DOCUMENT RESUME Student Feedback as a Tool in …Education; v7 n2 p221-302 1970. EDRS PRICE EDRS...

DOCUMENT RESUME

ED 047 497 EM 008 697

AUTHOR Meredith, Joseph C.; Ferguson, DouglasTITLE Student Feedback as a Tool in Computer-Assisted

Instruction (CAI) Frame Development.INSTITUTION Indiana Univ., Bloomington. Research Center for

Library and Information Science.; Stanford Univ.,Calif. Libraries.

PUB DATE 70NOTE 83p.JOURNAL CIT International Journal of Experimental Research in

Education; v7 n2 p221-302 1970

EDRS PRICE EDRS Price MF-$0.65 HC-$3.29DESCRIPTORS *Computer Assisted Instruction, *Feedback, Programed

Instruction, Programed Units

ABSTRACTIn order to measure the value of student feedback in

contributing to the refinement of instructional frames incomputer-assisted instruction (CAI), a stand-alone CAI frame waswritten around a question designed to elicit a wide variety ofreplies. Half of a real sample of fifty handwritten responses weresubmitted to the frame, with a match score of 96% and a sense scoreof 48%. After revision, the frame attained a match score of 100% anda sense score of 88%, on the other half of the test set. Conclusionsdrawn frcm the study were that (1) for an open question with a largenumber of legal answers and an adult, heterogeneous student set, astriking improvement in meaningful intercept can be obtained throughclose analysis of a modest number of live responses; and (2) as acorollary, the rate of improvement to be derived from close analysisbeyond this mcdest number must decline quite rapidly. In the reportof the study, selected-string-match CAI strategies and the mechanicsof the CAI "macrcframe" are discussed. Appendices contain supportingmaterials. (Authcr/MF)

O

iSCIENTIA PAEDAGOGICA

EXPERIMENTALISINTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCHIN EDUCATION REVUE INTERNATIONALE DE PEDAGOGIEEXPERIMENTALE INTERNATIONALE ZEITSCHRIFT FOREXPERIMENTALE PADAGOGIK INTERNATIONAAL TI JD-

SCHRIFT VOOR EXPERIMENTELE PEDAGOGIEK

EDITED BY - EMMA PARHERAUSGEGEBEN VON - UITGEGEVEN DOOR

R. VERBIST / Gent

WITH THE CO-OPERATION OF - AVEC LE CONCOURS DE

UNTER MITWIRKUNG VON - MET DE MEDEWERKING VAN

A. CLAUSSE / Liege E. FELDMANN / BonnJ. T. HUNT / Athens, Georgia (U.S.A.)

VII, 2

UITGEGEVEN MET DE STEUN VANHET BELGISCHE MINISTERIE VAN NATIONALE OPVOEDING,

EN VAN DE UNIVERSITAIRE STICHTING VAN BELGIE

GENT, HENRI-DUNANTLAAN, 1197o

U.S. DEPARTMENT OF HEALTH, EDUCATIONIS WELFARE

OFFICE OF EDUCATIONTHIS DOCUMENT HAS BEEN REPRODUCEDEXACTLY AS RECEIVED FROM THE PERSON ORORGANIZATION ORIGINATING IT. POUTS OFVIEW OR OPINIONS STATED DO NOT NECESSARILY REPRESENT OFFICIAL OFFICE OF EDUCATION POSITION OR POLICY

STUDENT FEEDBACK AS A TOOL INCOMPUTER-ASSISTED INSTRUCTION (CAI)

FRAME DEVELOPMENTby

JOSEPH C. MEREDITH (Research Center for Library and Infor-mation Science, Graduate Library School, Indiana University)and DOUGLAS FERGUSON (Editor, Automation Department, Stan-

ford University Libraries)

BACKGROUND AND OBJECTIVE

This paper represents an attempt to measure the value ofstudent feedback in contributing to the refinement of instructionalframes in Computer-Assisted Instruction (CAI). In our ownwork in CAI for graduate students (specifically for students oflibrary science) (1) we have concentrated on the development ofa « machine tutorial mode » suitable for presentation of con-ceptual material to an adult and very mixed student population.We must keep in mind students whose sole preparation is aliberal arts degree others who hold degrees in the sciencesand still others with abundant experience as practicing librarians.Primary interests range from children's libraries to the upperreaches of information theory. Ages range quite evenly throughthe twenties, thirties, and forties.

In order to be effective on an individual basis, we have. feltthat we must make the machine side of the program as versatileas possible, with a very broad capability for dealing with idio-syncratic responses. Whatever we do will fall Far short of humandiscourse; we cannot pretend to cope with meanings which havenot been anticipated; but we can at least take advantage ofobserved patterns of speech and thought in order to avoid moreformal constraints on dialogue than are absolutely necessary.

(1) Described in the interim report on Project No. 7-1085 (0EG-1-7-071085.4z86) : An Information processing laboratory for education and research inlibrary science : Phase I (M. E. MAaoN, A. 3. FIUMPHREY, and J. C. MEREDITH,DHEW (OE) July, 1969).

222 J. C. MEREDITH & D, FF,R6DSON

What about these patterns ? It is sometimes said that pre-planned conversational CAI quickly reaches an area of diminishedreturn, because of the massive intellectual and clerical effortrequired to prepare even a short sequence (r). This might betrue if the CAI author were to work in a vacuum, drawing onthe single resource of his own semantic store. But what if he wereto demand less of this for-the-most-part intuitive, remembered,and general resource and rely more on live results derived frompreliminary testing ? We know that student feedback is usefulin revising programmed learning, but we need to know moreabout the extent of this usefulness and to decide when and howbest to exploit it.

This, quite simply, was the idea behind what we thought wouldbe a straightforward exercise in recursive improvement of asingle CAI frame. If we had had an earlier indication of howstriking the results would be, we would have structured the testmore rigorously. Even so, we feel that the controls are adequateto justify reasonable confidence. Also, we did take the precautionof drafting and documenting Part I of this paper before proceedingwith the second part of the experiment, which is described inPart II.

PART I

PROCEDURE

Fifty individuals were polled in writing (Appendix A) with arequest that they indicate what single response they might havegiven to a certain statement/question, in an on-line situation ata student terminal. Individuals were chosen at random from amonglibrary school students, faculty, and technical and clerical staff ofthe Institute of Library search, University of California, Ber-keley, California. In using subjects from this environment itwas hoped that we could compensate in part for the lack of

(t) William R. UTTAL et al. represent this view in Generative ComputerAssisted Instruction (Communication #2,43 of the Mental Health ResearchInstitute, University of Michigan, January, 1969).

3

STUDENT FEEDBACK IN CAI 223

instructional preparation leading up to the statement /question.At the same time we realized that a residuum of this lack ofcontext would need to be accommodated within the frame itself,to an uncommon degree. « Success », as we viewed it, would bemeasured as much by the effectiveness of this accommodationas by the efficiency of the match and fail processes of the system.

Respondents were asked to write down their first response tothe question as worded. No explanation or comment was providedat the time. The fifty answers were serially divided into sets oftwenty-five each to make up an « A » list and a « B » list, and theformer was submitted to a previously-written program calledthe « A-Frame ». The results are described and analyzed in thesections which follow.

We then revised the A-Frame in the light of what we hadlearned, and tested the effectiveness of this version (the « B-Frame ») against the « B » list of anwers. Appendices A throughI document the steps which were taken before conducting thissecond test.

THE PROGRAM

In designing the A-Frame we assumed programming languagecapabilities corresponding with PILOT (1), the CAI language inuse in the Institute's Information Processing Laboratory, withminor technical augmentations provided in DISCUS (2). Bothlanguages employ selected character string match as the basicanswer-matching routine. That is, student input of any lengthwill be successfully matched in a scanning operation if the elementsspecified in the scan are present. The remainder of the input isignored.

(i) See R. ICAapu4sia et al., PILOT; a Conversational Language User'sGuide (Office of Information Systems, University of California Medical Center,San Francisco, California, December, 1968).

(a) Developed by Steven S. Silver, ILR staff member, and now implementedat both the Berkeley and Los Angeles campuses of the University of California.Documentation available from the Institute of Library Research, University ofCalifornia at Los Angeles, California. User's Manual in Press.

224 J. C. MEREDITH & D. FERGUSON

A successful match between a scan element and an input element(student answer) triggers a unique set of program =actions.A match does not signify rightness or wrongness of a response.It refers only to the detection of a sought element.

In practice, the reaction provided for a response which iscorrect from the teacher's viewpoint will consist of displaying tothe student a statement to that effect, then making a record of thefact for scoring and control purposes and then exiting tothe next frame. The reaction provided for incorrect matchedresponses may consist of a critical or cueing display followed by areturn to the question. Or it may be followed by a critical statementincorporating the correct answer and then exiting to the nextframe. The same scoring and control tallies can accompany thisoperation.

Lastly, the program reaction on failure to match any of the inputis usually the display of a succession of two or three cueing re-sponses ending with one which gives the correct answer. Thestudent is then passed to the next frame, or in rate cases he maybe advised to sign off and seek the help of an instructor. Fail-match tallies are also recorded for control and scoring purposes.Naturally, dependence on fail routines must be kept to a minimumif meaningful instructional dialogue is to be attained.

THE FRAME

The frame, as such, is primarily an instructional concept. Itmay be programmed in a variety of ways, but usually it is com-posed of a question or other requirement displayed to the student(and usually preceeded by a didactic statement of some kind),followed by a computer-waiting condition which allows the stu-dent to insert his response, which is in turn followed by a pro-gram reaction to that response. The program-reactive displaymay be subjoined by text which is really the beginning of thenext frame, but the transition is not always apparent to the stu-dent : only the author/instructor needs to be aware of frameboundaries, in order to assemble his planned dialogue properly.

For the purpose of this study, we envisaged a single framewith five « satellite » frames. Each of the satellites would be used


to supply an element of information requested by the student.,and to ask him if the information sufficed. If so, he would bereturned to the main frame; if not, he could elect either to passto the next frame or to sign off. We call the whole complex ofmain frame and satellites a « macro-frame ».

By calling up the satellites one by one, it is possible to provideindependent subroutines which look like frames, i.e., each has aquestion, a wait mode, and a reaction, but which would never beused as instructional units by themselves. It would be possibleto merge these mechanisms in the main frame, but this wouldrequire elaborate controls to prevent mutual interference.

Another alternative is to maintain an instructional glossary ina separate file, to which the student has access at all times. Thistype of facility can be programmed in the more powerful systems,but its greater scope needs to be weighed against the advantagesof defining a term in context.

A flowchart of the A-Frame is given in Appendix B, and anabridged version of the actual machine program which wouldimplement it in Appendix C. (Scoring devices and reiterative textare omitted throughout).

SCANSION

Student input can be scanned for certain words or parts ofwords, in specified order or (if desired) in any order. Scan specifi-cations can also be nested (DISCUS) in such a way that input istested in its entirety first for one set of elements, and, if successful,for additional sets of elements. This permits very eaboratedecision trees.

A « not » facility can also be used (DISCUS) to destroy anymatch condition when a specified unwanted element crops up inthe response. This could lead to a solution of the problem ofdealing with negatives, double negatives, triple negatives, etc.,if an effective means of dealing with negative affixes can bedeveloped.

A series of scan specifications is always designed with a seriesof answers in mind. We have found it useful to categorize thesealong the following lines :

dr,


AL

AI

ALP

ALP m

ALPF

ALU

F

(Answer, Legal)

(Answer, Illegal)

(Answer, Legal, Perceived)

(Answer Legal, Perceived,Matched)

(Answer, Legal, Perceived,Failed)

(Answer, Legi.l, Unperceived)

A block of student input(a) of limited length,(b) containing at least one

term relevant to thequestion or its context,and

(c) in syntactical order.Any block of student inputnot meeting the above re-quirement.

Any AL which can bepredicted by the authox /instructor, as possible input.Any ALP which the author/instructor chooses to dealwith uniquely.Any ALP which the authordeems so remote or so in-consequential or so difficultto deal with in its existingform that it should be failed.Any AL which the authorfails to predict as a possi-bility.

(The sum of failed answers) AI ALU ALpF

The type of answer which principally occupies our attentionis ALPM. It is not necessary that the author have in mind every-thing that the student might say embodying the Au, m element(s).It is necessary that he select scan elements which are likely tomatch with a broad range of responses relevant to the question.In other words, he formulates a key word or phrase which willmatch members of a set of semantically similar input, and whichwill distinguish semantically dissimilar input by failing to matchwith it.

The range of ALPM is directly influenced by the form of thequestions which precede them. A question posing a very rigidrequirement (e.g., True-False) reduces the number of ALPM toa minimum, but at the same titne it vitiates the conversationalmode. Our current goal is to open up questions and to give the


student as much freedom in formulating his response as we can,without letting match probability fall below acceptable levels.

Appendix D lists he statements which would be displayed tostudents in response to various ALPM and F answers in theA-Frame, both initially (first pass) and recursively (subsequentpasses). Other elements of the program's reactions may beconstrued by referring to Appendix B and to the supportingmaterial in Appendix C. As demonstrated by the imaginarydialogues at the end of Appendix D, it would be possible for thestudent to get through the frame very rapidly indeed. On theother hand, it would also be possible for him to flounder aroundin it for several minutes.

THE STATEMENT/QUESTION

In a pamphlet entitled Microfiche Systems Planning Guide (I)(J. M. Baptie, ed.), appear the following statements :

There are three basic types of unit microforms. These are : (f) thetransparent film sheet, or microfiche, with a moderate to high pagecapacity; (a) the opaque MicrocardR, with a moderate to high pagecapacity; (3) the aperture card, with a low page capacity... (p. 4)

andTo date, the significant microfiche applications appear to be : (I) distri-bution of technical report literature... (p. 3)

These statements were reformulated into the statement/questionon the student form (Appendix A), i.e.,

There are three basic types of unit microforms. These are :(1) The transparent film sheet, or microfiche,(z) The opaque microcard,(3) The aperture card.

Which of these do you think is the most suitable for dissemination oftechnical literature ? Why ?

In the absence of extensive preparation, this is obviously avery faulty combination. What does the teacher mean by unitmicroforms ? Is microfiche the same as transparent film sheet, or are

(z) Copyright 1965 by Microcard Corporation, West Salem, Massachusetts.

2

5


they offcrod as two different things ? What are the microfiche,the opaque microcard, and the aperture card really like, anyway ?By « which » does the teacher mean « which one ? ); What scopeof dissemination is meant ? Suitable for whom ? What are theconditions that generally govern the dissemination of technicalliterature ?

To cap it all, the instructor asks « Why ? », which would be anexcellent question element if it weren't preceded 5y such a con-fusing array.

It might be possible for a person armed only with an amateur'sknowledge of photography and with the nature of technicalliterature to come up with the correct answer. However, it wasnot our purpose to establish this, but rather to see whether allrespondents might be led to a degree of understanding whichwould permit them to formulate a correct response. This seemsto accord with the theory that (at the problem-solving level)learning acquired through active participation by the student ismore useful and durable than that which is passively ingestedby him.

EXPERIENCE WITH THE g A » SET OF ANSWERS

As stipulated, twenty-five answers from the 50-answer samplewere used to test the scans (SC) and their associated reactive textsin the « A-Frame ». It was expected that the « A » set of answerswould reveal omissions and unforeseen ambiguities latent in thescan specifications. For example we might have specified

(SC) CLARITY, LEGIBLE, CLEAR,EASY EYES, EASIER EYES

with the idea of picking up something likeIt's better because it's more legible.

But then the occurrence of« It's better because it's more readable »

among the sample answers would show the need to add READ-ABLE to the specification.

Also the answers were expected to turn up instances whereinthe program reaction text should be modified to accommodate


variant input expressions. For example suppose we wanted totreat alike all answers expressing preference for microfichebecause it is readily reproduceable. We might write the reactivetext in the following form :

« YES, IT IS ».

Then if one of the samples turned out to be« I prefer microfiche because s simple copy machine can make copies veryquickly and cheaply »

clearly the reactive text would need to be adjusted. (Sometimesthe reaction logic goes completely awry and new subroutineshave to be added. In other cases the change of a single word willsuffice to make the reactive text more broadly suitable).

Both of these expectations were fulfilled, but we also discoveredthat the means for dealing with the « I don't know » family werecompletely inadequate. And this led to one or the most interestingdevelopments to come out of the exptriment.

THE « I DON'T KNOW » CONDITION

The idea had been to scan for versions of « I don't know »occurring with any of the following essential elements :

DISSEMINATION (TECHNICAL LITERATURE)

APERTURE (CARD)

UNITIZED (MICROFORM)(MICRO) FICHE (TRANSPARENT SHEET)

(MICRO) OPAQUE (CARD, MICROCARD)and to deal with them in the five satellite frames. If more thanone of the sertrate elements occurred in an «I don't know »answer, an appropriate succession of satellites would be called up.

Expressions of plain « I don't know » or its variants wouldinvoke a response asking the student to identify the terms heneeded to have explained.

Expressions of « I don't know » together with unidentifiedmatter would invoke a similar response, also letting him knowthat a portion of his answer had not been understood.

Altogether, eight of the twenty-five answers fell into these lasttwo categories I

10


Group I Don't know no further clue#1 I don't know.#1 I don't know.#9 I really haven't the slightest idea.

Group II Don't know plus unidentified matter4h I don't know enough to decide.#4 I don't know what these are exactly. So can't tell.# 5 I don't know. They all seem about equivalent to me.

%to Not enuf info to answer this would need to know propertiesof microforms and kinds/quality of available photocopy.

# I I I've only had occasion to use the microfilm on reels, so I can'treally say.

A ninth answer, « I can't honestly compare because I am notfamiliar with all 3 of the media », really belongs in Group II,except that it was caught accidentally by the scan for « I don'tknow » « aperture card (3) ».

On revision, we sought to identify the clues in # 5, #and # I x and to modify the scans as follows :

# 5 scan for « all, same, equivalent » and provide explanatoryroutine (probably a satellite).

# io add « enuf » and « info » to the « don't know » specification.#i t scan for « reels, rolls », etc. Although microfilm on reels is

outside of the problem, it is relevant and may provide a goodinsertion point inside of ;.he problem; so we would write anadditional statement covering it.

The original programmed response to Group I was. « I'll beglad to explain any of the terms I've used, if you wish. Justtype 'Clarify' ».

The original programmed response to Group II was « I don'tunderstand exactly what you need to have clarified ».

Answer # To would have fallen through to FAIL (anotheraccident) because « enuf » and « info » had been omitted fromthe scan specifica'ion. However, as luck would have it, theprogrammed response was not too bad : « How's that again ?I don't understand your response as worded ».

But all three responses fail to differentiate between cognitivestates which would be fairly obvious to a human interlocutor.Granted that we can patch up # 5, #8, # Jo, and # I r, the


,:maining five raise some interesting theoretical points whichwe are now obliged to deal with.

INFORMATION-STATE, MOTIVATION, AND MOTIVATIONAL INDICATORS

For the purpose of this study it has seemed reasonable to makecertain assumptions regarding the information-state signalledby « I don't know ». Without really plumbing the depths ofcognitive theory, we might say that :

1. Information, in order to be recognizable as such by more than one person,must be capable of being expressed in terms which mean approximatelythe same thing to others. This set of agreed meanings is the basis forsemantic transfer.

2. Awareness of information requires a substrate, comprised of relatableunits... (t) the presence, absence, and quality of which tend to bound theultimate knowledge state.

3. The cognitive process (either self-generated or derived from an outsidesource) consists of a perception of some kind of logical structure formedby the relationships between the information units. The integrity of thisperception depends on the individual's synthetic skill and on theaccuracy/completeness of the substrate. Its transferability depends onsemantic equivalence (I, above).

4. Self-sustaining cognition can be triggered by an outside stimulus or byan influx of critically related information units. In either case it representsa learning proces6 thich engages the student's whole attention. Derivedcognition, on the other hand, while it may be accepted, probably doesnot command the student's attention to the same degree.

The order in which the first three assumptions are stated isnot meant to imply an unvarying sequence. For example, gapsin the informational substrate do not preclude formation of aknowledge structure, even though the structure may be somewhattentative and unstable until the gaps are filled in. Nevertheless, theexistence of some substrate appears to be prerequisite to structuralperception, so a progression from (2) to (3) can be stated as afifth assumption. (In fact, we might think of structural perceptions

(t) Ausubel might speak of these as « anchoring ideas ». See D. P. Ammo,Educational Pgebology ; a Cognitive View (New York, Holt, Rinehart andWinston, 1968) (esp. « Learning and the availability of relevant anchoringideas », p. 132 et seq.).


being packed down in use until they can be employed as in-formational units, to serve as substrate elsewhere. But the reverseof such a process is hard to imagine).

Probably none of the assumed conditions could be broughtabout in the absence of some kind of motivation, and it appearsreasonable to suppose that this element is strongest at the momentof creation, i.e., when the structure is about to be perceived.Can we, then, use motivational indicators to tell us somethingabout the student's cognitive state ? For example, from expressionsof disinterest that a semantic block exists, or that the informationalsubstrate is so fragmentary that nothing can be made of it ?

Does a request for specifics indicate that the structure thoughshaky is beginning to take shape ?

These questions impel us to see if we can detect motivationalindicators imbedded in « I don't know »s, hoping that from thesewe might get an idea of information-state. This in turn wouldpermit us to react to individual «I don't know » answers withmore discrimination.

Reverting to the Group I answers :# r I don't know. Curt. No evidence of interest.# a I don't know. Possibly none of the terms in the

statement/question connect withany information unit.

#9 I really haven't the slightest idea. Almost explicit disinterest, plusovertones of distaste for thesystem itself. There is a possibilitythat the student has some infor-mation but prefers to dismiss it,for the sake of being expressive.In any case, we have to take himat his word.

In the face of zero motivation implying a near-zero informationalsubstrate, we have two choices :

We could say « All right, let's talk about something else »,and go to another part of the program trying to find somethingwhich the student knows enough about to get off to 'a fair learningstart. Later, he might become interested and knowledgeableenough in « microphotography », for example, to allow us toreturn him to the A-Frame on a better footing. (This faculty of

13


letting the student influence his own learning path is of courseone of the most promising features of CAI technology).

We could start building, bit by bit, the necessary collection ofinformation units by literally imposing them on the student adifficult and thankless task, and one which promises no earlysurge of motivation on his part.

In this particular study we attempted a modification of thesecond tactic in order to be in a position to observe second-passperformance later on. We tried to pinpoint critical informationunits which are necessary in dealing with the question, i.e.,characteristics of microform types. Then we pointed out ourawareness of the difficulty of answering the question withoutthese units. Finally we presented a tactful offer of assistance.

« I realize that it is almost impossible for you to answer without knowingsomething about the types of microforms mentioned, and about the factorsinvolved in dissemination of technical literature. I'd like to help you tacklethe problem from the beginning by supplying the information you need.Are you willing to work with me toward a correct solution ) o

Second-pass reaction to a « yes » from the student would be :« Which item would you like me to describe first ? Type « De-scribe ». Second pass reaction to a « no » from the studentwould be a signoff, with a consultation recommended.

Turning to the answers in Group II :# 3 I don't know enough to decide.

#4 I don't know what these are ex-actly so can't tell.

Polite regret implies some wil-lingness, so there may be a fewscraps of information upon whichto build.

Rather than trying to ascertain the areas of complete ignorance,the first move was to ascertain and strengthen the areas ofknowledge, then to fill in the gaps. The following text was aimedat accomplishing this :

« Well, suppose we start by examining the items you may already have heardabout. I may be able to add some useful information about them. Howabout microfiche ? Microcard ? Aperture card ? The problem of dissemi-nation of technical literature ? Are you acquainted with any of these ?

Second-pass reaction to « Yes » would be : « Which ? » Second-pass reaction to « No » would be to step down to the text given

/4


for plain « I don't know » answers. Second-pass reaction to aspecific would be to branch to a satellite. Third-pass reactionto another « I don't know » would be something like this :

« Shall we take them up one at a time ? Suppose we start with the problemof dissemination of technical literature ».

This would be followed by at least four of the satellite sub-routines.

To continue with Group II :# 5 I don't know. They all seem about

equivalent to me.

# 8 I can't honestly compare becauseI am not familiar with all 3 of themedia.

Here the student is clearly makingan effort of some kind. He mayalready have a fair idea of thedissemination problem, since hisresponse shows a narrowing tothe three forms suggested.

In this case we are justified in concentrating initially on a de-finition of the three types.

A-FRAME PERFORMANCE WITH OTHER TYPES OF ANSWER

In the remainder of this section are listed the other answers,and an indication of how the program handled them. In order tosave space at this point, we simply refer opposite each to theAppendix D code number which defines the reaction that wouldhave occurred. An assessment as to whether or not that reactionwould have been appropriate appears in the right-hand column.

Group#

Answer# Student input

AppendixD

channel

Appro-priate

1st pLss ?

IV

IV

6

8 (a)

I don't know what an aperturecard is. Nor (1) or (2).

I can't honestly compare becauseI am not familiar with all 3 of themedia.

8.a

7.d

Yes

No

(a) Repeated here for statistical purposes. (The program would have interpreted the inputas meaning «I don't know what an aperture card is ». The outcome might have beensatisfactory in this case, but no credit can be taken).

15

STUDENT FEEDBACK IN 235

Group#

Answer.# Student input

AppendixD

channel

Appro-priate

1st pass ?

IV 12 Please define opaque microcardand aperture card.

8.a Yes

V 13 I believe the opt que microcardis the most useful.

4.a Yes

V 14 Transparent film sheet; justseems easy to store and use.

2.a No (b)

V 15 Aperture card too awkward for abig file, easier to enlarge micro-fiche for convenient viewing.

4.b No (c)

VI 16 None. I believe in hard copy. 10.a YesI dislike sitting in dark roomsand looking at idiot boxes.

V 17 I think the microfiche would bebest because I think it can con-tain more info than the otherforms.

2.a No (d)

IV 18 3. The aperture card can behandled mechanically.

4.b No (e)

IV

IV

19

20

Aperture card, because the cita-don infro (sic) is readable withoutresorting to a reader,

3. Aperture cards because theyare processed easily.

4.b

4.b

No (f)

No (g)

(b) Should have gone to a.b.z. « Store a was overlooked in the scan specification.(c) Program fault. On revision, something to screen correct negative judgements is needed.(d) Should have gone to 2.b.z. « Contain more info a was overlooked in the scan specification.(e) « Why ? a will be misinterpreted. A scan for the sense of machine handling and processing

is required.(f) « Why ? a will be misinterpreted. A scan for the sense of eye-legible titling is needed

in any case, and perhaps « citation » should be one of the specified keywords, althoughits occurrence seems remote.

(g) Same as (0).

110


Group#

Answer# Student input

AppendixD

channel

Appro-priate

1st pass ?

IV 21 The opaque microcard is ade-quate.

4.a Yes

IV 22 The microfiche offers better re-production (visually) than themicrocard.

2.a No (h)

IV 23 I would suggest the microfiche,since complete information canbe stored in a small space.

2.b.2 Yes

IV 24 #2 is most suitable. As apositive copy of the text it isleast objectionable to the user.

2.a No (i)

IV 25 The microfiche has proved quiteuseful for cartographic informa-tion.

2.a Yes (j)

(h) The answer itself is ambiguous, but suggests the need for scans covering (z) manualcopying, (a) rendering into hard copy, and (3) clarity of display in a reading machine.

(i) A scan is needed to dispose of the positive-negative issue, if possible, followed by areturn to other considerations.

(3) « Cartographic a is too far out to provide for at present. One can only hope that the« Why ? » will elicit something which the system can process more meaningfully.

EVALUATION AND REVISION OP THE A-FRAME

Appendix F shows in tabular form the characteristics of eachof the «A » answers; how it fared in the «A -Frame »; and anevaluation of the A-Frame as to each, on a scale of i -o. Thesummary I ,f this data leads to one conclusion, namely, that althoughthe percentage of matches on the first pass was high (96 %), frameperformance as a whole was not commensurately good, since programreactions were suitable in only about half of the cases. Accordingly wewere obliged to evaluate the frame as a whole at less than 5o %.

Let us examine this more closely. The A-Frame was planned andwritten by an experienced CAI programmer, and his work was

I


checked and emended by a second person, also experienced inthe field. It is not surprising that between them they should haveforeseen a range of keywords such that 24 out of the 25 answerswere intercepted. But they failed to anticipate many of the sec-ondary elements which were needed to refine the meanings of theanswers. For example, every response containing a primaryelement : microfiche, microcard, or aperture card, was identified assuch, but further processing did not occur, because of variousomissions and other faults in dealing with such things as « me-chanical » and « info ».

This in itself was not unexpected. The original purpose of thetest was to see how much of an improvement could be madein the frame, using ideas suggested by operational experience.Admitting that no one person nor even two persons in collabora-tion could imagine the whole scope and variety of possible re-sponses to a fairly open question, we wanted to gauge the directutility of student responses in augmenting the range of ourperceptions in this regard. We might even decide that the bestway to design a frame was to erect only the barest structure,without trying to cover every eventuality, and to rely on aninflux of on-line responses from live students to point the wayto the frame's further development.

But the experience with « I don't know » answers raisedanother problem. It was not that the program reactions wereparticularly bad. They just weren't particularly good. They madeno account of attitudinal overtones which to a human teacherwould have clearly indicated shades of knowledge-condition inthe student. The effort to improve this situation needs to beconsidered apart from the simple refinements mentioned in thepreceding paragraph, because, as indicated in the discussion ofcognitive states, we believe that such differentiations might provegenerally useful as tools in CAI methodology.

In order to investigate this, all the words and phrases whichmight be expected to have key status in a « I don't know »statement were listed. Items were clustered in various ways, andthe clusters (of which there were finally ten) were mapped intoeach other on a trial and error basis until it was possible to discernroughly three levels of attitude. These combinations were then


assigned programmatically, yielding an array of scans we call the« M-series », set forth on page 276, Appendix G.

Taking the scan labelled « MO » (motivation zero) as an ex-ample, we see that it is composed of elements from the scanlabelled «-A », or from scans « Az » plus « Bi » plus « CI ».Any answer which had not been screened out by an earlier scanwould thus be matched by « MO » if it contained any of thefollowing :

... don't know...

... can't say...

... can't imagine...you expect...

. you think...

. at sea...

... with it...search me...

or a combination of the following :... don't have...... not have...... haven't...

expect have...... think have...

plus

plus

least..... vague...

vaguest...slightest...fuzzy...fuzziest...foggy...foggiest...remote...remotest...

. idea...notion...interest...concept...conception...inkling...

STUDENT FEEDBACK IN CAI

Thus

239

« What makes you think I could possibly answer that ? »« Search me 1 »« I'm sure I don't know ».« I haven't the foggiest notion ».

would all be matched in « MO » and be treated accordingly.The « M-Series » of scans would not need to bulk as large in

the program as might be supposed from the foregoing, becauseof the common device of forgiving parts of words which maybe missing or misspelled. For example the form FOGG* (PILOT) orFOGG' (Discus) accepts « foggy », « foggiest », « foggier »,

and « fogged ». A third language, FOIL (r), forgives a specifiedpercentage of characters : FOGGIEST (50 %).

The adoption of an additional line of inquiry did not seem toprejudice the original idea of the model, because it is possible toassess performance according to AL groups individually or incombination. Thus in the A-Frame the following breakdown canbe made :

Score Possible %

Groups I and II (« I don't know ») 5 1/2 8 69Groups III, IV, & V (all others) 6 1/2 17 38

Composite 12 25 48

The A-Frame's superficial success with Group I and II restssolely on the fact that forms of « I don't know » can be ex-haustively predicted, compared with all the other things that canbe said by a student, and a response that makes some sense can

(I) HESSELBART, J. C., « FOIL a file-oriented interpretive language » inProceedings of the 23rd National Conference of the Association for Computing Machinery(ACM Publication P-68, Princeton, Brandon, 1968), pp. 93-98. See also in thesame publication the review of CAI languages by Karl ZINN, « Programmingconversational use of computers for instruction » (pp. 83-92).

20


be trotted out. The M-series of scans represents a new object' .ve,the hope that we could significantly refine this ongoing sense.

As for Groups III, IV, and V, we could continue to pursuethe general intent of improving scan specifications word by word,so a?. to catch and channel a greater variety of unique responses.In some cases it would be necessary to write separate new scanstatements and program reactions to deal with aspects we hadcompletely missed in the first version, such as the idea of me-chanical handling.

PART II

PROCEDURE

Appendix G (Flow Chart) and Appendix H (Program) showthe model as revised and redesignated the « B-Frame D.

Very little was done to alter the flow, except to insert theM-series of scans just ahead of « DUNNO + GARB », and to adda satellite to deal with mechanical handling. The A-Frame flow-chart omitted certain controls specified in the program, and theB-Frame does the same.

Appendix I corresponds with Appendix D, in specifyingexactly what would happen to an identifiable type of input marchedby a particular scan statement.

All this work was completed before we allowed ourselves tolook at the « B » set of answers. When we had everything ready,we proceeded to drop them one by one into our paper « hopper ».It may be asked why we didn't actually put the program in thecomputer and test it on-line, in order to avoid any possibility ofshading the results. We would have liked to do this, of course,but as mentioned in the beginning we had assumed an augmenta-tion of an existing language (PILOT) with features of one whichwas not yet operational (DISCUS). In other words, we wereusing a non-existent CAI language. More recently, DISCUS hasindeed been implemented at the University of California Berkeleyand UCLA campuses, and we expect to use it for most of ourCAI programming.

Anyone monitoring the project might have objected to the

e-I


fact that we did not seem to require an absolute fit of some inputterms with the scan specifications, as originally laid down inAppendix C and Appendix H. The reason for this is that weassumed the « forgive » convention whereby « stor* », forexample, would match « storing », « storage », or « stored», butwrote one of these out for better understanding by the reader.This is obviously misleading, though, so we revised these twoAppendices to show such a specification as « storage » with onlythe underlined characters representing the absolute requirement.Thus a reader can perceive both the real specification and theterm idea behind it. Except for this reformatting, no retrospectivechanges of any consequence were made in any of the supportingdocuments.

COMPARISON OF RESULTS

The summaries of A-Frame and B-Frame observations arecompared in Table I. Clearly a remarkable improvement tookplace in respect to what we call « sense score », i.e., the percentageof student answers to which the program would have respondedsatisfactorily.A satisfactory response in every case is deemed to be one whichnot only makes sense on the first pass but offers a good chanceof leading to a sensible response on the next pass. (This rulesout the plausible but misleading response of « Why ? » to ques-tions A-I8 and A-x 9 (see page 230 where the student is talkingabout one thing and the program is talking about another).

In these assessments the student viewpoint controls; we cannotexcuse awkward « noncommunicating » responses ongrounds that they suit our pedagogical plan or that they will leadin the end to some kind of complete exposition.

Before looking into the significance of the results, we need toask if enough similarity existed between the A set of answersand the B set of answers to validate a comparison of A-Frameperformance and B-Frame performance. I think we can say thatthe motivation scores and the proportions of Group V answersshow that the two sets were quite alike certainly close enoughfor our purposes.

a R.

142 J. C. MEREDITH & D. VERGUSON

Table ICOMPARISON OF A-FRAME AND B-FRAME PERFORMANCE

ON SIM;LAR SETS OF ANSWERS

A-Frame B-Frame Change

Match score 24 (96 %) 25 (100 %) + 4 %

Sense scoreGroups I & II 5 1/2 / 8 (69 %) 5 1/2 / 6 (91 %) +22 %Groups III, IV, V, & v I 6 1/2 / 17 (38 %) 16 1/2 / 19 (86 %) +48 %Composite 12/25 (48 %) 22/25 (88 %) +40 %

Characteristics of the sets :

Motivation to 4x0.--- 0 2x0= 0Motivation 1 4 x 1 = 4 6 X 1 = 6 +2Motivation 2 3 X 2= 6 3 X 2= 6Motivation 3 14 x 3 = 42 14 x 3 = 42

Composite 52 54 +2Possible maximum 75 75Factored 52/75 54/75 +2/75

Answer Type Groupings :

I 3 1 2II 5 5III 1 0 1IV 3 6 +3V 12 13 +1VI 1 0 1

25 25

SIGNIFICANCE

The total number of unique AL which might be entered inresponse to a given question is a function of the openness of

as


that question, and can range from a very small set (as in responseto a True-False) to one which is very large (as in response to« What's it all about ? »). Let us call this number K and charac-terize it as finite (because of the constraint of our AL definition)but indeterminate (because no matter how many different legalanswers we can imagine in response to a question, someone maycome along with one we had not thought of).

If we were to scan randomly for AL the curve of match pro-bability would be a straight line as shown in Figure I. But wedo not scan randomly, because we know that the frequency ofoccurrence of some AL is far greater than that of others, and thatit is possible to attain a better match probability for K scans bytailoring them to the :core frequent AL. The uppermost curve(line 4) represents the cumulative frequency of occurrence ofAL and at the same time serves to represent the match proba-bility (P) which a CAI programmer could obtain for N scans ifhe knew the makeup of the items to the left of N.

But of course he doesn't know this. He can only guess, on thebasis of his perception of (t) the topic, (z) the student set, (3) theposition of the question in context, and (4) the language in use,which AL will occur more frequently than others, and howmuch more frequently. The richer these variables, the harderit will be for him to predict not only the identity of the AL tothe left of N but the sha-le of the curve of their frequency. Line zshows how his prediction will always be better than randombut less than perfect.

Line 3 represents a closer approach to Line 4, attained by virtueof revising and improving upon the original, highly subjective,work of the programmer. What is the most efficient way of doingthis ? By bolstering his predictions with external data furnishedhim by the system in the course of its operations.

We would like to get a better idea of the role of this externaldata (feedback) what does it look like ? how can it best becategorized ? how many iterations should be considered beforethe utility of the process is overshadowed by the tasks associatedwith it ? These questions have a direct bearing on the refinementof conversational CAI techniques.


F1.0

AL

Fig. I

We are still a long way from being able to scale curves z, 3,or 4, for even one set of variables, but at least the results obtainedfrom this study indicate that

(1) for an open question with a large K and an adult, heterogenous studentset, a striking improvement in meaningful intercept can be obtainedthrough close analysis of a modest number of live responses, and

(z) as a corollary, the rate of improvement to be derived from close analysisbeyond this modest number must decline quite rapidly.

It is only fair to ask how much of the improvement in meaning-ful intercept (which is a better measure than raw match proba-bility) should be credited to the M-series of scans, representinga completely new conceptual element in the B-Frame. Lookingback to the evaluations given for responses to Group I andGroup II answers in the A-set, it will be seen that none of them

g.5


was downgraded for failure to detect nuances of « I don't know »,but only for mech inistic faults. Thus the conditions determiningsense score are the same for both frames.

ACKNOWLEDGEMENTS

Our thanks to Miss Helen Yen, ILR staff, for gathering « stu-dent » input for our model frame, and to the fifty individualswho contributed to the sample.

APPENDIX A

FORMS

1. Attachment (I) is the form of answer sheet used by all Ely« students ».

2. Attachment (2) is the form used by the authors for analysisof student responses and program reactions.

Attachment (I)From : J. C. Meredith, Institute of Library Research

(Information Processing Laboratory Project)You are asked to participate in a survey of response patterns to a singlestatement/question typical of programmed instruction, as follows :

There are three basic types of unit microform.These are :i. The transparent film sheet, or microfiche.2. The opaque microcard(19.3. The aperture card.Which of these do you think is most suitable fordissemination of technical literature ? Why ?

(Assume that the program is capable of dealing with responses consisting ofsymbols, numerals, or English words, singly or in combinat ion, up to two linesin length. Your response may be in the form of a statement or a question orboth, but should at least be relevant. Try to decide on your answer quickly.)Response :

a'


Thank you. Please check the appropriate box(es)

Attachment (2)

STUDENT ANSWER

PROGRAM RESPONSE

I. First passz. Likely znd pass

SATISFACTORY ?

TYPE OF MODIFICATION INDICATED

REVISED PROGRAM ELEMENT

APPENDIX 3

[11 LS

1:1 ILROther

«A -FRAME FLOWCHART

I. Attached.

z. The chart represents in abbreviated form the decision sequencesenvisaged for the «A -Frame ». The following notes pertain :

i) The «Q entry (large circle) represents student input, terminatingwith a « send » signal which causes the computer to resumeprocessing.

z) « Q » returns (small circles) imply a textual display of some kind,generated according to the specifications laid down in the programitself, and held on the CRT while the computer is waiting for thenext input. Consult Appendix D to ascertain actual text planned foruse in various situations.

3) « OUT » (small circles) signifies exit to the next frame.4) To avoid excessive complication of the flowcart, several tests

(diamonds) have been omitted. For example, the first five major

0 4.1

0 0t.0

WM

'

00

NM

I15

1K

Z?

05A

TE

LL:

"UN

I T Z

"

.11.

7 tL

e

T.. 0

1"o

r I

IIER

Z5,

331.

1!

on.

00

KA

LI.

O.o

t

tusi

f 'Le

!

00

7.4

v.,

v.. 0

FI1

11_

11.1

1 .3

(Sr

JO

73%

? 44

0, T

i. C

Z, 1

3,35

0. (

6YP

AS

SI

z3,

35. I

I, lt,

to

A-F

ram

e Fl

owch

art

11?

0

7t.

ca A,

Um

, 'N

C

Trr

ts

4

Fir

ER

!

Itt 0


tests to the left of the « DUNNO » diamond entail additionalsubordinate tests and decisions associated with recursive scan of thesame input string, under a series of « nested » scan parameters.These are omitted from the diagram.

5) The group of operations to the left of « DUNNO » is devoted toprovision of information intended to help the student solve theproblem. One of five information « satellites » UNITZ isshown in the large box. It should be visualized as occupying theaction box marked « CALL UNITZ » (hand).

6) The group of operations to the right of « DUNNO » is designedto accommodate a student's attempt to make the right choice. Eachchoice and the first recognized valid reason, or a combination(not shown) of the two or three best reasons, call forth a uniqueblock of text. Note that if the student fails to give a recognizedreason, the program loops through to a display of « Why ? » to Q,and his second response is scanned without requiring him to repeathis choice. (This is another case where the diagram omits specifictests).

7) The remaining operations (bottom center) are miscellaneous tests andutility operations, ending with two fail-match displays terminatingthe frame.

8) The typed numerals refer to answers in the «A » set, located so asto show how they would have been treated in a first pass throughthe program.

APPENDIX C

THE « A-FRAME N PROGRAM

i. Attached.

2. The program listing resembles actual source input in a high-level CAI programming language (selected string match type),such as PILOT. We have, however, omitted all delimiters (whichserve only to notify the compiler whether it is dealing with alabel, an operation code, or an operand) and have stylized thescan specification in logical form (e.g., a comma means « or »).

3. The following notes will assist the reader in interpreting theprogram :

x) All operations are sequential from left to right and from top tobottom.

Rq


a) For purposes of this sample, « G » signifies a display occurringonly if a preceding scan specification has been successfully matched.« F » signifies a display occurring only if all match attempts havefailed; and « T » signifies a display occurring regardless of matchcondition if execution reaches that point.

3) Reference to a label in quotes means that everything in the operandso labelled is to be inserted at that point. The expanded versions ofall labelled alphabetic strings are listed toward the end of the program.

4) The control statements in the right hand column would in manycases be constrained to operate only under a « match » condition,the same as a « G » textual display. The necessary qualifiers for thishave been omitted for the sake of simplicity.

5) Control statements ADD, TALLY, and MARK mean « add »,« add one », and « ensure one and one only in » respectively. Theyaffect the contents of numerical counters whose existence is estab-lished implicitly rather than by formal definition.

6) The sequence in which various match attempts are made reflects therule that normally scans terminate immediately on successful match

the remainder of student input is ignored. This favors puttingthe larger, more difficult specifications ahead of those which areeasier to satisfy. For example, the scan for « I » is placed at thevery end, so that it will not ensnare emphatic statements whichcontain sought-for elements specified in other scans.

7) Many of the planned unique text displays ste 3o similar in structurethat they can be assumed. or example the response to «I likemicrofiche because it's convenient » would call forth exactly thesame kind of text as « I like microfiche because it's cheap ». To avoidneedless repetition, many of these have been omitted from the listing.

8) A sample satellite is shown on the last two sheets of the program.The other four satellites would resemble it very closely. It mighteven be feasible to use just one satellite with a choice of texts to be« USED » depending on which indicators (variables such asMICROY) had been activated.

0

« A

» S

ourc

e Pr

ogra

m

Lab

els

Cod

eD

ispl

ay te

xt a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

EX

POT

(Ini

tial s

tate

men

t and

que

stio

n)

MIC

RO

.AA

(stu

dent

inpu

t)

R(«

DU

NN

O »

, CL

AR

IFY

) ±

(U

NIT

, UN

ITIZ

ED

)T

AL

LY

(U

NIT

Y, P

RO

BY

),IF

UN

ITY

2 A

DD

UN

ITX

,>...

.

TO

DU

PY

R.

(« D

UN

NO

», C

LA

RIF

Y)

+ (

TR

AN

SPA

RE

NT

, FIL

M. S

HE

ET

,FI

CH

E, M

ICR

OFI

CH

E, «

N.1

»)

TA

LL

Y (

MIC

RO

Y, P

RO

BY

),IF

MIC

RO

Y >

2 A

DD

MIC

RO

XI

TO

DU

PY

(« D

UN

NO

», C

LA

RIF

Y)

+ (

OPA

QU

E, M

ICR

OC

AR

D, «

N.2

»)

TA

LL

Y (

OPA

QY

, PR

OB

Y),

IF O

PAQ

Y e

2 A

DD

OPA

QX

TO

DU

PY

R(«

DU

NN

O »

, CL

AR

IFY

) +

(A

PER

TU

RE

, « N

.3 »

)T

AL

LY

(A

PER

Y, P

RO

BY

),IF

APE

RY

?...

. 2 A

DD

APE

RX

TO

DU

PY

IF (

UN

ITY

, MIC

RO

Y, O

PAQ

Y,

DIS

SEY

) >

2 U

SE D

UP,

AD

D«

UN

ITZ

» T

O P

RO

B U

NIT

Y,

AD

D «

MIC

RO

X »

TO

PR

OB

ON

MIC

RO

Y, A

DD

« O

PAQ

X »

TO

PRO

B O

N O

PAQ

Y, A

DD

« D

ISSE

X »

TO

PR

OB

ON

DIS

SEY

GI

UN

DE

RST

AN

D Y

OU

NE

ED

MO

RE

IN

FOR

MA

TIO

N A

BO

UT

« PR

OB

»IF

PR

OB

Y _

>...

. 2 U

SE T

AK

EU

P,C

AL

L U

NIT

Z O

N U

NIT

Y, C

AL

LM

ICR

OZ

ON

MIC

RO

Y, C

AL

LO

FAQ

Z O

N O

PAQ

Y, C

AL

L A

PER

ZO

N A

PER

Y, C

AL

L D

ISSE

Z O

ND

ISSE

Y, J

UM

P T

O M

ICR

O.A

bk

Lab

els

Cod

eD

ispl

ay te

xt a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

R«D

UN

N°

» +

AN

Y

GO

NE

OF

TH

EM

IS

SUPE

RIO

R I

N R

EPR

OD

UC

EA

BIL

ITY

, CO

ST,

GE

NE

RA

L U

SE, E

TC

. DO

YO

U N

EE

D M

OR

E I

NFO

RM

AT

ION

?IF

SO

, TY

PE a

CL

AR

IFY

0JU

MP

TO

MIC

RO

.A

R«D

UN

NO

» -

I- «

GA

RB

»

GI

DO

N'T

UN

DE

RST

AN

D E

XA

CT

LY

WH

AT

YO

U N

EE

D T

OH

AV

E C

LA

RIF

IED

.M

AR

K C

LA

RA

, JU

MP

MIC

RO

.A

R«D

UN

NO

a, N

OT

CO

MPE

TE

NT

, IN

CO

MPE

TE

NT

, NE

ED

INFO

RM

AT

ION

, RE

QU

IRE

IN

FOR

MA

TIO

N

GI'L

L B

E G

LA

D T

O E

XPL

AIN

AN

Y O

F T

HE

TE

RM

S I'V

EU

SED

, IF

YO

U W

ISH

. JU

ST T

YPE

a C

LA

RIF

Y0

MA

RK

CL

AR

A, J

UM

P M

ICR

O.A

R«N

.1 »

« P

LU

SCA

P »

a PL

USC

AP

» «

N.1

MIC

RO

CA

PG

YO

U A

RE

CO

RR

EC

T I

N P

ICK

ING

MIC

RO

FIC

HE

, BU

T T

HE

RE

ASO

N I

T I

S SU

PER

IOR

FO

R T

HE

DIS

SEM

INA

TIO

N O

FT

EC

HN

ICA

L L

ITE

RA

TU

RE

IS

NO

T T

HA

T I

T H

AS

GR

EA

TE

RC

APA

CIT

Y (

WH

ICH

DE

PEN

DS

ON

TH

E R

ED

UC

TIO

N R

AT

IO

R

USE

D)

BU

T O

N T

HE

i A

CT

TH

AT

a F

ICH

UP

»

c N

.1 »

+ «

PL

USE

ASE

» +

« P

LU

SCH

EA

P »

JUM

PT T

O N

EX

TFR

AM

E

MIC

RO

BO

TH

GG

OO

D. Y

OU

CO

RR

EC

TL

Y P

ICK

ED

TW

O O

F T

HE

IM

POR

TA

NT

AD

VA

NT

AG

ES

OF

FIC

HE

. a F

ICH

UP

aJU

MP

TO

NE

XT

FRA

ME

MIC

RO

EA

SEG

YO

U A

RE

CO

RR

EC

T I

N P

ICK

ING

MIC

RO

FIC

HE

, AN

D E

ASE

OF

DIS

TR

IBU

TIO

N I

S O

NE

OF

ITS

CO

MPE

LL

ING

AD

VA

N-

TA

GE

S. a

FIC

HU

P »

JUM

P T

O N

EX

TFR

AM

E

I

Lab

els

Cod

eD

ispl

ay te

xt a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

R«

N.1

s «

PL

USC

HE

AP

», «

PL

USC

HE

AP

» c

N.1

»

MIC

RO

CH

EA

PG

YO

U A

RE

CO

RR

EC

T I

N P

ICK

ING

MIC

RO

FIC

HE

, AN

DC

OST

CO

NSI

DE

RA

TIO

NS

CE

RT

AIN

LY

EN

TE

R I

NT

OT

HE

PIC

TU

RE

. «FI

CH

UP

»JU

MP

TO

NE

XT

FRA

ME

R«N

.1 »

« P

LU

SEX

PER

IEN

CE

», a

PL

USE

XPE

RIE

NC

E »

« N

.1 »

GC

AN

YO

U C

OM

PAR

E T

HE

M M

OR

E O

N T

HE

BA

SIS

OF

INFO

RM

-A

TIO

N?

IF Y

OU

WO

UL

D L

IKE

TO

KN

OW

MO

RE

AB

OU

T T

HE

OT

HE

R F

OR

MS,

OR

AB

OU

T T

HE

GE

NE

RA

L R

EQ

UIR

EM

EN

TS

FOR

DIS

SEM

INA

TIO

N O

F T

EC

HN

ICA

L L

ITE

RA

TU

RE

, JU

STT

YPE

a C

LA

RIF

Ya

JUM

P T

O M

ICR

O.A

(rep

eat s

erie

s fo

r a

N.2

» a

nd «

N.3

)

RR

EJE

CT

, HA

TE

, DIS

LIK

E, «

NE

GA

T »

LIK

E

GA

SID

E F

RO

M Y

OU

R P

ER

SON

AL

FE

EL

ING

S IN

TH

E M

AT

TE

R,

CA

N Y

OU

EX

PRE

SS A

N O

PIN

ION

AS

TO

WH

ICH

MIG

HT

BE

LE

SS U

NSU

ITA

BL

E F

OR

TH

E S

TA

TE

D P

UR

POSE

?JU

MP

TO

MIC

RO

.A

R«

NE

GA

T »

EX

PER

IEN

CE

GI'D

LIK

E Y

OU

TO

CH

OO

SE O

N T

HE

BA

SIS

OF

AV

AIL

AB

LE

INFO

RM

AT

ION

, IF

POSS

IBL

E. P

LE

ASE

TY

PE «

CL

AR

IFY

» FO

R E

XPL

AN

AT

ION

OF

IND

IVID

UA

LE

LE

ME

NT

S.JU

MP

TO

MIC

RO

.A

R«N

EG

AT

»T

AL

LY

HO

PEL

ESS

, IF

HO

PEL

ESS

= 1

JU

MP

CH

AN

CE

GPL

EA

SE S

IGN

OFF

NO

WJU

MP

TO

EN

D

Lab

els

Cod

eD

ispl

ay te

st a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

CH

AN

CE

GW

OU

LD

YO

U L

IKE

MO

RE

IN

FOR

MA

TIO

N ?

IF

SO, P

LE

ASE

IND

ICA

TE

BY

TY

PIN

G «

CL

AR

IFY

D

(WH

ICH

EV

ER

TE

RM

BO

TH

ER

S Y

OU

).JU

MP

TO

MIC

RO

.A

R«

N.1

» +

« N

.2 D

+ «

N.3

» (

any

2, a

ny o

rder

)

RA

LL

, AN

Y

GSE

E I

F Y

OU

CA

N P

ICK

JU

ST O

NE

, ON

TH

E B

ASI

S O

FW

HA

T Y

OU

CA

N L

EA

RN

FR

OM

ME

ON

TH

E S

UB

JEC

T.

JUM

P T

O M

ICR

O.A

R«

N.1

»

GW

HY

?T

AL

LY

(A

NSY

, MIC

RO

B),

JUM

P T

O M

ICR

O.A

R«

N.2

»

GW

HY

?T

AL

LY

(AN

SY, O

PAQ

B),

JUM

P T

O M

ICR

O.A

R«

N.3

»

GW

HY

?T

AL

LY

(A

NSY

, APE

RB

),JU

MP

TO

MIC

RO

.A

R«

PLU

SCA

P »

IF M

ICR

OB

JU

MP

MIC

RO

CA

P,IF

OPA

QB

JU

MP

OPA

QC

AP,

IF A

PER

B J

UM

P A

PER

CA

P

R«

PLU

SCL

AR

oIF

MIC

RO

B J

UM

P M

ICR

OC

LA

R,

IF O

PAQ

B J

UM

P O

PAQ

CL

AR

,IF

APE

RB

JU

MP

APE

RC

LA

R

Lab

els

Cod

eD

ispl

ay te

xt a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

R G F F

1 RE

AL

LY

? I

'M S

UR

PRIS

ED

YO

U F

EE

L S

O S

TR

ON

GL

Y A

BO

UT

IT

.IF

YO

U'D

LIK

E C

LA

RIF

ICA

TIO

N O

F A

NY

EL

EM

EN

T O

F T

HE

STA

TE

ME

NT

AB

OU

T U

NIT

MIC

RO

FOR

MS,

OR

OF

TH

E Q

UE

STIO

N,

JUST

ASK

. OT

HE

RW

ISE

PL

EA

SE S

IGN

OFF

AN

D N

OT

IFY

TH

EL

AB

ASS

IST

AN

T

HO

W'S

TH

AT

AG

AIN

? I

DO

N'T

UN

DE

RST

AN

D Y

OU

R R

ESP

ON

SEA

S W

OR

DE

D.

WE

LL

, I'M

LIC

KE

D. G

OO

DB

Y.

JUM

P T

O M

ICR

O.A

JUM

P T

O M

ICR

O.A

JUM

P T

O E

ND

(Not

e : T

he f

ollo

win

g ar

e th

e ex

pand

ed v

ersi

on o

f th

e st

ring

s re

ferr

ed to

by

quot

ed la

bels

)

UN

ITZ

MIC

RO

X

OPA

QX

APE

RX

DIS

SEX

DU

NN

O

GA

RB

T T T T T T T

UN

IT O

R U

NIT

IZE

D M

ICR

OFO

RM

S

TR

AN

SPA

RE

NT

FIL

M S

HE

ET

, OR

MIC

RO

FIC

HE

OPA

QU

E M

ICR

OC

AR

D

APE

RT

UR

E C

AR

DS

DIS

SEM

INA

TIO

N O

F T

EC

HN

ICA

L L

ITE

RA

TU

RE

DO

N'T

KN

OW

, HA

VE

N'T

, LE

AST

, SL

IGH

TE

ST, L

OST

, OU

TO

F IT

, SE

AR

CH

, CO

NFU

SED

, CA

N'T

DE

CID

E, C

AN

'TC

HO

OSE

, CA

N'T

CH

OIC

E, C

AN

'T S

AY

, HA

VE

NE

VE

R,

(etc

.)

(alp

habe

t)

Lab

els

1

Cod

eD

ispl

ay te

xt a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

DU

P

TA

KU

P

POSI

T

NE

GA

T

RPT

PLU

SCA

P

PLU

SCL

AR

NE

GC

LA

R

PLU

SEA

SE

PLU

SEX

PER

-T

IEN

CE

NE

GN

OT

ES

T

PLU

SCH

EA

PT

NE

GL

OSE

I B

FUR

VE

YO

U A

LR

EA

DY

ME

NT

ION

ED

YO

UR

PR

OB

LE

M W

ITH

a D

UPY

» A

ND

I T

RIE

D T

O E

XPL

AIN

. IF

YO

U A

RE

ST

ILL

AT

AL

OSS

, PE

RH

APS

YO

U S

HO

UL

D S

IGN

OFF

AN

D S

EE

K H

EL

P.

WE

WIL

L T

AK

E T

HE

M U

P IN

TH

AT

OR

DE

R.

(For

ms

of a

ffir

mat

ion)

, MO

RE

MO

ST

(For

ms

of n

egat

ion)

, LE

AST

, LE

SS

HE

RE

IS

TH

E O

RIG

INA

L Q

UE

STIO

N A

GA

IN :

« E

XPO

»

CA

PAC

ITY

, NU

MB

ER

, VO

LU

ME

, SPA

CE

, CO

NT

AIN

, HO

LD

CL

AR

ITY

, CL

EA

R, F

INE

, FIN

EN

ESS

, VIE

WIN

G, R

ESO

LU

TIO

N

HA

RD

TO

RE

AD

, PO

OR

RE

SOL

UT

ION

, DIM

, DA

RK

, HU

RT

S E

YE

S,PR

INT

TO

O S

MA

LL

, FU

ZZ

Y

EA

SE O

F D

IST

RIB

UT

ION

, EA

SE O

F M

AIL

ING

, EA

SE O

FSE

ND

ING

EX

PER

IEN

CE

IN

USE

, KN

OW

MO

RE

AB

OU

T

HA

RD

NO

TE

S, D

IFFI

CU

LT

NO

TE

S, U

ND

ER

LIN

E, W

RIT

E O

N

CO

ST, C

HE

AP,

IN

EX

PEN

SIV

E, S

AV

E M

ON

EY

TO

O S

MA

LL

, LO

SE, L

OST

, MIS

PLA

CE

, MIS

PLA

CE

D,

LO

OSE

TR

AC

K O

F

Lab

els

NE

GT

RU

B

FIC

HU

P

N.1

N.2

N.3

Cod

e

T T T T

Dis

play

text

and

inpu

t sca

n sp

ecif

icat

ions

Con

trol

sta

tem

ents

TO

O M

UC

H T

RO

UB

LE

, TIM

E C

ON

SUM

ING

MIC

RO

FIC

HE

CA

N B

E D

UPL

ICA

TE

D V

ER

Y R

EA

DIL

Y A

ND

INE

XPE

NSI

VE

LY

, AN

D I

N F

AC

T I

S T

HE

ST

AN

DA

RD

IN

USE

BY

TH

E F

ED

ER

AL

GO

VE

RN

ME

NT

FO

R D

IST

RIB

UT

ION

OF

CL

EA

RIN

GH

OU

SE, D

OD

, NA

SA, A

EC

, AN

D E

RIC

DO

CU

ME

NT

SW

HIC

H I

NC

LU

DE

A L

AR

GE

PO

RT

ION

OF

OU

R T

EC

HN

ICA

L L

ITE

RA

TU

RE

OU

TPU

T. R

EC

IPIE

NT

SC

AN

RE

AD

ILY

USE

TH

EM

IN

A W

IDE

VA

RIE

TY

OF

RE

AD

ER

S, O

R (

UN

LIK

E O

PAQ

UE

CA

RD

S) R

EN

DE

R T

HE

MB

AC

K I

NT

O E

YE

-LE

GIB

LE

FO

RM

TH

RO

UG

H R

EA

DE

R/

PRIN

TE

RS.

WIT

H A

PER

TU

RE

CA

RD

S SE

VE

RA

L U

NIT

SA

RE

RE

QU

IRE

D T

O C

ON

VE

Y A

MU

LT

I-PA

GE

DO

CU

ME

NT

,H

EN

CE

TH

EY

AR

E M

OR

E S

UIT

AB

LE

FO

R S

ING

LE

IT

EM

S,SU

CH

AS

EN

GIN

EE

RIN

G D

RA

WIN

GS.

(A

PER

TU

RE

CA

RD

SA

RE

MO

RE

OFT

EN

USE

D I

N A

ME

CH

AN

IZE

D S

YST

EM

BU

T T

HIS

IN

IT

SEL

F D

OE

S N

OT

MA

KE

TH

EM

LE

SSSU

ITA

BL

E T

HA

N F

ICH

E, O

R M

OR

E S

O.)

1, O

NE

, FIR

ST, M

ICR

OFI

CH

E, T

RA

NSP

AR

EN

T, F

ILM

SHE

ET

2, T

WO

, SE

CO

ND

, OPA

QU

E, M

ICR

OC

AR

D

3, T

HIR

D, L

AST

, APE

RT

UR

E

(Not

e : T

he f

ollo

win

g is

a ty

pica

l « s

atel

lite

» fr

ame,

cal

labl

e at

any

poi

nt).

TA

UN

IT M

ICR

OFO

RM

IS

ON

E W

HIC

H C

ON

TA

INS

A S

ING

LE

DO

CU

ME

NT

, UN

LIK

E R

OL

L M

ICR

OFO

RM

(M

ICR

OFI

LM

) W

HIC

HM

AY

CO

NT

AIN

SE

VE

RA

L D

OC

UM

EN

TS

ON

ON

E R

OL

L, O

R R

EE

L.

EA

CH

UN

IT M

AY

BE

LA

BFT

T.F

D, I

ND

EX

ED

, ST

OR

ED

, CO

PIE

D,

Lab

els

Cod

eD

ispl

ay te

xt a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

OR

. TR

AN

SFE

RR

ED

TO

A U

SER

IN

DE

PEN

DE

NT

LY

OF

AN

Y O

TH

ER

DO

CU

ME

NT

. SO

UN

IT M

ICR

OFO

RM

IS

SPE

CIA

LL

Y W

EL

LSU

ITE

D F

OR

DIS

SEM

INA

TIO

N O

F T

EC

HN

ICA

L L

ITE

RA

TU

RE

WH

ICH

IS

USU

AL

LY

PU

BL

ISH

ED

IN

TH

E F

OR

M O

F JO

UR

NA

LA

RT

ICT

PS

OR

DO

CU

ME

NT

S O

F SI

MIL

AR

LE

NG

TH

. (U

SUA

LL

YSE

PAR

AT

E D

OC

UM

EN

TS)

.D

OE

S T

HIS

EX

PLA

IN T

HE

TE

RM

AD

EQ

UA

TE

LY

FO

R Y

OU

?A

DD

-1

TO

PR

OB

Y, A

DD

-10

TO

UN

ITY

UN

IT.A

A(s

tude

nt in

put)

R«

POSI

T »

IF P

RO

BY

JU

MP

TO

MO

RE

DU

N

GFI

NE

. IF

YO

U A

RE

IN

DO

UB

T A

BO

UT

AN

Y O

F T

HE

OT

HE

R T

ER

MS,

DO

N'T

HE

SIT

AT

E T

O A

SK. «

RPT

»JU

MP

RE

TO

UR

MO

RE

DU

NG

IN T

HA

T C

ASE

WE

WIL

L P

ASS

TO

TH

E M

EA

NIN

G O

FIF

MIC

RO

Y =

I U

SE M

ICR

OX

,IF

MIC

RO

Y J

UM

P R

ET

OU

R,

IF O

PAQ

Y =

1 U

SE O

PAQ

X,

IF O

PAQ

Y J

UM

P R

ET

OU

R,

IF A

PER

Y =

I U

SE A

PER

X,

IF A

PER

Y J

UM

P R

ET

OU

RIF

DIS

SEY

= I

USE

DIS

SEX

,IF

DIS

SEY

JU

MP

RE

TO

UR

R«

NE

GA

T »

GR

AT

HE

R T

HA

N S

PEN

D M

OR

E C

AI

TIM

E O

N I

T N

OW

, ISU

GG

EST

TH

AT

YO

U E

ITH

ER

SIG

N O

FF A

ND

SE

EK

OU

TA

N I

NST

RU

CT

OR

, OR

TY

PE 0

PA

SS

» A

ND

I W

ILL

GO

ON

TO

TH

E N

EX

T Q

UE

STIO

N.

JUM

PT T

O U

NIT

.A

Lab

els

Cod

eD

ispl

ay te

xt a

nd in

put s

can

spec

ific

atio

nsC

ontr

ol s

tate

men

ts

RE

TO

UR

R G R G F E

PASS

AL

L R

IGH

T. B

UT

PL

EA

SE G

ET

CL

AR

IFIC

AT

ION

AT

TH

EFI

RST

OPP

OR

TU

NIT

Y.

SIG

N O

FF, Q

UIT

YO

U A

RE

NO

W S

IGN

ED

OFF

BE

G P

AR

DO

N ?

(ter

min

ates

cal

l, re

turn

s to

poi

nt w

henc

e it

cam

e).

JUM

P T

O N

EX

TFR

AM

E

JUM

P T

O E

ND

JUM

P T

O U

NIT

.A

NE

XT

FRA

ME

EN

D


APPENDIX D

REACTION TEXT DISPLAYS - K A-FRAME

259

1. List attached.z. This Appendix lists the textual responses programmed in the« A-Frame », that is, the responses which are called forth fromthe computer as part of its reaction to student input. Romannumerals indicate « pass status » as follows :

I. Responses which can occur on the first pass (i.e., in answer to a student'sfirst input), or on subsequent passes.

II. Responses which can occur only on the second or subsequent passes(depending on certain conditions being set in an earlier pass).

III. Responses occurring on the second or subsequent pass if a satellite hasbeen called.

All student answers are classified ALPM (legal, perceived, matched)except numbers I2, 27, z8, and 31, which are F (failed), beingeither AI (illegal), ALU (legal but unperceived), or AuF (legal,perceived, but intentionally failed).3. Types of student answer are indicative only. For example, eachtype or subtype listed represents hundreds or even thousands ofdifferent inputs all containing a required key element. For theexact specification of these elements, consult Appendix C. Thedesignations « Right » and « Wrong » indicate the instructor'sviewpoint and have no mechanistic significance in the program.4. For the sake of brevity, many unique responses are left to thereader's imagination, in situations where they would requireonly slight rewording of a response which has been written out.S. None of the text should be interpreted substantively. Weexpect an expert in the field might think of many counter-argu-ments.We are concerned here with dialogue rather than with factualdistinctions, and in regard to the latter defer entirely to copy-righted reference 5 .

6. At the end of this Appendix are included two sample dialogueswhich might occur in an on-line situation, to illustrate the kindof interchange which might take place in the « A-Frame ».

z6o J. C. MEREDITH & D. FERGUSON

I. First pass.Type of answer (numbers and letters Program display reactionestablish categories referred to in (« E » = « exit to next frame »)Appendix F)

z. Right both elements

a. Reason : ease of distribution, You are correct in pickingof mailing, of sending microfiche, and ease of distri-

bution and use is one of itscompelling advantages.(use FICHUP text)FICHUP :Microfiche can be duplicatedvery readily and inexpensively,and in fact is the standard in useby the federal government fordistribution of Clearinghouse,DOD, NASA, AEC, and ERICdocuments which include alarge portion of our technicalliterature output. Recipientscan readily use them in a widevariety of readers, or (unlikeopaque cards) render them backinto eye-legible form throughreader-printers. With aperturecards, several units are requiredto convey a multi-page docu-ment, hence they are moresuitable for single items, suchas a file of engineering draw-ings. (Aperture cards are moreoften used in a mechanizedsystem, but this in. itself doesnot make them more suitablethan fiche, or less suitable). (E)

b. Reason : cost, cheap, inexpens- You are correct in pickingive, save money microfiche, and cost consider-

ations certainly enter into thepicture. Microfiche can be du-plicated very readily and in-expensively (etc. seeFICHUP in La above). (E)

At 1


c. Reason : r.a i.b

z. Right first element only

a. No reason givenb. Wrong reason given

experience (« because I'vehad more experience withit »).

(a)

(3)(4)

capacity (« because it hasgreater capacity »).

clarityknowledge (« the only oneI know about »).

3. Wrong Two first elements

4. Wrong wrong first elementa. Microcardb. Aperture card

5. Wrong Don't knowa. Alone

6. Wrong Don't know + uniden-tified element

44

Good. You correctly pickedtwo of the important advan-tages of fiche. (FICHUP) (E)

Why ?

Can you compare them moreon the basis of information ?If you would like to knowmore about the other forms, orabout the general requirementsfor dissemination of technicalliterature, just type« Clarify ».

You are correct in pickingmicrofiche, but the reason it issuperior for dissemination oftechnical literature is not that ithas greater capacity than otherunit forms (which depends onthe reduction ratio in use) buton the fact that microfiche canbe duplicated very readily andinexpensively. (FICHUP) (E)(similar to 2.b.a above) (E)(similar to z.b.z above) (E)

See if you can pick just one, onthe basis of what you can learnfrom me on the subject.

Why ?

Why ?

I'll be glad to explain any of theterms I've used, if you wish.Just type « Clarify ».

z6z j. C. MEREDITH & D. FERGUSON

a. « Don't know because( ?) ».

7. Wrong Don't know + singleidentified

a. Unit, unitized

b. Transparent, film, sheet,fiche, microfiche

c. Opaque, microcardd. Aperturee. Dissemination, technical,

literature

8. Wrong Don't know + multipleidentified

a. Any combination of 7.a, 7.b,7.c, 7.d, and 7.e)

9. Wrong Don't know general

a. Any

I don't understand exactly whatyou need to have clarified.

(First pass only)I understand you need moreinformation about « unit a or« unitized » microforms. A unitmicroform is one which...(etc.) Does this resolve youruncertainty concerning thedefinition of unitized micro-forms ?

(Second pass et seq.) I believeyou already mentioned yourproblem with « unit » or « u-nitized » microforms, and Itried to explain. If you are stillconfused, perhaps you shouldsign off and seek help.(Similar to 7.a)

(Similar to 7.a)

(Similar to 7.a)

(Similar to 7.a)

I understand you need moreinformation about (7.a, 7.b,7.c, 7.d, 7.e). We will take themup in that order. (Stacked sub-routines of 7.a-e)

One of them is superior in re-produceability, cost, generaluse, etc. Do you need moreinformation ? If so, type « Clar-


b. No experience

to. Wrong Rejection

a. Dislike

b. Expletive

x. Wrong Request

a. Clarify, define

x z. Fail

I'd like you to choose on thebasis of available information,if possible. Type « Clarify »for explanation of individualelements.

Aside from your personal feel-ings in the matter, can youexpress an opinion as to whichmight be least unsuitable forthe stated purpose ?Really ? I'm surprised you feelso strongly about it. If you'dlike clarification of any elementof the statement about unitmicroforms, or of the question,just ask. Otherwise please signoff and notify the lab assistant.

(Same as 7 and 8)

How's that again ? I don'tunderstand your response asworded. Please elucidate.

II. Second pass

13. Right both elements Not applicable (frame com-pleted on first pass)

14. Right Second Element(first element right)a. Reason : ease of distribution, of (Same as t.a) (E)

mailing of sending

b. Reason : cost, cheap, inex- (Same as t.b) (E)pensive, save money

c. Reason : x.a 1.b (Same as t.c) (E)

15. Right Second Element(first element wrong)a. Microcard 1st element

(i) experience (Similar to z.b.x, a.b.;

4 -r-A-

3


(z) capacity(3) clarity(4) knowledge

b. Aperture card ist element

i6. Wrong Second Element(1st element right)

17. Wrong Don't know.

18. Wrong Don't know + un-identified

19. Wrong Don't know +a. Single identifiedb. Multiple identified

zo. Wrong Don't know general

r. Wrong Rejectiona. Dislikeb. Expletive

22. Wrong Request

SUBROUTINE FOLLOWING

and 2.b.4, with special wording)(E)

(Same as i5.a) (E)

(ditto)

(Repeat 5)

(Repeat 6)

(Similar to 7)(Similar to 8)

(Repeats 9)

(Repeats io.a)(Repeats 1o.b)

(Same as 7, 8, and i)

23. Righta. « YES »

24. Wronga. «NOu

Fine. If you are in doubt aboutany of the other terms, don'thesitate to ask. Here is the originalquestion again. (repeats question)

Rather than spend more CAI timeon it now, I suggest that youeither sign off and seek out aninstructor, or type < Pass o and Iwill go to the next question. (E)

SUBROUTINE FOLLOWING 8

25. Righta. « YES »

26. Wronga. s NO

In that case we will pass to themeaning of ... (next occurringitem of requested information).

(Same as 24).

46'


27. FAIL If *12 unused on 1st (Same as la).pass

z8. FAIL If #iz used on 1st pass (Use rz (admits frame failure;may quote all instructive text)).

III. Third pass(Predictable only if a subroutine was used on the '1st or and pass)

SUBROUTINE FOLLOWING 8 and 9

z9. Wrong Continuea. « PASS a All right. But please get clarifica-

tion at the first opportunity. (E)3o. Wrong EXIT

a. « I quit » You are now signed off. (E)

3t. FAIL *27 & # 28 used (Program falls through to nextframe) (E)

A SHORT DIALOGUE AS ENVISAGED IN THE S< A » VERSION

Instructor (Statement and question)Student I don't know what a microcard is, really.Instructor (Defines microcard) Does this resolve your

uncertainty concerning the definition ofmicrocards ?

Student Yes.Instructor Fine. If you are in doubt about any of the

other terms don't hesitate to ask me. (Repeatsstatement and question).

Student Microfiche, because it is cheaper to reproduceand distribute in quantity.

Instructor You are correct in picking microfiche, andease of distribution and use is one of itscompelling advantages. Microfiche can beduplicated very rapidly and inexpensively ...(etc.)

(Exit to next frame)

4'

266 J. C. MEREDITH (.3c D. FERGUSON

A LONGER DIALOGUE AS ENVISAGED IN THE «A 0 VERSION

Instructor (Statement and question)Student I dislike all three.Instructor Aside from your personal feelings in the

matter, can you express an opinion as to whichmight be the least unsuitable for the statedpurpose ?

Student How can I ? I don't know much about anyof them.

Instructor One of them is superior in reproduceability,general use, cost, etc. Do you need more in-formation ? If so, type « Clarify D

Student O.K. Clarify what is a microcard. Also 3.Instructor I understand you need more information about

opaque microcard aperture cards. We willtake them up in that order. (Defines micro-card) Does this resolve your uncertaintyconcerning microcards ?

Student AffirmativeInstructor In that case we will pass to the meaning of

aperture card : (Defines aperture card). Doesthis resolve your uncertainty concerning aper-ture cards ?

Student YesInstructor Fine. If you are in doubt about any of the

other terms, don't hesitate to ask. Here is theoriginal question again. (Repeats statementand question).

Student The whole question seems unreasonable to me.Instructor How's that ? I don't understand your response

as worded. Please elucidate.Student Well, I'll make a guess. Microcards and aper-

ture cards.Instructor See if you can pick just one, on the basis of

what you can learn from me on the subject.Student Microcards.Instructor Why ?Student Because they're cheaper, I guess.Instructor No, as a matter of fact, they are more expensive

than either of the others. Microfiche is betterfor dissemination of technical literature than

47

(Exit to next frame).


either microcard or aperture cards, because(ctc.) ... Microfiche can be duplicated

... (etc. FICHUP).

APPENDIX E

K A » STUDENT ANSWERS AND PROGRAM RESPONSES

This Appendix lists the student answers of the « A » set,together with the reactive text displays which would have beencalled forth in the « A-Frame ». All items are grouped accordingto « sense » from the instructor's viewpoint. The several evidentinconsistencies are discussed in the body of the paper.

STUDENT INPUT REACTION TEXT DISPLAY

Group I : No clue

# 1 I don't know. I'll be glad to explain any of the termsI've used, if you wish. Just type« Clarify »

# 2 I don't know. (same as #1)

# 9 I really haven't the slightest idea. (same as # 1)

Group II : Polite regret, some narrowing of information need

# 3 I don't know enough to decide. I don't understand exactly what youneed to have clarified.

# 4 I don't know what these are exactly,so can't tell.

(same as #3)

# 7 Don't have enough experience tojudge.

I'd like you to choose on the basis ofavailable information, if possible.Type « Clarify » for explana-tion of individual elements.

4g



410

411

Not enuf info to answer this wouldneed to know properties of micro-forms and kinds/quality of availphotocopy.

I've only had occasion to use themicrofilm on reels, so I can't really say.

How's that again ? I don't understandyour response as worded. Please eluci-date.

(same as 41)

Group III : « I don't know » + attempt

4 5 I don't know. They all seem aboutequivalent to me.

(same as #3)

Group IV : Specific information need

4 6

4 8

412

I don't know what an aperture card is.Nor (1) or (2)

I can't honestly compare because I amnot familiar with all 3 of the media,

Please define opaque microcard andaperture card.

I understand you need more informa-don about aperture card - microfiche -opaque microcard. We will take themup in that order. Aperture cards arenot true unit microforms since they are« assembled », generally from cardstock, adhes'e, and film. They atewell suited to applications such asengineering drawings... (etc.) Doesthis explain the term adequately foryou ?

I understand you need more informa-don about aperture cards. Aperturecards are not true unit rnicroformssince they are « assembled »,... (etc.)Does this explain the term adequatelyfor you ?

I understand you need more informa-don about opaque microcard - aper-ture card. We will take them up in that

49



order. Microcard is a registered trade-mark for the form which is also calledthe a microopaque )), and is charac-terized by ... (etc.) Does this explainthe term adequately for you ?

Group V : Solution offered

#17

#18

#19

#20

4P21

4t22

#23

I believe the opaque microcard is mostuseful.

Transparent film sheet; just seems easyto store and use.

Apf tire cards too awkward for a bigfile, easier to enlarge microfiche forconvenient viewing.

I think the microfiche would be bestbecause I think it can contain moreinfo than the other forms.

3. The aperture card can be handledmechanically.

Aperture card, because the citationinfro (sic) is readable without resortingto a reader.

3. Aperture cards because they areprocessed by machine easily.

The opaque microcard is adequate.

The microfiche offers better reproduc-tion (visually) than the microcard.

I would suggest the microfiche, sincecomplete information can be stored ina small space.

Why ?

Why ?

Why ?

Why ?

Why ?

Why ?

Why ?

Why ?

Why ?

Why ?



*24 #2 is most suitable. As a positivecopy of the text it is least objectionableto the user.

Why ?

Group VI : Other

#16 None. I believe in hard copy. I dislikesitting in dark rooms and looking atidiot boxes.

Aside from your personal feelings inthe matter, can you express an opinionas to which form might be leastunsuitable for the stated purpose ?

APPENDIX F

(< A-FRAME » PERFORMANCE SUMMARY

I. Attached.

z. Explanation of column headings follows :

1) Student .answer number

z) Studenttnswer

3) Sense group

4) Scan fate

5) Answer type

generally in random order, except thatwhen typing the list from the answerforms, « don't know »s were entered first.

except for line breaks, a literal versionin each case.

ordered, I through VI (see Appendix Efor designations).

an identification of the particular scanspecification which is believed would haveachieved match condition, for the studentanswer given. (Note : seeming incon-sistencies can be resolved by considering(a) sequence of scans, together with (b)sequence of elements in the answers).

classification of answers according tocategories discussed on page 6.


6) First pass program reaction refers to item numbers in Appendix D.

7) Sense score an evaluation of performance (o-t) basedon the following criteria : Does the firstpass reaction make sense, and is the pros-pect of sensible second-pass reactiongood ?

8) Motivation rating assigned this is a rating intuitively assigned by theby the instructor instructor, based on the content and

overtones of the student's answer :Ma = no apparent motivationMx = polite regret, some motivationMz = fair-to-good motivationM3 = weighting factor for Group V

answers, in which good to excel-lent motivation can be assumedregardless of wording.

9) Revision required(I) + scan

(z) M-Group

decision to improve scan specification, orto scan for additional items and ideas.answers which it is expected might bebetter handled through motivationalclustering. (See pages 237-2.38)

3. Initially we tried to distinguish between sense from the programviewpoint and sense from the student's viewpoint. This raisedproblems of assigning credit for fortuitous first-pass performance,and in the assessment of partial relevance, so it was decided tocombine the two on an empirical basis. We consider that if theframe performs meaningfully for the student it is immaterialwhether or not the operation was planned in exactly the wayit takes place, and whether or not all of the expository materialone might have liked to call forth in a given situation actuallyappears. Credit is taken for situations where slight discrepanciesin meaning stand a very good chance of being corrected in thenext pass.

5 24

A-F

ram

e Pe

rfor

man

ce S

umm

ary

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

Num

ber

Ans

wer

Sens

eG

roup

Scan

Fate

Ans

wer

Typ

e

1st P

ass

Sens

ePr

ogra

mq

Rea

ctio

nco

re

Mot

iva-

lion

Rat

ing

Rev

isio

n re

quir

ed

+SC

AN

M-G

roup

# 1

I do

n't k

now

.I

Dun

noA

LPM

51

0x

# 2

I do

n't k

now

.I

Dun

noA

Lpm

5I

0x

1# 9

I re

ally

hav

en't

the

slig

htes

t ide

a.I

Dun

noA

Lpm

51

0x

#11

I've

only

had

occ

asio

n to

use

the

mic

rofi

lmon

ree

ls, s

o I

can'

t rea

lly s

ay.

Iln

Dun

noA

LU

51/

22

x

# 3

I do

n't k

now

eno

ugh

to d

ecid

e.II

Dun

no +

Gar

bA

LPM

61/

21

x#

4I

don'

t kno

w w

hat t

hese

are

exa

ctly

so

can'

t tel

l.Il

nD

unno

+ G

arb

AL

pm6

1/2

1x

# 7

Don

't ha

ve e

noug

h ex

peri

ence

to ju

dge.

IID

unno

+ G

arb

AL

pm9b

1

#10

Not

enu

f in

to to

ans

wer

this

wou

ld n

eed

to k

now

pro

pert

ies

of m

icro

form

s an

d ki

nds

/qu

ality

of

avai

l pho

toco

py

Iln

Fail

# 1

AL

U12

01

xx

# 5

I do

n't k

now

. The

y al

l see

m a

bout

equi

vale

nt to

me.

III

Dun

no +

Gar

bA

LU

61/

22

xx

#12

Plea

se d

efin

e op

aque

mic

roca

rd a

ndap

ertu

re c

ard.

IVD

unno

+ o

paqu

eA

LPM

7b1

3

# 6

In'

t kno

w w

hat a

n ap

ertu

re z

ard

is.

IVD

unno

+ap

ert.

AL

p m

7d1

3N

or (

1) o

r (2

)#

8I

can'

t hon

estly

com

pare

bec

ause

I a

m n

otfa

mili

ar w

ith a

ll 3

of th

e m

edia

.IV

Dun

no +

aper

t.A

LU

7d0

2x

x

#13

I be

lieve

the

opaq

ue m

icro

card

is m

ost

usef

iil.

VW

hy 2

AL

PM2a

13

#14

Tra

nspa

rent

film

she

et; j

ust s

eem

s ea

syto

sto

re a

nd u

se.

VW

hy 1

AL

U2a

03

x

#17

#18

#19

#20

#21

4t22

#23

#24

#25

#15

#16

I th

ink

the

mic

rofi

che

wou

ld b

e be

stbe

caus

e I

thin

k it

can

cont

ain

mor

e in

foth

an th

e ot

her

form

s.3.

The

ape

rtur

e ca

rd c

an b

e ha

ndle

dm

echa

nica

lly.

Ape

rtur

e ca

rd, b

ecau

se th

e ci

tatio

n in

fro

(sic

)is

rea

dabl

e w

ithou

t res

ortin

g to

a r

eade

r.3.

Ape

rtur

e ca

rds

beca

use

they

are

proc

esse

d by

mac

hine

eas

ily.

The

opa

que

mic

roca

rd is

ade

quat

e.T

he m

icro

fich

e of

fers

bet

ter

repr

oduc

tion

(vis

ually

) th

an th

e m

icro

card

.I

wou

ld s

ugge

st th

e m

icro

fich

e, s

ince

com

plet

e in

form

atio

n ca

n be

sto

red

ina

smal

l spa

ce.

#2 is

mos

t sui

tabl

e. A

s a

posi

tive

copy

of th

e te

xt it

is le

ast o

bjec

tiona

ble

to th

e us

er.

The

mic

rofi

che

has

prov

ed q

uite

use

ful f

orca

rtog

raph

ic in

form

atio

n.A

pert

ure

card

s to

o aw

kwar

d fo

r a

big

file

,ea

sier

to e

nlar

ge m

icro

fich

e fo

r co

nven

ient

view

ing.

Non

e. I

bel

ieve

in h

ard

copy

. I d

islik

e si

tting

in d

ark

room

s an

d lo

okin

g at

idio

t box

es.

RE

CA

PIT

UL

AT

ION

:

Ove

rall

mat

ch s

core

24 (

96 %

)Se

nse

Scor

e12

(48

%)

Mot

ivat

ion

0(4

)W

eigh

t :0

Mot

ivat

ion

1(4

)4

Mot

ivat

ion

2(3

)6

Mot

ivat

ion

3(1

4)42

Mot

ivat

ion

com

posi

te52

Mot

ivat

ion

poss

ible

max

.75

Mot

ivat

ion

fact

or52

/75

VW

hy 1

AL

T T

2a0

3x

VW

hy 3

AL

U4b

03

x

VW

hy 3

AL

U4b

03

x

VW

hy 3

AL

U4b

03

x

VW

hy 2

AL

P M

4a1

3V

Why

1A

LU

2a0

3

VW

hy 1

AL

U2a

03

x

VW

hy 2

AL

U4a

03

x

VW

hy 1

AL

U2a

13

VW

hy 3

AL

U4b

03

x (s

peci

al)

VI

RE

JEC

TA

Lp

m10

a1

0

Gro

up V

res

pons

es12

(48

%)

All

othe

rs13

(52

%)

Scan

impr

ovem

ent r

equi

red

13 (

52 %

)fo

r «

B-F

ram

e »


APPENDIX G

B-FRAME » FLOWCHART

I. Attached.

z. The attached flowchart reflects program changes incorporatedin the « B-Frame >+. Appendix B notes apply (p. 246-248), exceptas provided below :

5) (revised) The group of operations to the left of « DUNNO » is devotedto the provision of information intended to help the student solve theproblem. If no specific information need is detected, his response istested for motivational indicators (Mz, Mi, & Ma) in three separatescans whose specifications are given in Appendix H. The Q-returndisplay for each of these differs from the other two in the way it ap-proaches the problem. Thc. general « I don't know » + GARB, and« I don't know » scans are retained in order to catch all such answersthat escape earlier match.

9) (new) Satellites are increased from five to six in number, and a slightstructural modification is provided : Instead of a single block of textending with « Does this explain the term adequately for you ? » adifferently worded ending is used for students who have been assessed« MI » and are first being reinforced on something they may have someknowledge of : «I hope this confirms your understanding of ... (arWhat would you like to follow up next ? »

55

r

00

00

CA

LL

VS

SC

L

T.. I

Tos N

IKE

Mto

kr

GA

LLA

l KR

...

5AT

EL

LiT

E "

UN

IT 2

:

Ti, C

ALL

0,4X 0

Y.I MC

RBIT lilt'

eXt

Der

enc

CA

L

To

T -

00

eAL

FA

L A

L'

CA

LLio

di X

"

4111

$*41

1* Ter

Q (

DT

ftss

l 4.11

1Itz

. ti

0

Ves

FA

:L. I

(L e

Lx.

)

B-F

ram

e Fl

owch

art

Y. 0

4.*

Wit

larL

AA

Yes

00

In 0


Yn

B-Frame Flowchart (Continuation)

5?


APPENDIX H

THE (< B-FRAME )) PROGRAM

277

I. Attached.

2. The notes to Appendix C apply (p. 248-249) except as providedbelow :

8) (omitted)

9) (new) A word partially underlined, in a scan specification, means thatonly the underlined portion is stipulated as a match requirement. Thecharacters which are not underlined are included only to show thereader what term group the programmer has in mind. For example« STORAGE » will accept « store », « stored », «storing», and «sto-rage» but not «stop». It will also accept the name of the bird, but theprogrammer is willing to take the risk.

3. It should not be imagined that a CAI programmer wouldoften invest in a single frame the amount of work which hasclearly gone into both the A-Frame and the B-Frame. As a stand-alone unit we had to provide it with features which normallywould need to be written only once far an entire course, e.g.,the « POSIT » and « NEGAT » expansions, and the M-seriesof scan specifications. Also it should be mentioned that thecommand statements can be put together quite rapidly after onehas worked with a language for a while. As a matter of record,one person designed and wrote the « A-Frame » in a little overfour hours.

58

<4

B »

Sou

rce

Prog

ram

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

EX

POT

(Ini

tial s

tate

men

t and

que

stio

n)

MIC

RO

.AA

(stu

dent

inpu

t)If

PU

TA

AD

D D

ESC

RIB

E T

O«

MIC

RO

.A »

, CL

EA

R P

UT

A

R(«

DU

NN

O s

, CL

AR

IFY

, IN

FOR

MA

TIO

N, D

ESC

RIB

E)

+T

AL

LY

(U

NIT

Y, P

RO

BY

), I

F(U

NIT

, UN

ITS,

UN

ITIZ

ED

UN

ITY

2 A

DD

UN

ITX

TO

..>...

DU

PY

R(«

DU

NN

O »

CL

AR

IFY

, IN

FOR

MA

TIO

N, D

ESC

RIB

E)

+T

AL

LY

( M

ICR

OY

, PR

OB

Y),

IF

(TR

AN

SPA

RE

NT

, FIL

M, S

HE

ET

, FIC

HE

, MIC

RO

FIC

HE

,FI

LM

S, «

N.1

»)

MIC

RO

Y >

2 A

DD

MIC

RO

X T

OD

UPY

R(«

DU

NN

O »

CL

AR

IFY

, IN

FOR

MA

TIO

N, D

ESC

RIB

E)

+T

AL

LY

(O

PAQ

Y, P

RO

BY

), I

F(O

PAQ

UE

, AIC

RO

CA

RD

, «N

.2 »

OPA

QY

_>

... 2

AD

D O

PAQ

X T

O D

UPY

R(«

DU

NN

O »

, CL

AR

IFY

, IN

FOR

MA

TIO

N, D

ESC

RIB

E)

+T

AL

LY

(A

PER

Y, P

RO

BY

), I

F(A

PER

TU

RE

, «N

.3 »

)A

PER

Y >

2 A

DD

APE

RX

TO

DU

PY

R(«

DU

NN

O »

, CL

AR

IFY

, IN

FOR

MA

TIO

N, D

ESC

RIB

E)

±T

AL

LY

(D

ISSE

Y, P

RO

BY

), I

F(D

ISSE

MIN

AT

ION

, TE

CH

NIC

AL

, LIT

ER

AT

UR

E)

DIS

SEY

2 A

DD

DIS

SEX

TO

_.-.

DU

PY; I

F (U

NIT

Y +

MIC

RO

Y ±

OPA

QY

+ A

PER

Y +

DIS

SEY

= 2

USE

DU

P, A

DD

« U

NIT

X »

TO

PR

OB

ON

UN

ITY

, AD

D «

OPA

QX

P T

OPR

OB

ON

OPA

QY

, AD

D «

APE

RX

»T

O P

RO

B O

N A

PER

Y, A

DD

« D

ISSE

X »

TO

PR

OB

ON

DIS

SEY

Lab

elC

ode

G R G R G R G R G R G

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

I U

ND

ER

STA

ND

YO

U N

EE

D M

OR

E I

NFO

RM

AT

ION

AB

OU

T «

PR

OB

»

« D

UN

NO

» +

(R

EE

L, R

EE

LS,

RO

LL

, RO

LL

S)

AS

TO

RE

EL

MIC

RO

FOR

M V

S. S

HE

ET

MIC

RO

FOR

MS

«M2

»

IN O

RD

ER

TO

EN

AB

LE

YO

U T

O D

ISC

RIM

INA

TE

... (

etc.

)...

TY

PE «

IN

FOR

MA

TIO

N

« M

I »

WE

LL

, SU

PPO

SE W

E S

TA

RT

BY

EX

AM

ININ

G...

(et

c.)

...A

RE

YO

U A

CQ

UA

INT

ED

WIT

H A

NY

OF

TH

ESE

?

« M

O »

I R

EA

LIZ

E I

T I

S A

LM

OST

IM

POSS

IBL

E...

(et

c.)

AR

E Y

OU

WIL

LIN

G T

O W

OR

K W

ITH

ME

TO

WA

RD

AC

OR

RE

CT

SO

LU

TIO

N ?

«DU

NN

O »

(A

LL

, AN

Y)

ON

E O

F T

HE

M I

S SU

PER

IOR

IN

RE

PRO

DU

CE

AB

ILIT

Y,

CO

ST, G

EN

ER

AL

USE

, ET

C. D

O Y

OU

NE

ED

MO

RE

INFO

RM

AT

ION

? I

F SO

, TY

PE «

CL

AR

IFY

a

IF P

RO

BY

> 2

USE

TA

KU

P,C

AL

L U

NIT

Z O

N U

NIT

Y, C

AL

LM

ICR

OZ

ON

MIC

RO

Y, C

AL

L O

PAQ

ZO

N O

PAQ

Y, C

AL

L A

PER

Z O

NA

PER

Y, C

AL

L D

ISSE

Z O

N D

ISSE

Y,

JUM

P M

ICR

O.A

USE

RE

EL

Z, J

UM

P T

O M

ICR

O.A

MA

RK

(M

OT

C, P

UT

A),

JU

MP

TO

MIC

RO

.A

MA

RK

(M

OT

B, P

UT

A),

JU

MP

TO

MIC

RO

.A

MA

RK

MO

TA

, JU

MP

MIC

RO

.A

JUM

P T

O M

ICR

O.A

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

R«

DU

NN

O »

« G

AR

B »

GI

DO

N'T

UN

DE

RST

AN

D W

HA

T Y

OU

NE

ED

TO

HA

VE

JUM

P T

O M

ICR

O.A

CL

AR

IFIE

D.

RN

OT

CO

MPE

TE

NT

, IN

CO

MPE

TE

NT

, IL

L E

QU

IPPE

D, N

OT

PRE

PAR

ED

, NO

T E

QU

IPPE

D, I

LL

PR

EPA

RE

D

GI'L

L B

E G

LA

D T

O E

XPL

AIN

AN

Y O

F T

HE

TE

RM

S I'V

EJU

MP

TO

MIC

RO

.AU

SED

, IF

YO

U W

ISH

. JU

ST T

YPE

« C

LA

RIF

YD

RPO

SIT

IVE

, NE

GA

TIV

E

GM

ICR

OC

AR

D I

S A

LW

AY

S D

IST

RIB

UT

ED

AS

A P

OSI

TIV

E :

JUM

P T

O M

ICR

O.A

MIC

RO

FIC

HE

S A

ND

APE

RT

UR

E C

AR

DS

MA

Y B

E H

AD

IN

EIT

HE

R P

OSI

TIV

E O

R N

EG

AT

IVE

VE

RSI

ON

S. A

SID

EFR

OM

TH

IS C

ON

SID

ER

AT

ION

, WH

AT

WO

UL

D B

E Y

OU

RC

HO

ICE

?

R(«

N.I

», «

N.2

», «

N.3

»)

-1-

(NO

T, I

NSU

FFIC

IEN

TL

Y,

INA

DE

QU

AT

E, T

OO

, OV

ER

, OV

ER

-, U

NA

CC

EPT

AB

LE

,E

XC

ESS

IVE

, UN

DU

E, U

ND

UL

Y)

+ $

, (no

t inv

erte

d)(N

ote

: $ s

igni

fies

lite

ral)

GPL

EA

SE R

EPH

RA

SE Y

OU

R R

EPL

Y, S

TA

TIN

G T

HE

PR

E-

JUM

P T

O M

ICR

O.A

FER

RE

D T

YPE

FIR

ST

R«N

.1 »

, « N

.2 »

, « N

.3 »

(re

quir

e 2,

any

ord

er)

RA

LL

, AN

Y

GSE

E I

F Y

OU

CA

N P

ICK

JU

ST O

NE

, ON

TH

E B

ASI

S O

FJU

MP

TO

MIC

RO

.AW

HA

T Y

OU

CA

N L

EA

RN

FR

OM

ME

ON

TH

E S

UB

JEC

T.

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

MIC

RO

CA

PR

« N

.1 »

« P

LU

SCA

P »,

« P

LU

SCA

P »

« N

.1 »

MIC

RO

CA

PG

YO

U A

RE

CO

RR

EC

T I

N P

ICK

ING

MIC

RO

FIC

HE

, BU

T T

HE

JUM

P T

O N

EX

TFR

AM

ER

EA

SON

IT

IS

SUPE

RIO

R F

OR

TH

E D

ISSE

MIN

AT

ION

OF

TE

CH

NIC

AL

LIT

ER

AT

UR

E I

S N

OT

TH

AT

IT

HA

S G

RE

AT

ER

CA

PAC

ITY

(W

HIC

H D

EPE

ND

S O

N T

HE

RE

DU

CT

ION

RA

TIO

(USE

D)

BU

T O

N T

HE

FA

CT

TH

AT

e F

ICH

UP

a

MIC

RO

EA

SER

«N.1

» «

PLU

SEA

SE »

, « P

LU

SEA

SE »

« N

.1 »

MIC

RO

EA

SEG

YO

U A

RE

CO

RR

EC

T I

N P

ICK

ING

MIC

RO

FIC

HE

, AN

D E

ASE

JUM

P T

O N

EX

TFR

AM

EO

F D

IST

RIB

UT

ION

AN

D U

SE C

ER

TA

INL

Y D

OE

S FA

VO

R I

TA

S A

ME

DIU

M F

OR

TH

E D

ISSE

MIN

AT

ION

OF

TE

CH

NIC

AL

LIT

ER

AT

UR

E. «

FIC

HU

P »

R«

N.I

» «

PL

USC

HE

AP

», «

PL

USC

HE

AP

» «

N.1

MIC

RO

CH

EA

PG

YO

U A

RE

CO

RR

EC

T I

N P

ICK

ING

MIC

RO

FIC

HE

, AN

D C

OST

JUM

P T

O N

EX

TFR

AM

E

R

CO

NSI

DE

RA

TIO

NS

AR

E A

N I

MPO

RT

AN

T F

AC

TO

R. «

FIC

HU

P »

e N

.1 »

« P

LU

SEA

SE a

« P

LU

SC.H

EA

P a

(any

ord

er)

GG

OO

D. Y

OU

CO

RR

EC

TL

Y P

ICK

ED

TW

O O

F T

HE

IM

POR

TA

NT

JUM

P T

O N

EX

TFR

AM

EA

DV

AN

TA

GE

S O

F M

ICR

OFI

CH

E. «

FIC

HU

P a

R«N

.1 »

«PL

USC

IT a

(JU

MP

TO

NE

XT

FRA

ME

OR

TO

R«

N.1

» «

PL

USC

OP

»M

ICR

O.A

, DE

PEN

DIN

G O

NR

N.1

» o

PL

USC

ON

V »

(sim

ilar

disc

ussi

on o

f ty

pes

and

thei

r as

pect

s)SE

NSE

)R R

« N

.2 o

« P

LU

SCA

P »

o N

.2 »

a P

LU

SCH

EA

P a

etc.

R«

PLU

SEX

ER

»

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

GC

AN

YO

U C

OM

PAR

E T

HE

M, N

OT

ON

TH

E B

ASI

S O

FJU

MP

TO

MIC

RO

.AE

XPE

RIE

NC

E, B

UT

MO

RE

ON

TH

E B

ASI

S O

F IN

FOR

-M

AT

ION

? IF

YO

U W

OU

LD

LIK

E T

O K

NO

W M

OR

E A

BO

UT

TH

E O

TH

ER

FO

RM

S, O

R A

BO

UT

TH

E G

EN

ER

AL

RE

QU

IRE

-M

EN

TS

FOR

DIS

SEM

INA

TIO

N O

F T

EC

HN

ICA

L L

ITE

R-

AT

UR

E, J

UST

TY

PE «

CL

AR

IFY

a

RR

EJE

CT

, HA

TE

, HA

TE

D, D

ISL

IKE

, « N

EG

AT

» L

IKE

D,

«NE

GA

T »

LIK

E, «

NE

GA

T a

CA

RE

FO

R

GA

SID

E F

RO

M Y

OU

R P

ER

SON

AL

FE

EL

ING

S IN

TH

E M

AT

TE

RJU

MP

MIC

RO

.AC

AN

YO

U E

XPR

ESS

AN

OPI

NIO

N A

S T

O W

HIC

H M

IGH

TB

E L

EA

ST U

NSU

ITA

BL

E F

OR

TH

E S

TA

TE

D P

UR

POSE

?

R«

POSI

T »

IF M

OT

A M

AR

K P

UT

A,

IF M

OT

A J

UM

P M

OT

AA

GW

HIC

H ;

IF M

OT

B M

AR

K P

UT

A,

IF M

OT

B J

UM

P M

ICR

O.A

MO

TA

AG

WH

ICH

ON

E W

OU

LD

YO

U L

IKE

ME

TO

DE

SCR

IBE

FIR

ST ?

TY

PE «

DE

SCR

IBE

aJU

MP

MIC

RO

.A

MO

TB

BG

SHA

LL

WE

TA

KE

TH

EM

UP

ON

E A

T A

TIM

E ?

SU

PPO

SEW

E S

TA

RT

WIT

H T

HE

PR

OB

LE

M O

F D

ISSE

MIN

AT

ION

OF

TE

CH

NIC

AL

LIT

ER

AT

UR

E. F

OL

LO

WIN

G T

HA

T, Y

OU

CA

NA

SK F

OR

IN

FOR

MA

TIO

N O

N O

TH

ER

EL

EM

EN

TS

OF

TH

EPR

OB

LE

M.

R«

NE

GA

T »

IF M

OT

B U

SE M

OT

BB

,C

AL

L D

ISSE

Z, J

UM

P M

ICR

O.A

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

GPE

RH

APS

YO

U S

HO

UL

D S

IGN

OFF

, TH

EN

.JU

MP

TO

SIG

NO

FF

R«

N.I

»

GW

HY

?T

AL

LY

(A

NSY

, MIC

RO

B),

JUM

P M

ICR

O.A

R«

N.2

»

GW

HY

?T

AL

LY

(A

NSY

, OPA

QB

),JU

MP

MIC

RO

.A

R«

N.3

D

GW

HY

?T

AL

LY

(A

NSY

, APE

RB

),J'

JMP

MIC

RO

.A

Re

PLU

SCA

P »,

« N

EG

AT

» e

NE

GC

AP

oIF

MIC

RO

B J

UM

P M

ICR

OC

AP,

IF O

PAQ

B J

UM

P O

PAQ

CA

P,IF

APE

RB

JU

MP

APE

RC

AP

R«

PLU

SCO

NV

», «

NE

GA

T »

e N

EG

CO

NV

aIF

MIC

RO

B J

UM

P M

ICR

OC

ON

V,

IF O

PAQ

B J

UM

P O

PAQ

CO

NV

,IF

APE

RB

JU

MP

APE

RC

ON

V

R«

PLU

SCIT

»(s

imila

r to

abo

ve)

R«

PLU

SCL

AR

»R

e PL

USC

OP

»R

o PL

USD

IST

»(s

imila

r do

uble

neg

ativ

es)

Re

PLU

SME

CH

»R

« PL

USE

XP

»

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

RN

OT

ES

(sim

ilar

to p

rece

ding

)R

LO

SE, L

OST

, MIS

PLA

CE

D(s

imila

r to

pre

cedi

ng)

GD

ID Y

OU

SPE

CIF

Y O

NE

OF

TH

ET

YPE

S ?

JUM

P T

O M

ICR

O.A

FH

OW

'S T

HA

T A

GA

IN?

I D

ON

'TU

ND

ER

STA

ND

YO

UR

IJU

MP

TO

MIC

RO

.AR

ESP

ON

SE A

S W

OR

DE

D

FW

EL

L, I

'M L

ICK

ED

. GO

OD

BY

.JU

MP

TO

SIG

NO

FFU

NIT

ZT

UN

IT (

OR

UN

ITIZ

ED

) M

ICR

OFO

RM

MIC

RO

XT

MIC

RO

FIC

HE

OPA

QX

TO

PAQ

UE

MIC

RO

CA

RD

APE

RX

TA

PER

TU

RE

CA

RD

DIS

SEX

TD

ISSE

MIN

AT

ION

OF

TE

CH

NIC

AL

LIT

ER

AT

UR

E

DU

PT

I B

FLIF

VE

YO

U A

LR

EA

DY

ASK

ED

AB

OU

T «

DU

PY »

, AN

DI

TR

IED

TO

EX

PLA

IN. I

F Y

OU

AR

EST

ILL

AT

A L

OSS

,PE

RH

APS

YO

U S

HO

UL

D S

IGN

OFF

AN

D S

EE

Y: H

EL

P.T

AK

UP

TW

E W

ILL

TA

KE

TH

EM

UP

IN T

HA

TO

RD

ER

.R

PTT

HE

RE

IS

TH

E O

RIG

INA

LQ

UE

STIO

N A

GA

IN :

« E

XPO

aM

0T

« A

l », (

« A

2 »

+ «

BI

» +

« C

I ».

!

MI

T(«

Al »

, «A

2 »)

+ «

D3

», «

El »

+ U

NL

ESS

+ (

« C

2 »,

«A

3 »)

M2

T«D

I »

+ «

C2

», «

D2

»A

lT

DO

N'T

KN

OW

, CA

N'T

SA

Y, C

AN

'TIM

AG

INE

, YO

U E

XPE

CT

,Y

OU

TH

INK

, YO

U I

MA

GIN

E, A

TSE

A, W

ITH

IT

, SE

AR

CH

ME

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

A2

TD

ON

'T H

AV

E, N

OT

HA

VE

, HA

VE

N'T

, EX

PEC

T H

AV

ET

HIN

K H

AV

E

A3

TA

BSE

NC

E, W

ITH

OU

T, L

AC

KIN

G, E

VE

R, N

EV

ER

, EN

OU

GH

,E

NT

JF, S

UFF

ICIE

NT

, IN

SUFF

ICIE

NT

, IN

AD

EQ

UA

TE

B1

TL

EA

ST, V

AG

UE

. SL

IGH

TE

ST, F

UZ

ZIE

ST, F

OG

GIE

ST,

RE

MO

TE

ST

Cl

TID

EA

, NO

TIO

N, I

NT

ER

EST

, CO

NC

EPT

, IN

KL

ING

C2

TE

XPL

AN

AT

ION

, DE

FIN

ITIO

N, D

EFI

NE

, CL

AR

IFIC

AT

ION

EL

UC

IDA

TIO

N, E

XPL

ICA

TIO

N, I

NFO

RM

AT

ION

DI

TN

EE

D, 1

1EQ

UIR

E, H

EL

P

D2

TD

EFI

NE

, FU

RN

ISH

, GIV

E, E

XPL

AIN

, TE

LL

, SU

PPL

Y,

AC

QU

AIN

T, E

LU

CID

AT

E, E

XPL

ICA

TE

D3

TK

NO

WL

ED

GE

, FA

CT

S, I

NFO

RM

AT

ION

, DA

TA

, EX

PER

TIS

E,

INC

OM

PET

EN

CE

El

TN

OT

AB

LE

, UN

AB

LE

, CA

N'T

, CA

NN

OT

, CA

N N

OT

DU

NN

OT

DO

N'T

KN

OW

, CO

NFU

SED

, CA

N'T

DE

CID

E, C

AN

'T C

HO

OSE

,U

ND

ER

STA

ND

, UN

AC

QU

AIN

TE

D, M

AK

E C

HO

ICE

, NO

T S

UR

E,

UN

SUR

E, D

ON

'T U

ND

ER

STA

ND

, WH

AT

GA

RB

T(a

lpha

bet)

POSI

TT

(for

ms

of a

ffir

mat

ion,

exc

ludi

ng w

ord

« PO

SIT

IVE

»)

NE

GA

TT

(for

ms

of n

egat

ion,

exd

udia

g w

ord

«NE

GA

TIV

E »

)

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

PLU

SCA

PT

CA

PAC

ITY

, SIZ

E, V

OL

UM

E, S

PAC

E, C

ON

TA

IN, H

OL

D, H

OL

DS,

HE

LD

, MO

RE

IN

FOR

MA

TIO

N, S

TO

RE

, CO

NV

EY

, CO

NV

EY

S

PLU

SCIT

TT

ITL

E, H

EA

DIN

G, C

ITA

TIO

N, M

EN

/11-

ICA

TIO

N, B

IBL

IOG

RA

-PH

ICA

L, R

EFE

RE

NC

ES

PLU

SCL

AR

TC

LE

AR

, VIE

IFIN

G, R

ESO

LU

TIO

N, R

EA

DE

R, D

ISPL

AY

, SC

RE

EN

PLU

SCO

PT

CO

PY, C

OPI

ES,

CO

PYIN

G, B

LO

W U

P, B

LO

W B

AC

K, B

LO

WB

AC

KE

NL

AR

GE

ME

NT

, RE

EN

LA

RG

EM

EN

T

PLU

SDIS

TT

EA

SY (

DIS

TR

IBU

TE

, MA

ILIN

G, S

EN

DIN

G, D

ISSE

MIN

AT

ION

)

PLU

SME

CH

TM

AC

HIN

E, A

UT

OM

AT

ION

, ME

CH

AN

ICA

L

PLU

SEX

PER

TI

HA

VE

USE

, I K

NO

W A

BO

UT

, I E

XPE

RIE

NC

E

PLU

SCH

EA

PT

CH

EA

PER

, CO

ST, C

OST

S, E

CO

NO

MIC

AL

PLU

SCO

NV

TC

ON

VE

NIE

NT

, « N

EG

AT

» D

EL

AY

, QU

ICK

LY

, SPE

ED

, FA

STE

Ra

NE

GA

T »

TR

OU

BL

ESO

ME

PLU

SSO

MP

TR

OO

M, C

OM

PAC

T, S

TO

RA

GE

, SPA

CE

NE

GC

LE

AR

TD

IFFI

CU

LT

RE

AD

, PO

OR

RE

SOL

UT

ION

, BL

UR

RE

D, D

IM, D

AR

K,

FUZ

ZY

, HU

RT

S E

YE

S, I

LL

EG

IBL

E, P

RIN

T S

MA

LL

, SM

AL

L P

RIN

T

NE

GL

OSE

TL

OSE

, LO

ST, M

ISPL

AC

ED

, TO

O S

MA

LL

NE

GC

ON

VT

INC

ON

VE

NIE

NT

, AW

KW

AR

D, T

IME

CO

NSU

MIN

G, D

EL

AY

,T

RO

UB

LE

SOM

E

N.1

T1,

ON

E, F

IRST

, MIC

RO

FIC

HE

, FIC

HE

, TR

AN

SPA

RE

NT

, FIL

MSH

EE

T

,,,N

ripr

. .1*

.rvi

rrtg

7":7

^

6%-

oc)

Lab

elC

ode

Dis

play

text

and

inpu

t ter

m s

peci

fica

tions

Con

trol

sta

tem

ents

N.2

T2,

TW

O, S

EC

ON

D, O

PAQ

UE

, MIC

RO

CA

RD

N.3

T3,

TH

RE

E, «

NE

GA

T »

AL

L (

3, T

HR

EE

), T

HIR

D, L

AST

TR

AN

SFIC

HE

TT

RA

NSP

AR

EN

T (

SAM

E, D

thE

bRE

NT

) FI

CH

E, F

ICH

E (

SAM

E,

DIF

FER

EN

T)

TR

AN

SPA

RE

NT

MO

TS

TI

HO

PE T

HIS

CO

NFI

RM

S Y

OU

R U

ND

ER

STA

ND

ING

OF

« M

ICR

O.A

».

WH

AT

WO

UL

D Y

OU

LIK

E T

O F

OL

LO

W U

P N

EX

T ?

AL

TE

RE

XPO

TW

HA

T I

ME

AN

IS

: OF

TH

E T

HR

EE

CO

MM

ON

EST

TY

PES

OF

NO

N-R

OL

L M

ICR

OFO

RM

, WH

ICH

LE

ND

S IT

SEL

F B

EST

TO

TH

ED

ISSE

MIN

AT

ION

OF

TE

CH

NIC

AL

LIT

ER

AT

UR

E ?

FIC

HE

?M

ICR

O O

PAQ

UE

S ?

APE

RT

UR

E C

AR

DS

? B

EA

R I

N M

IND

CO

ST A

ND

EA

SE O

F M

AK

ING

AN

D D

IST

RIB

UT

ING

CO

PIE

S.

288 . C, MEREDITH HL D. FERGUSON

APPENDIX I

REACTION TEXT DISPLAYS - < B-FRAME >>

I. List attached.

z. Appendix D comments (p. 259) apply, with the exceptionof para. 6 (sample dialogues omitted).

I. First passType of answer (numbers and lettersestablish categories referred to inAppendix K).

1. Right both elementsa. Reason : ease of distribution, of

mailing, of sending.

b. Reason : cost, cheap, inex-pensive, save money.

c. Reason : z.a t.b

2. Right first element only

a. No reason givenb. Wrong reason given

(z) experience (« because hadmore experience with it o).

C99

Program display reaction(«E» = « exit to next frame »)

You are correct in pickingmicrofiche, and ease of distribu-tion and use is one of its com-pelling advantages. Microfichecan be duplicated very readilyand inexpensively... (etc.) ...(same as corresponding« FICHUP » text in AppendixD) (E)You are correct in pickingmicrofiche, and cost considera-tions certainly enter into thepicture. Microfiche can be du-plicated ... (etc., same asabove). (E )

Good. You correctly pickedtwo of the important advan-tages of fiche. Microfiche canbe duplicated... (etc.; same asabove). (E)

Why ?

Can you compare them moreon the basis of information ? Ifyou would like to know moreabout the other forms, or


(z) capacity

(3)(4)

(5)

clarityknowledge (« the only oneI know about »)(n) (other aspects providedin the program)

3. Wrong Two first elements

4. Wrong Wrong first element

(a) Microcard

(b) Aperture card

5. Wrong Don't knowa. « don't know » with no addi-

tional element following, andno match on motivational in-dicators.

6. Wrong Don't know + (un-identified matter)

a. « don't know » with addedmaterial, but no match onmotivational indicators

b. Ms

O

about the general requirementfor dissemination of tochnicalliterature, just type « Clarify

You aw correct in pickingmicrofiche, but the reason it issuperior for dissemination oftechnical literature is not that ithas greater capacity (whichdepends on the reduction ratioin use) but on the fact thatmicrofiche can be duplicated... (etc.) (E)(similar to 1.b.z above) (E)(similar to z.b.z above) (E)

(similar to z.b.1 above, exceptthat some will exit from theframe while others will requirerecursion).

See if you can pick just one, onthe basis of what you can learnfrom me on the subject.

Why ?

Why ?

I'll be glad to explain any of theterms I've used, if you wish.Just type « Clarify >>

I don't understand exactlywhat you need to have clarified.

I realize it is almost impossiblefor you to answer withoutknowing something about thetypes of microform mentioned


c. Mt

7. Wrong Don't know + singleidentified

and about the factors involvedin dissemination of technicalliterattre. I'd like to help youtackle the problem from thebeginning by supplying all theinformation you need. Are youwilling to work with me towarda correct solution ?Well, suppose we start byexamining the items you mayalready have heard about, I maybe able to add some usefulinformation about them. Howabout microfiche ? Microcard ?Aperture card ? The dissemina-tion of technical literature ?Are you acquainted with anyof these ?

a. Unit, unitized I understand you need more in-formation about unit, or uni-tized, microforms. A unit mi-croform is one which... (L tc.,same as corresponding text inAppendix D)

b. Transparent, film, sheet, fiche, (similar to 7.a)microfiche

c. Opaque, microcard (similar to 7.a)

d. Aperture (similar to 7.a)

e. Dissemination, technical (similar to 7.a)literature

8. Wrong Don't know + multipleidentifieda. Any combination of 7.a, 7.b, I understand you need more in-

7.c, 7.d, 7.e formation about (7.a, 7.b, 7.c,7.d, 7.e). We will take them upin that order. (Stacked sub-routines of 7.a-.e)

9. Wrong Don't know + misc-ellaneous

71


a. Any, all One of them is superior inreproduceability, cost, generaluse, etc. Do you need moreinformation ? If so, type « Clar-ify » for explanationof individual elements.

b. No experience I'd like you to choose not onthe basis of experience but onthe basis of available informa-don, if possible. Type « Clar-ify » for explanationof individual elements.

c. Positive, negative Microcard is always distributedas a positive; microfiches andaperture cards may be had ineither positive or negativeversions. Aside from this con-sideration, what would be yourchoice ?

d. Not competent, ill equipped, I'll be glad to explain any of thenot prepared, not equipped terms I've used, if you wish.

Just type « Clarify

Jo. Wrong Rejection

a. Dislike

b. Expletive.

Aside from your personal feel-ings in the matter, can youexpress an opinion as to whichmight be least unsuitable for thestated purpose ?

(same as corresponding text inAppendix D. No evidence yetas to how this might work).

I1. Wrong Request

a. Clarify, define, info, describe (same as 7 and 8. The variationbetween request words for Meand Mi is provided to facilitatechanneling of operations calledforth by these terms, and forscoring if and when incor-porated).

1. Fail How's that again ? I don'tunderstand your response as


IL Seem' pass

13. Right Both elements

14. Right Second Element, firs,:element right

a. Reason : ease of distribution, ofmailing, of sending

b. Reason : cost, cheap, inex-pensive, save money

c. Reason : r.a .b

worded. Please explain morefully.

(not applicable. (Frame com-pleted on first pass)).

(same as i.a) (E)

(same as r.b) (E)

(same as x .c) (E)

15. Right Second Element first element (same as correspondingwrong tions in Appendix D)

x6. Wrong Second Element, first ele-ment right.

17. Wrong Don't know

x8. Wrong Don't know Un-identified

18. a. Wrong Don't know + singleidentified

19. Wrong Don't know + multipleidentified

zo. Wrong Don't know + misc-ellaneous

z1. Wrong Rejection

a. Dislike

zz. Wrong Request

75

(ditto)

Repeats 5.

Repeats 6.

Similar to 7.

Similar to 8.

Repeats 9.

Repeats x o.a

(same as 7, 8 and it)

reac-



23. Right

a. « yes

24. Wronga. a. « no »

Fine, if yon are in doubt aboutany of the other terms, don'thesitate to e.sk. Mere is the originalquestion again. (Repeats question).

Rather than spend more CAItime on it now, I suggest that youeither sign off and seek out aninstructor, or type « PASS » andI will go on to the next question.(E)


2 5 . Right

a. « yes » In that case we will pass to themeaning of (next occurring itemof requested information).

z6. Wronga. « no » (same as 24)

SUBROUTINE FOLLOWING 6b (Handled within main frame)

27. Righta. « yes » alone Which one would you like me to

describe first ? Type « Describe

b. « yes » with item, I understand you need more in-or item alone formation about... (item)... (etc.)

z8. Wronga. « no a Perhaps you should sign off, then.

74


SUBROUTINE FOLLOWING 6c (Handled within main frame)

29. Righta. « yes » alone

b. « yes s with item,or item alone

3o. Wrong.a. «no»

Which ?

« DIJNNO » is automatically in-setted in rtst,nnse, so that pro-gram goes direct to satellite

Shall we take them up one at atime ? Suppose we start with theproblem of dissemination of tech-nical literature. Following that,you can ask for other elements ofthe problem.

31. FAIL If unused on 1st pass

32. FAIL Second instance

III. Third pass

(same as 12, above)

Admit frame failure.

(Predictable only if a subroutineis used on the 1st or znd pass).

SUBROUTINE FOLLOWING 8 and 9

33. Wrong continuea. « PASS »

34. Wrong Exit

a. « I quit »

All right. But please get clari-fication at the first opportunity.(E)

You ate now signed off. (E)

35. FAIL If #31 and #32 havebeen used, program falls throughto the next frame.

STUDENT FeEDBACK IN CAI 295

APPENDIX J

B » STUDENT ANSWERS AND PROGRAM RESPONSES

This Appendix lists the student answers of the « B » set,together with the reactive text which would have been calledforth in the « B-Frame ».


# 1 I don't even know what a micro- I understand you need more informa-fiche is. tion about microfiche - (test of

MICROZ satellite) - Does this explainthe term adequately for you ?

# 2 I don't know. I realize it is almost impossible for youto answer without knowing somethingabout the types of microforms men-tioned and about the factors involvedin dissemination of technical literature.I'd like to help you tackle the problemfrom the beginning by supplying allthe information you need. Are youwilling to work with me toward acorrect solution ?

# 3 I am not familiar with programming. I don't understand what you need tohave clarified.

# 4 I have not had any experience with Can you compare them not on theany, so cannot decide. basis of experience but on the basis of

information ? If you would like toknow more about the general require-ments for dissemination of technicalliterature, just type a Clarify a

# 5 I am not sufficiently acquainted with Well, suppose we start by examiningthe characteristics of these different the items you may already have heardmicroforms to answer. about. I may be able to add some

useful information about them. Howabout microfiche ? Microcard ? Aper-ture card ? The dissemination oftechnical literature? Are you acquaint-ed with any of these ?

'7



# 6 Lack of knowledge prevents a con-sidered answer.

(Same as #5)

7 What's an aperture card. To tell the I understand you need more informa-truth, 1 strongly dislike microforms ton about apri.cure cards. - (text ofand readers. APERZ satellite) - Does this explain

the term adequately for you ?

# 8 Microfiche, since it is easy to handle,store, and reproduce.

You are correct in picking microfiche,and ease of storing is certainly an ad-vantage, but one which is also sharedby the other two forms. Microfiche canbe duplicated very readily and in-expensively, and in fact is the standardused by the Federal government fordistribution of Clearinghouse, DOD,NASA, AEC, and ERIC documents,which include a large portion of ourtechnical literature output... (etc.)

I don't understand all the distinctions One of them is superior in repro-among the three types. duceability, cost, general use, etc. Do

you need more information ? If so,type « Clarify ---»

#10 1. Because I don't quite know what the See if you can pick just one, on theother two are I basis of what you can learn from me

on the subject.

#11 1. as it contains more information(data) readily & quickly available,

You are correct in picking microfiche,but the reason it is superior for thedissemination of technical literature isnot that it has greater capacity (whichdepends on the reduction ratio in use)but on the fact that microfiche can beduplicated... (etc.) (same as #8)

#12 I want aperture cards. Aperture cardsare fun.

Why ?

'17



#13 1 - easy storage (same as # 8)

#14 Fiche. Equipment feasibility, cost, andeventual wide availability.

You are correct in picking microfiche,and cost considerations are an impor-tant factor. Microfiche can be dupli-cated very readily and inexpensively...(etc.) (same as # 8)

# 15 Type 1. Largest amount of experience You are correct in picking microfiche.with this type. Can you compare them not on the

basis of experience but on the basis ofinformation ? (etc.) (same as #4)

#16 It depends on the dissemination systemand the ultimate user(s), but 1 seemsbest.

Why ?

#17 Microfiche, because of Federal stand-ard, cost, and availability ,)f readers.

(same as #14)

#18 1. Density and handling are superiorto 3. copying superior to 2.

(same as #10 FAULT !)

# 19 Microfiche. Large amount of informa-tion stored by unit and a flexible »display.

(same as # 11)

#20 I have used microfiche with a projectorand it seems to be quite satisfactory.

(same as # 10)

I have not used the other two so haveno opinion of their usefulness.

#21 What is opaque microcard ? I understand you need more informa-What is aperture card ? tion on microcard - aperture card. We

will take them up in that order. (Textof OPAQZ satellite) - Does this ex-plain the term adequately for you ?

1 g



#22 Don't know what aperture card is. I understand you need more informa-What is microform card with all the tion on aperture card - (text of APERZpages on one card called ? satellite) - Does this explain the term

adequately for you ?

#23 What is an aperture card ? ksame as #22)

#24 I don't know what a microcard oraperture card are. I prefer hard copyanyway.

(same as #21)

#25 Microfiche, I guess. Access time in See if you can pick just one, on the basismicrofilm is too long I have never of what you can learn from me on theused opaque microcard. subject.

APPENDIX K

B-FRAME » PERFORMANCE SUMMARY

. Attached.

z. For explanation of column headings, see Appendix F (p. 270-271).

'79

Wif

ieO

RM

A-

71M

mrs

trs1

*71.

4mA

rTrt

rpt-

0--

00

B-F

ram

e Pe

rfor

man

ce S

umm

ary

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Ans

wer

Scan

Gro

upSc

an F

ate

Ans

wer

Typ

e

1st p

ass

Pg.tn

reac

tion

sens

ese

nse

Scor

e

hlfi

°.11

Rat

ing

2 31

4

* 5

* 6

* 9

* 7

* 1

*21

*23

424

#22

*10

* 8

*11

I do

n't k

now

.I

am n

ot f

amili

ar w

ith p

rogr

amm

ing.

I ha

ve n

ot h

ad a

ny e

xper

ienc

e w

ithan

y, s

o ca

nnot

dec

ide.

I am

not

suf

fici

ently

acq

uain

ted

with

the

char

acte

rist

ics

of th

ese

diff

eren

tm

icro

form

s to

ans

wer

.L

ack

of k

now

ledg

e pr

even

ts a

con

side

red

answ

er.

I do

n't u

nder

stan

d al

l the

dis

tinct

ions

amon

g th

e th

ree

type

s.W

hat's

an

aper

ture

car

d. T

o te

ll th

etr

uth,

I s

tron

gly

disl

ike

rnic

zofo

rms

and

read

ers.

I do

n't e

ven

know

wha

t a m

icro

fich

e is

.W

hat i

s op

aque

mic

roca

rd?

Wha

t is

aper

ture

car

d ?

Wha

t is

an a

pert

ure

card

?I

don'

t kno

w w

hat a

mic

roca

rd o

r ap

ertu

reca

rds

are.

I p

refe

r ha

rd c

opy

anyw

ay.

Don

't kn

ow w

hat a

pert

ure

card

is. W

hat

is m

icro

form

car

d w

ith a

ll th

e pa

ges

onon

e ca

rd c

alle

d ?

1. B

ecau

se I

don

't qu

ite k

now

wha

t the

othe

r tw

o ar

e 1

Mic

rofi

che,

sin

ce it

is e

asy

to h

andl

e,st

ore,

and

rep

rodu

ce.

1. a

s it

cont

ains

mor

e in

form

atio

n (d

ata)

read

ily &

qui

ckly

ava

ilabl

e.

I II II II H II IV IV IV IV IV I" V V V

MO

Dun

no +

Gar

bD

unno

+ E

ver

M1

Ml

Dun

no +

all

Dun

no +

ape

r

Dun

no +

Fic

heD

unno

+ o

paqu

e+

ape

rD

unno

+ a

per

Dun

no +

opa

que

+ a

per

Dun

no +

ape

r

N1

+ N

2

N.1

+ P

lusc

omp

N.1

+ c

ap

AL

PMA

L,p

mA

Lpm

AL

p m

AL

pm

AL

PM

AL

pm

AL

p m

AL

pm

AL

PMA

Lpm

AL

PM

AL

PU

AL

PM

AL

pm

6b 6a ob 6c.

6c 9a 6d 7b 8a 7d 8a 7d 3

2b (

2)

1 1/2 1 1 1 1 1 1 1 1 1 1 1 1 1

0 1 1 1 1 2 1 1 3 3 2 3 2 3 3N 0

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Num

ber

Gro

upSc

anG

roup

Scan

Fat

eA

nsw

ert

ass

1s p

prog

ra m

reac

tion

Sens

etio

nM

oon

Rat

ing

*12

*25

#13

*14

*15

#16

*17

#18

*20

*19

I w

ant a

pert

ure

card

s. A

pert

ure

card

sar

e fu

n.M

icro

fich

e I

gues

s. A

cces

s tim

e in

mic

rofi

lm is

too

long

. I h

ave

neon

zus

ed o

paqu

e rn

icro

card

.1

Eas

y st

orag

e.Fi

che.

Equ

ipm

ent f

easi

bilit

y, c

ost,

and

even

tual

wid

e av

aila

bilit

y.T

ype

1. L

arge

st a

mou

nt o

f ex

peri

ence

with

this

type

. Rep

rodu

ctio

n an

d en

larg

emen

tqu

ality

is g

ood.

It d

epen

ds o

n th

e di

ssem

inat

ion

syst

em &

ultim

ate

user

(s),

but

1 s

eem

s be

st.

Mic

rofi

che,

bec

ause

of

exis

ting

Fede

ral

stan

dard

s, c

ost,

and

avai

labi

lity

ofre

ader

s.1.

Den

sity

and

han

dlin

g ar

e su

peri

or to

3.

copy

ing

supe

rior

to 2

I ha

ve u

sed

mic

rofi

che

with

a p

roje

ctor

and

it se

ems

to b

e qu

ite s

atis

fact

ory.

Iha

ve n

ot u

sed

the

othe

r tw

u so

hav

e no

opin

ion

of th

eir

usef

ulne

ss.

Mic

rofi

che.

Lar

ge a

mou

nt o

f in

form

atio

nst

ored

by

unit

and

« fl

exib

le a

dis

play

.

V V V V V V V V V V

N.3

N.1

N.1

+ P

LU

Scom

pN

.1 ±

PL

USc

heap

N1

+ P

LU

S ex

p

N.1

N.1

N.1

N.1

+ N

.2

N.1

+ G

arb

AL

PM

AL

U

AL

PMA

LPM

AL

pm

AL

PU

AL

PM

AL

PU

AL

PU

AL

PU

2a 3 in lb

2b (

1)

2a lb - 3 6a

1 1/2 1 1 1 1 1 0 1 0

3 3 3 3 3 3 3 3 3

RE

CA

PIT

UL

AT

ION

:O

vera

ll m

atch

sco

reSe

nse

Scor

eM

otiv

atio

n 0

Mot

ivat

ion

1M

otiv

atio

n 2

25 (

100

%)

22 (

88 %

)W

eigh

t0 6 6

Mot

ivat

ion

3(1

4)M

otiv

atio

n co

mpo

site

Mot

ivat

ion

poss

ible

max

.M

otiv

atio

n fa

ctor

Gro

up V

res

pons

es

42 54 7554

/75

13 (

52 %

)

O


ABSTRACT

A stand-alone CAI frame was written around a question designed to elicita wide variety of replies, in order to measure the utility of reply analysis inrefining individual frames. Half of a real sample of fifty handwritten responseswere submitted to the frame, with a match score of 96 % and a sense scoreof 48 %. After revision, the frame attained a match score of too % and asense score of 88 %, on the other half of the test set.

The authors, discuss selected-string-match CAI strategies, including theuse of attitudinal scan combinations to indicate cognitive state. Mechanics ofthe CAI « macroframe A and its « satellites » are detailed.

RAstnit

A propos d'un projet sur l'instruction par les machines a calculer, un seulcadre isole fut daft sur une question precise afire d'obtenir une grande varietede reponses, et d'etre a meme de mesurer la valeur de l'analyse des reponsesfournie par les cadres individuels. La moitie d'un vrai modele de cinquantereponses ecrites a la main furent soumises au cadre, avec une congruencestatistique de 96 % et un systeme de comprehension de 48 %. Apres avoirete revise, le cadre atteignit un resultat de too %, et un systeme de comprehen-sion de 88 % sur moide du contrOle d'essai. Les auteurs discutent lesstrategies de « selected string match CAI » comprenant l'usage des combinai-sons parcourues pour indiquer Petat des connaissances. On discute les meca-niques en grand cadre et ses satellites.

ZUSAMMENFASSUNG

Fez amine Frage der Programmierten Instruktion wurde ein unabhangigerRahmen geschrieben, der entworfen wurde um eine Reihe von Resultatenzu bekommen und um die Brauchbarkeit der Antwortenanalyse zur Ver-feinerung einzelner Rahmen zu messen. Die Halfte der Originalserie vonfunfzig handgeschriebenen Antworten wurde in den Rahmen gefuttert underreichte einen Erfolgsgrad von 96 % und einen Verstandnisgrad (von derSeiten des Systems) von 48 %. Nach der Revision erzielte der Rahmen mitder anderen Halfte der Versuchsserie einen Erfolgsgrad von too % unterVerstandnisgrad des Systems von 88 %.

Die Autoren beschreiben die Auswahl-Aufeinanderfolge-Anpassung (se-lected-string-match) Strategien des Programmierten Unterrichts, einschliesslichdie Anwendung von verhaltensskandierenden Kombinationen, urn auf dasErkennungsstadium hinzuweisen. Die Mechanik des « Makrorahmens » undseines « Satelliten » werden beschrieben.

4


JOSEPH C. MEREDITH was born at Oakland, California on June 29th,1914. B. A. Olivet College, Olivet, Michigan (i934) and M.L.S. School ofLibrarianship, University of California (1967). Between 1941 and 1962 :

Commissioned service, U.S. Navy; studies in International Law and Strategy& Tactics, U.S. Naval War College, Newport; R.I. NATO staff duties;R&D in amphibious warfare, various commands. From 1967 to 1968 Projectstaff member and from 1968 to 1969 Project Manager, CAI component of aproject entitled « An information Processing Laboratory for Education andResearch in Library Science », (0EG-I-7-71o85-4286), Co-Director of In-formation Processing Laboratory in the School of Librarianship, Universityof California. Is now Assistant Director (Acting) of the Research Center forLibrary and Information Science, Graduate Library School, Indiana Uni-versity, Bloomington, Indiana, 47401.

Publications :« The CAI Author/Instructor an Introduction and Guide » (EnglewoodCliff , New Jersey, Educational Technology, Inc., 197o).« Machine as Tutor » in Educational Technology IX (September 1969).« DISCUS A New CAI Programming Language » (with StevenS. Silver) Institute of Library Research Technical Paper Nom (February 197o).

DOUGLAS FERGUSON was born on July 4th, 1935. B.A. (Philosophy)Macalester College (1956). M.L.S. School of Librarianship, University ofCalifornia (1967). Post-master's course work, ditto. From 1967 to 1968 TeamMember, Indian National Scientific Documentation Centre, New Delhi.Study and proposal for the National Electronics Information Network. FromJanuary to September 1969 specialist in CAI Course Development, Instituteof Library Research, University of California. Is now editor of the AutomationDivision, The Stanford University Libraries, Leland Stanford, Jr., University,Stanford, California 943o5.

43

DOCUMENT RESUME Student Feedback as a Tool in …Education; v7 n2 p221-302 1970. EDRS PRICE EDRS...

Documents

Transcript of DOCUMENT RESUME Student Feedback as a Tool in …Education; v7 n2 p221-302 1970. EDRS PRICE EDRS...