC Document Reproduction Service, 7420 Fullerton Road,Suite 110, Springfield, VA 22153-2852; some are availableonly in microfiche. Information needed for orderingincludes the ED number, the number of pages, the numberof copies wanted, the unit price, and the total unit cost.Sales tax should be included on orders from Maryland,Virginia, and Washington, DC.The prices of paper copy are based on units of 25 pages(and/or any fraction thereof) at the rate of $3.12 per unit.The prices for microfiche are based on the number ofmicrofiche for each document. The price for one to fivemicrofiche for a single document is $1.15 (up to 480 pages)plus $.25 for each additional microfiche (96 pages) for thatdocument.Shipping charges for microfiche via first class mail begin at$.29 for one to seven fiche and $.52 for 8-19 fiche; addpostage for an additional ounce for each additional ten oreleven microfiche up to a total of 80. Documents and ordersfor more than 80 microfiche are shipped via UPS in the con-tinental United States, and charges should not exceed $2.79for one pound (81-160 microfiche or 1-75 pages). Estimateone pound for each additional 170 microfiche or 75 pages ofpaper copy.For additional information about ordering call 1-800-443-3742 or 703-440-1400.

75 67

Journal ArficlesCopies of journal articles (EJ numbers) are not included inthe ERIC Microfiche Collection. Photocopies of articlesfrom many of the journals indexed by ERIC are availablefrom the UM1 Article Clearinghouse, 300 N. Zeeb Rd., AnnArbor, MI 48106. Information needed for ordering includesthe author, title of article, name of journal, volume, issuenumber, page numbers, date, and EJ number for eacharticle. The price is $10.75 for each article and must beprepaid. Additional copies of the same article are $2.25each. (Lower prices are available to deposit accountcustomers.) For additional information, call UMI at 1-800-521-0600 ext. 533 or 534.

68 76

Kent Gustafson is a Professor in the Departmentof Instructional Technology at the University ofGeorgia in Athens, Georgia. Former positionsinclude being a professor at Michigan StateUniversity, director of media services for aregional school district, and classroom teacher ofmathematics. He holds degrees from WorcesterState College, the Univerc:ty of Massachusetts,and Michigan State University. His current areasof interest include managing the instructionaldesign process, analyzing performanceproblems, and designing computer-based toolsto support the instructional design process. Dr.Gustafson has authored numerous publicationsincluding two books: Instructional Technologywith Fred Knirk, and the second edition ofInstructional Design with Murray Tillman and thelate Leslie Briggs. He is an active member ofseveral professional associations includingAECT, AERA, and NSPI. He is also a member ofthe Editorial Board of the Development sectionof a quarterly journal, Educational TechnologyResearch and Development.

Gary C. Powell has a BA in Psychobiology fromDrew University and an MA in Industrial/Organizational Psychology from the Universityof New Haven. He is currently a third yeardoctoral student in the Department ofInstructional Technology at the University ofGeorgia in Athens, Georgia. Professionally, Garyaspires to te an instructional designer workingin business and industry.

BEST COPY AVAILABLE

7 7