'UTHOR Laey, George F.; And Others r 'TITIE Learner ...

ED 125 543 .

DOCIJNBI1T RESIINB

. 003;627

'UTHOR Laey, George F.; And Others r.

'TITIE = Learner Control of Lesson Strategy:44. Model for-PLATO.

IV System Lessons. Tebhnical Report, NovAmber 1973 toJune 1975. Final 'Report..

. INSTITUTION Navy personnel Research and Development Center, San

'REPORT NOPUB DATENOTE.

DESCRIPTQRS

IDENTIFIERS'

Diego, Calif. .

NPRDC-T-R-76.-36Jun 76.

aitsRostage.

:4 *Computer Assisted'instruction; Computer_Graphics;Computer priented-Piograts; Educational Research;Individualized Instruction;'*InstrictionalImprovement; InstrActional Materials; IjastructionalSystems; Learniing.Tleoiies; *Loons of Control;Military.Training;.*Models; On Line Systems; *TimeFactors (Learning) '

4

Authoring; *Learifei Controlled.Instructioh;,Navy;o:PLATO IV;,Pragiammed_Logic for AittolWtip TeachingOperations

ABSTRACT . ,

A technique-:was developed to facilitate creatingcomputer-based instruction (CBI) with a Minimum of effort On the partof the-author/coder, and 'the feasibility of using a structure whichputs control of the lesson strategy into the hands of the_studentlearner was investigated. A model which relieve* tire author' of CBIstrategy selection was developed for the PLATO IV sySted and wastestedin twoinstructional situations: trainins students to use a

multimeter to measure resistance and current flow, and assisting Navymess management specialists in recipe conVersion. Student performanceusing the learner control mode was compared tolthet. of Other studentsusing the iegular CBI methOch Students using the learner control modeexperienced a savings.in'tiaining time and the use of an authoringaid was shown to save time over.ther methods of lesson preparation.

p

*********i******#***************4**********************************Documents acquired by ERIC include many inforpal unpublished

* materials not available from oilier sources. ERIC makes every effort *.* to obtain the best copy. available. Nevertheless, items of marginal, s* reproducibility are-often encountered and this affects the quality ** of the microfiche and h#dcopy,reproductpns ERIC makes available', *,

* via the ERIC Document RePrceduction Service, (EDRS). EDRS is not* responsible for the quality of the origiAal document. ReprodActions** supplied by EDRS are the best that can be made from the original. *

********4*******************4****************************************

NiRDC TR 76-36 .

, ,

,

-LEARNER. CONIRQL 07 -LESSON STRATEGY:

A MODEL FOR PLATO' ri4YSTEM-LissoNs4George F. LaheYMice M. Crawford .

.';10chard E.'Hurlock -

J ,

U SICIEPAATMENT OF HEALiH.. - .

EDUCATION WEL t0110E *4.141101MAL 1.16TITUTE 9F .

,EDUCATION

e

June 1476

fr

a

. .

THIS -DOCUMENT HAS '66E/I REPRO.1 OUCEO 'EXACTLY AS RECiIVED FROM

---- --TNE-P,EDSON OR ORGANIZATION ORIGIN.* ,.. .. . .ATING IT pOyas OF VIEW OR OPINIONS, l''SYATECr00 11101 'NECESSARILY riEPRE!.

"

SENT OEFIGIALMAT,IONM.INST INT COF ; e 4

E0vc4a loN,P4:4S T't ON ON Pot ICY .r.,

I 4-. .; ,

,R.evfeWed -by

'John Ford; '.Jr.

Approv64;hy.Jarneg 3. R,41011

Teahniaal,Director, . ,

4

;.4

At

lk I

.

. ,

Navy' Personnel .Research and 'D-Oio-lopmententer...SE11:1 .Dien, Callforrtia,,,423:52

\ 5

0

'W..,i?,,:.-.,'1. - ...

''.;r .A.- .1 .. , ..

I :::... Ft SWORD

Management Spec lists chool.

4...e, 1

.. . i p

The reSeai5Ch'anddevelopment destribedherain was performed underTechnical Development Plan.43-0 (Edueation and Training Devellpment),

,Work Unit 0-03,03A, (Experimental tliAlva41*qf PLATO IV Technology).

. This effort is part of an Advanced Researdr Projects Agency /Joint Services,

Training Technology Erograth,for development of computer-based training. -

. ..,technology. .

. -..-

''';. -The coopdration and assistande,of the,followingare gratefully acknow-

ledged,, Personnel athe Epic Electricity/Electronics School and Mess

Management Specialists Schodl, Service School Command, Naval Training Center

.(NTC),_Saa Diego, .provided students kpr the-test of the format described .

j herein. Mr._ P. H._Ellfott provided valuable feedbackfi5E-Ris experience

with a metric Conveirsion71esson Ms,--1cS.idericks and' Mr. L. 11-45eVet-aice

provided valuable data on tleir experience with_students from the Mhss

ik

Ms% Betty_Whitehilfassisted the authors'.

during student uns.-- .

JAMES J. 6LARKINCommanding Officet

ibt.

"`""

0 .

4

. r

3

- O

. ,

,

r

-Problem'

'4

SUMMARY

ti

%r

,.

.

Preparation of lesson materials for tomputer-based instruction :.

ii a time-consuming task reqpIring specia_skills. Authori mast,

usually develop fhe lesson nlieeNs:t cre-iie--Thiffda_oterall -strategy,

.2 -.....= eirerta_ for mastery of in ividual-IlUca segments, etc:

Once -may to reduce this,workigad and'uhke lessoa.gregaraeion less

. costly and time -consu4tiglionld-be to develop authoring -aids, such as

-an instructional strategy whiC-1-todld be applied universally:,Iiirespective-,of subject matter. This strategy miiiii,7-then be preprogrammed and packaged

__ as a"driver" available for general use. The driver would p'rovide the_....

branching, routing and record-keeping fUnctionvnecessarf-r6716s-gon

- gresenta.luu. ----7mtbors- PaTI-e-SEcias- could' ttiatioolidelZrZtemprepara-

tion of content compatible with.the nature of the driver.

,?

Objectives.- °

,

ti

The objectives Of this research were to develop a techniqiie.to facilitate

creating computerKbased instruction.(CBI) with a minimum of effort on the

pare of thp authorfcoder, and to investigate the feasibility of;

Ior this purpose, using a structure which puts.dontrol of lesson stra-

tegy into thehand6 of the student-learner.

Approach .

.`,The.eseg-atial-aspects of a learner control lessoriwere seen to be

*AprOposto development of a driver. The breakdow\of content into.

-tales, examples, and practice problems for. each objective, with provisions

for student selection of the basis of type of content, relieves the author

'of strategy selection. -A model using'this format was de,ieloped-fOr use

CR=the PLATO' IV system. The utilitY,of the model was then tested by pre-

paring instruction in two different application , use of a multimeter'

to measure resistance and current flow, and recipe conversion for mess

*managemenf specialists. These lessons were'then tested using students

for whom theqe materials were essential parts-of the r school instruction.

Their performance was obierva-and-Ebiiria-reff-ff-tair-.5 othey-gtamm-pstng -

tbeirregular modp,of instruction.'.

Results

TheutiIity.of the model was demonstrated by haViIig authors prepare t

lessons which made considerable demands on the model structure Two

of the authors-so involved had minimal exposure to the'sdeve went of

the model and to authoring on-the PLATO IV system. Stu tle 8,using

these instructional IessonS were compared to these using the %r'regular

modes of, instruction. Their test performance -d not dif sigii -

ficantly. Moreover, the students using the lla er c or modelof.

act-- 1

. .

.0/

vii

.1.0

_

.instructioft appeared to be highly motivated 'and generaly enthusiasticabout their experience. In the case where computercbased instructionwas compared to classroom instruction they experienced a very consider-able savings in training time.r

The time required to develop the lessons for mess mapiagement.% ,

specialists were found to be significantly less per hour of in-struction time than the time required by other authors prepar-ing CBI lessons of comparable difficulty., -

Conclusions

1

1

This project demonstrated,that the use of an authoring aid can indeed-save tithe over other methods of lesson paparation -rats ofcomputer-based instruction. The use of a 'universal" driver to controllesson strategy eliminated a major part of the.task of lesson.prepara-tion, offering a significant savings in both author training and lessonpreparation time.

if

5.

CONTENTS

..,,

-%% Page. .

-...

INTRODUCTION - ' ' I'

I.

..,14,,

.

Problem .. ,-.1

Objectilke1 .-

Background 14.% . .... ... . 1

PROCEDURERESULTS 4

LEARNER CONTROL MODEL. . . 49

`Ba-Sic_ Premises , , ..... c . . . . . . ... . . .

Format ... .

Functional Characteristics

94'

=

1

Access to 'Content c.1 .

Selection and Type of Content and Level of

Difficulty... 11

-=- Access of Special' Units. 16

Handling Student Input t . . 16

Instance Sequencing ,-. . . ,.. 14

Answer Processing .. '16 J

Overwrites . . . , 4.16

DISCUSSION - 4. .4 190

..,

CONCLUSIONS AND, RECOMMENDATIONS21 .4

REFERENCES. . : ... .1 .4 ... . ............. ,'. . . . 23 :it

DISTRIBUTION LIST e'25 '

,

FIGURESe

1.. Student,Terminal)

3. Lou-tilt-g4 ex -with Status Shown . . .

-3i-2-4---Con-t--en4 -ran-gement-1- . i- . . ., . . . ..... . . -----io-

12

'4. Table of- -Key Functions ag-Seen-with_a Medium Level of

.

Ditficult.Exampie on the. Screen ... ... . . ,-- -,.....! . 13

5. .iii-minal Keyboard ,

Iy. . ..... ,. 14___,

6. Typical' Objective page ,. 15

a. Typical Page of In stance File for Multimeter,Iesson 17

o

0

16:

9

.IMTRODUCTION

*.Problem

?reparation of Tesscm _for r computer-based instruction (CBI)

is a.time-clonsuraing_task requiring -special.' skills. Several elements are

involved. Authors must decide an a lesson strategy, prepare the subject

-material to fit the.strategy, develop mastery criteria,_ develop- the,

coding, debug t lesson from both a strategy and Content 'standpoint, try

it out on studen s, and finally revise the strategy and content based on

experience with he lesson. One way to reduce this workload and make lesson

t. preparation iesslicostly.and,time-consuming'Would be to develop authoring'aids,

such ab an instrtktional strategy which could be applied universally, ir-

respective of subject matter. This Strategy might then be preprogrammed and

packaged as a "dr er" available for general-uSeTh--&-driver mould--provzide_

the branching, rou in'g, and record- keeping functions neeessari to lesson

presentation. Aut rs of CBI lessons could.then concentrateon-preParation

of content compatib evith,the nature of the-driver.-.

.9

ObjectiveOS.

_Learner control, idh puts. lesson strategy in the `hands of the student,

is particularly aprqpo to the development of a driver and is of intense

as a method oiVinstrution. The major requirement of the driver

is to be responsiVe-to 6he student's choices. The student must select his

own instructional path7o -strategy and decide when he has achieved mastery.

The author's.responsibili y becomes one of preparing adequately instructive

content in-a form.compati le with the driver function.

The objectives'ofthis'research, then Were to develop a technique to

facilitate cresting cOmputebaseclinstruCtion (CBI) with a minimumLoLeffbrt-

on the part of the author/coder,-44-to inVe&tiigate.the-teasibiliEY of using

a structure which puts corktroL-of-lesson-S-tirategy into the hands of tt%student.

Background -f

-The hypothesis that' learners should. be given control of lesson strategy

traces its origins td experiments conducted by Hager and his associates in

.the-early 1960s (Mager, 1961; Mager & McCann, 19.61; Mager & Clark, 1963).

Their findings indicated that students given. complete control of lesson

.strategy-not only had a better learning experience but also completed train-

ing in a,shorter time han students Whose learning was controlled by teachers.

At about'the same time`; Evans, Glaser, and Homme (1960) presented researchers

th a conundrum which remains unsolved at this.writing: What should'be

p sented-first, rules or.examples?

Ga e' (1970) propoSed t-fit'lesson materials.be presented according to

rather gid hierarchical constructs. His theorems-require that authors

aialyze t e relationships of the lesson objectiveso asking-themselves whatf

,

the learner must know in order t chievethis new task given instructiononly. Further, his theorems quire materials within objectives,to be ,'.pr#sentedso that the student can first make associations, then discrimina-'dons, and finally recognize the rule(s) governing theconstruct-beingpresented. Gagne' ;agrees with other authors (e.g., Stolurow, 1955) thatlesson materials should be adaptive; that is; they should rive to thenext logical step as-soon as the learner has demonstrated mastery. How-ever, such advances Would presumably be controlled by the 'author on the

basis of criterion testing rather than bythe student.

4. Oae point that Gagne' makeS.quite strongly 'is .that learaing'depends upon ,

events in the learner's external environment: Since these events cannotnecessarily'be controlled by either the student or the teacher, the questhat arises is whether the student wh3 experienc s Ehem.or the_teacher,wh_Tresumes!--them-ii-better able to-assess them. -

, \Perhaps the most persuasive literature on'the subject of learner ,contro

da that prepared by Merrill and his associates (Merrill, 19734 Merrill &Boutweli, 1973)., Merrill(19735 stated that Strategy and content areAndependent,st,least to the extent that what is to be. taught is indepen-dent of-howit'is'to be taught. If, then,'oneyere to provide astandard-.izedistrategyor strategies which would have wide,'perhaps universalapplication, the task of the individual author could be reduced to de-velOping, encoding, debugging, and revising, content only, a considerablesaving over the full task. One such strategy is offered by Merrill.Giving control of lesson strategy to the student is an instructional devicewhich relieves the author of concera,for strategy, and permits developmentOf a lesson model in which tie author's sole concern is in developingthe approiltiate content.

The Merrill papers were based largely on theoretical constructs, whichformed the. theoretical basis for the development of the Time-Shared Inter-,active Computer Controlled Instructional Television (TICCIT) system. Thissystem is being used operationally at Northern Virginia Community College,Jhoenix,Jdniortollege, North Island.NavalAir Station at San-Diego; andErigham Yourd-UniversitY. -

. ,

1,,.

In a comprehensive review of previous research on learner control, Judd,O'Neil, and Spelt (1974) ,raised,serious questions about. the methods usedand the influence of, uncontrolled variabies.I They stated fdrther that learnercontrol can mean very) different things to different people. Even the TICCITsystem puts many, restrictions on the learner, such as prerequisite lessonobjectives and prerequisite criteria /or content selection. HoweVer, inspite of their objections, there appears to be a wealth of research indicat-ing some advantage for learner control,,-,at least from a motivational, if nota performance, standpoint.'

Research on learner control originated.at the NaVal Personnel and Training,Research Laboratory OPTRL), San Diego in the early 1960s. (NPTRL and theNaval' Personnel Research'and Develdpment Laboratory, Washington, were merged

2

in 1973 to form the y Personnel Research and Development.Centex).

Initial efforts were di ected at adaptive as opposed to linear programming,

but adaptive only to the\ xtent that students were branched past linearly

programmed materials as so as they demonstrated mastery. Later research

(Slough, Ellis, & Lahey, 19 ; ;Ilurlock, 1972; McCann, Lahey, & Hurlock,

1973) gave students more cont 1 reflecting increased awaipness of the Iadvantage of student participat n in selection of materials.

4

, 3

94

411

PROCEDURE

Late in 1973,'a beginning was made on a lea'rner control model along

lines largely consistent with Merrill's propositions. The following

prescriptions were adhered to in developing this model:

'V 1. The

'selecti the

I

tude t should havg complete control of theorder in which he

ess n1 segments or objectives.

2. Having selected an objective to study, the'student should have

complete. freedom of choice of type oftcontent (i.e., rules, examples, and

practice problems) for that oBjectIve.

3. HaVing selected a type of content, the student should have complete

freedom-of choice as to the level of difficulty (easy, medium, or hard) of

the materials presented.

4. The'final lesson segment, the lesson quiz, should be available as

an,option. That is, the student should be able to take it at any time,

omit it if he chooses, or repeat it if desired..

skeleton strategy (driver) and content sections were developed for'

delivering lessons on the PLATO IV System. This system; an interactive,

'general-purpose system as described by Bitter & JohnSon "(1971), is flexible

and fairly easily programmed, offering authors the opportunity to create

materials with few prerequisite constraints. A photo of a PLATO IV terminal

is preiented as Figure 1. Using a dummy lesson, the paradigm was examined '

for its utility and feasibility.

10

5

. 5

a

0

5

- .

Figitre 1. Student terminal..

ti

9

,

iI

.

gESULAIII

.. \[3. .

The original model'clearly indicated that preparing separate strategyand content sections for a CBI lesson wNeleasible for the PLATO IV:Sy0teM.The strategy section was reasonably economical, in terms of-hoth'the amount .

of computer storage required and the rapid, cogent response to learner selec-tion of content.

_,

. .

r ----.....,,,,....

Durj.ng 7114, the model was used to prepgre lessons-in two .fields: U.)

using a multimeter to measure electrical resistance and current flow, nd

.(2) using the mathematics of converting standardized Navy recipes to it

messes of widely differdht sizes. The multimeter lessons became the primary

, development vehicle and exposed several weaknesses in the original model/ which required changes in the driverand content formats. The recipe conversion

-lessons, which were developed later by authors who had had no previousexperieoce with the model, also exposed some weaknesses. Thus; during the

course of lesson developMent, the model Underwent signifioant'changes, although

the basic paradigm remained the same.

& 4 Initial tryoutsdndicated 'that students did not appreciate or understand

the options availablL Thus, revisions were made to the introduction and to

the form of the presentations used. Even during the'research effort whichfollowed, when many students (74 multimeter students, 40 recipe conversionstudents) were using the lessons, some shortcomings in the driver were evidentand minor adjustments were required. However, results of these efforts, as.reported by Lahey, Crawford, & Hurlock, 1975; and-Fredericks, & Hoover-Rice,

in press, indicate that authors are able to develop meaningful lessommaterials.

in the form required by the model. Further,, for both types of subject matter,learner control of lesson strategy proved to be an effective mode of Computer-

based instruction...

. ,.

. In the multiZeter study, there was.no significant difference in performance

on laboratory tests and module examinati,ons between students taking the Multi-.:

meter, lessons via CBI and others using regular individualized,instructiontechniques. The CBI students took longer to finish the modules than theregular students, partly because of the time required to get acquainted with

the new training mode, and partly dug to.±ncreased opportunity for (simulated),

practice. -'e'-

7 i,

In the_reape conversion study, experimental (CBI) students with 4 hours

of instruction e'bored as well on final examinations as controls who received

28 hours orclaseroom inst ion. This represented a highly significantt ----.-savings in training time. is also important to-mite that this test...

.

population included a large proportion of average aptitude trainees. As

to the author requirements for lesson,preparation, Fredericks and Hoover -Rice

were able to prepare their lesson materials with a minimum of gladance fromthe model developers, and to complete their lesson materials in less time

)than was required by authors using more conventional techniques (77 hoursper hour of instruation=asopposed to.315 hours per-hour):'

r 124 -7

S

J

t Si

tr.

, -The final versi.on of the rodel, which was' developed upon the conclusion

of these. research efforts in January of1975, is de-scribed in the followingsection. -

,......

, .^t

..:"$

"

Jd

v

a

1,41INER CONTROL MODEL

'Basic Premises.

The major premise usain developing, the learner control_modelWasthat lesson content and lesson strategy are independent functians; so thata common instructional' strategy can be used to present CBI lesson Materials

for a wide, perhaps limitlessi, variety of subject matters. This commoninstructional strategy,'which would inclUde all control and record-keepingfunctions, would allow the author to concentrate on the development ofinstructional pontent,-

Minor premises were:

1. Once the lesson content had been prepared, encoding the lessonshould not require intimate know/edge of 'he computer language but onlythe basic ability toledit cdde, in this i Lance the TUTOR language.

. .

2. Authors using the model for development of their own lessons shouldlearn the operations required (excepting, of 'con*, how to editfthe TUTORlanguage) from content within the model rather than having to look to an

external source. - 14

Format

The strategy section of the lesson --is segregated within three blocksof the model, which ate identified as the "driver.',', All Units essential,to controlling lesson strategies are contained in these three blocks,cluding those for the basic branching and record-keeping functions, and foraccessing units of the content gentian: sTo accomplish this latter function,'the basic units within the content section have'specified labels.

Lesson materials consistof three types of content (rules, examples,tandpractice problems) at riirbe levels of difficulty (easy, medium, and hard);

seeFigure The medium level of diffiPulty is aimed at the,"average"-student; easy, at those having trouble with meditii level materials; and hard,at those ready foran approach involving the use of technical terms. The

hardItvel is frequently Omitted. e

The model format calls*.for rules to be,subsuted within single .units '

for each level of difficulty.. Thus, Unit."ezru' contains ,the easy-to-read

rules; Unit "medru," the mediui rules; and Unit-"hdru," the rules involving

techniCal.terms.. On .the other hand, becauseeaciilexamPle set may include

.many examplesi. examples for each level of diffieuity are subsumgd within

400a.8ingle unit. Thus, the easy examples' for Objective 1 are 'Subsumed

, within Unit "ezinl, "; examples, within Unit !'iml.,".; and hard examples,

'within Unit "hdinl."

411

44

14"

9

An introduction used to indoctrinate students occupies the initial

blocks of the model lesson. It tells the student what to expect and how.

to use the model and is as instructive to new authors as it is to students.

In ,instructional lessons, the introduction should not be physically located

in'the author's bases lesson. If will usually be put elsewhere, followed

by ,a jumpout to the base. Leeson.

The last part of the model contains special units used to demonstate

special features and to display the paths students take through their lessons.

Most lessons will be similarly constructed, in that authors will add special

units to provide graphics and other illustrative ptesentations to supplement

the text contained in the basic content units, and to see what students

actually do in using their lessons.'. ,

Functional Characteristics

Access to Content

Access to content is Controlled by two key units which direct the

.student's responses and accept his chdice,of materials. One unit lets him

choose an objective from-ad index of the subject matter, (Figure 3) the

second shows'whilgh keyto press for the various types of content.

The student accesses the various types of content via touch panel or

keypress. A table.of key functi ?ns (preiented at the bottom of the plasma-:

screen) is, displayed at all times except when the student is referring to

s he index page or an ing a problem. As shown ,in Figure 4, it offers -.7.

%the student his, choice f rule, example, or.ptracice; easy, medium, or hard' .

versions of each; and elps and superhelps. The gtpdent can either press

the key indicted-*(see keyboard, Figure 5), on-touch the block on the Screen

with identifies the key function. Both'options are available at all times

during the lesson. , .

Selection of Type of Content and Level of Difficulty

After the student has selected the subject he wants to study from the

index page, he is'branched tg an expanded statement of the learning objective..

When properly worded, the statement is a behavioral description of what he

will be able to do upon completing the study of that lesson segment. From

the objective "page," the student chooses, a content' category; i.e., rule,

example,,orpractice (see Figure 6).

with

selecting Ehe,desired category.of

-content', the student is presented with an item at the medium-Avel of dff- "&;

ficulty He can then change the level of difficulty. to easy or hard or back .

to medium as he desires. .

'1511

t

...AcCesi of Special Units

Special units are used to ekpand the capabilities of the model.'Whereas the basic content units have .specified labels and a -Awritec-format, special units may have any format and label the author desires.The branching, commands required to access special units are placed within

-content units. The driver handles their accession by accessing the conteniunit. Using special -units, authors conversant with TUTOR coding can, createlessons to fit almOst any'instrUctional requirement,

Instance Sequencing -.

*

Since examples and practice problems differ only as to their expository/inquisitory nature, the facts pertaining to a_-particular instance may oftenbe used,in either form. The model therefore calls for a file of instancestoserve both as examples and practice prOblensA Access to individual 'instancesis controlled by indexing a counter. All files recycle (returntothe first instance) after presenting the last instance. --

:'. !; __.

Answer Processing .

, .. jI,.,

.

A five-part, multiple- choice format has been chosen ad-a standard format

for practice problems. In keWng with the rest of the model; it uses a-writec- command to present the multiple- choice' response alternatives. ,Unitsin the ,driver scramble the presentation of alternatives.;prepared by the authorso that the correct answer, inserted as the first altehativeby the author,'may appear on the screen-in,alay one-of'five alternative positions. The

driver units identify theeCorrect ansWei, 'so that appropriate fdedback-canbe,given to, the student when he responds. In the standard format,,leedback

Consists of either, "No. Try again'.," Or,"Very good!" -

. Authors who wish to use other than the five-paftmultiplelchoice format/

snbstitute their own_sliecial units for the standard units: Figure 7 shows

a c6nstrucied response practice problem used in'thelultimetet lesapn,,with'apptopriate feedback for the response indicated,' , . - , 3

Z1 = * .

,

Overwrites

Screen iodations have been dedicated to alleviate th Aroblem ef"oVet,-

writes. As for other portions of the model, authors who:arewilling t6make their,own provisions to prevent Overwrites Call modifythe space pto-7

to fit their materials. .Such modifications need"nOt affect thexdriverfunctions.

.

4. 1

16

16'

-;

4

0

. DISCUSSIONA

:The-constructs used to develop the model are derived from the premises

.stated by Merrill (1973). The model is' designed for developing lessons,

in which learner control of lesson strategy is accepted as an advantageous

mode of instruction for both the learner and the author of the instructional

lesson. Individuals desiring to develop learnpr control lessons on the

PLATO IV system may copy the model into their own lesgon spaces and then

insert their own lesson materials.

Hilie paradigm used, to develop this particular model has much in common

-.with that beingused to develop lessons for the TICCIT systdia. Since the

model is_not hardware limited, it could have been developed for the IBM

1500 system as easily as for the PLATO IV system.'' Further, 'it doesn't

depend heavily on system Configuration or software. Thus, the basic lessaR

structure could easily be applied on any number of hardware Systems., .

1,es n "learner ", the learner control model, makes no special deuhnds

e capabilities of the PLATO IV system. It uses the hardware.and scift-Q

re. capabilities available. The courseware which makes,up a learner control

lesson 'requires no special expertise. However, PLATO users familiar

with the problems of adequate extended core storage may encounter one

problem: -The driver(three blocks) must be attached to all parts of

the lesson,so that, lessons requiring several lesson spaces may require

extra, extended core storage. There are no data with which to eV4Uste_

the net effect. --It may be less-or greater than. 'lessons.notuSing-learner

control, but the. problem is worth,tentpnidg. On systems where the driver

could. be developed as a subroutine, the problem would not exist. ,

The development,of rules, examples,\ and practice'ProbletS for learner

Control lessons witll be a function of the type of material being presented.

Some-lesson mafrials will require little more than teXt,.andihence, d

minimi departure from the basic lesson. format. -Others will-require fiord.

The flexibility to accommodate diverse materials is very MOdh.,:a part of

the "learner" concept.;' Nothing in the driver or content "sections of:the model

precludes the developMent ofl'extremelyiy displays, constructed

response problets,IRr otter,complex instructional modes. What an author

does is. entirely up,to him. Same indicationOf.the,fletibility of The model 'isset forth the lesson "itself in the form .of "Stand-ihitCatitett. AdditiOnil

data may be obtained by examining one of the muitimetdeOr'rec;Pe::::600.VerSi*;::-:

lessons dev loped for :tie PLATO IV system.. Both, the timeteX "arid recipe =conversion 1 ss depended on graphicti forcanfent OtesentatiOn;-patticularlyr.

the former.--in .texr:

u

eloping- these materials was a, good bteti beyond simply replaor

-

CONCLUSIONS AND RECOMMENDATIONS.

On the basis of NA*USRANDCEN experience with the muttimeter andrecipe conversion lessons,. this model is recommended as a way to createCBI lessons. It Sou1d not only reduce CBI development time and.providemeaningful instruction to students, but also'Play,some part in bringingthe economics of CBI lesson development within reasonable bound. Thetwo adv.antages ofthe model are that it (1) provides a vehicle fOr learnercontrol of lessOn strategies and (2) frees the authors of the necessityfor strategy development and testing. There are 1r) decision points to beconsidered, no mastery criteria to be developed.%These become theresponsibility of the student, and interestingly enough, he usually makesrational if not optimal decisions.

Although the concepts of developing CBI lesson in which the strategycand content sections are completely independent providing learner

control both have been successfully demonstrated, 'need for additionalresearch is indicated.* NAVPERSRANDCEN currently, is using the multimeterlessons as a vehicle for evaltating the effectiveness of learner control.The learner control mode will be compared to, progkammed control' by ,stib

stituting.a normative programmed driver for the learner control driver inone set of dultimeter lessofi materials. At the same time, another comparisonwill test-the effect of telling the student what he should do via an "advisor"which uses the same normative driver used for programmed control,as-a' cueingdevice. -The "advisor" will not affect the student's freedom to do what he.Wants..

211

REFERENCES ,

Bitzer, D. L., & Johnson, R. L. PLATO: A computer-based system usedin the engineering of education. Proceedings of t e Institute ofElictrical and Electronics Engineers, 5; 960, 1971

Evans, J. H., Gla5er, R., & Homme, L.E. The RULEG system for theconstruction of programmed verbal learning sequen4es. Pittsburgh:

University of Pittsburgh, 1960.,___

Fredericks, 7. S., & Hoover-Rice, L. B. An evaluat on of.learner controlCAI and traditional instruction-classroom techniques for reciEe conversion.San Diego: Navy.Personnel Research and Development Center. (In preparation.)

Gagne', R. IL The conditions of learning. New Yolk: Holt, Rinehart'nidWinston, Inc. 1970 (2nd edition).

,

Hurlock, R. E. Applications of pretest branchins designs to CAI basicelectronics training. .San Diego: Naval Personnel and Training ResearchLaboratory, September 1972. (Research Report SRR 73-8)

Judd, W: A., O'Neil, H. H., Jr., & Spelt, P. F. Individual differences-andlearner control: Program development and investigation of ,control over

mnemonics In computer-assisted Instruction. Air Force Huddn Resources

Laboratory, July 1974.' '(Technical Report AFBRL-TR74-30)4

Lahey,-G. F., Crawford, A. M. & Hurlock, R. E. Use of an interactivegeneral-purpose computer terminal to simulate training e ment operation.

San Diego: Navy Personnel Research and Development Cente, ovember 1975..

(TechnicalReport TR 76-19)

Mager, R. P...On the sequencing of instructional content. Psychbl

Reports, 9; 1961.

Mager, R. F., & Clark, C. Exploration in student controlled instructioA.''Psychological Reports, 13, 1963. ,

Mager, R. F., &McCann, I.44Learar controlled instruction: Palo Alto CA:'

Vafian Associates, 1961.

I .

McCann, P. H., Lahey, G. F., & Hurlock, R. E. A comparison of Student option

versus Erogram controlled CAI training. San Diego: Navya Personnel and-

Training Research Laboratory, April 1973. (Research Repprt SRR 73-17)

Merrill, M. D. Pr emises, Propositions -, and research underlxing the design

of a learner controlled computer-assisted instruction systei: A summary

for the TICCIT system. Provo, Utah: Brigham Young University,' Division

of Evan Institutional Services, June 1973, (Working Pa er No. 44).

..

'1923

I ..._qierrill, M. 1):,? & Boutwell, R. C. Instructional development: Methodologyand teseard'h..\' In Berlinger, F. N. (Ed.) Review of research in education.Itas a, Illinois: Y.-E. Peacock, 1973. e.-

,

Slough D. A., Ellis, Bs D., & Lahey, G. F. Fixed sequence and multiple,bran.hin strate\ es in co n. uter assisted instruction. San Diego: NavalPers . nel and Traiksing Research Laboratory, September 1972. (ResearchRepor SRR 73-6) ,

\\ .

StolUroUnivenical

, L. 111, System's, apProach to instruction. Urbana, Illinois: -

sity of Illinois;,Training Research Laboratory, July 1965. (Tech-eport No. 7)

41b

2,0

24