Th is s tudy has been designed to answer specif ic hypo theses concerning the utility of contras ting approaches for teach ing word recogni t i on . Additionally , the f indings offer some prac tical in formation on c ogni tive characteris tics of the mentally retarded
learner . In o rder that the results migh t be d is cuss ed within a
theore t i cal f r amework , this s e c t ion has been o rganised to integrate
the experimental f ind ings w i th previous work undertaken on acquis i tion , r e t ention , and trans fer of l earning . The maj or f ocus o f this Chapt er is upon the t r ans lat ion of theory into practice . An attemp t has been mad e to h ighl igh t some principles of l earning wh ich were found to b e
importan t f o r teaching word recogni tion to mentally retarded l earners .
The maj o r findings of this res earch demons trate that microcomputer b as ed ins truc t ion and more convent ional tui tion result in s imi lar
l earning ou t c omes desp i te procedural variations ar ising f rom the two d if f erent me th ods of ins truct ion prov ided . As predicted , there were
no reliab l e d ifferences b e tween the two groups and mod es of ins truc tion in terms of p erformance on the cri terion measures of Word Recogni tion , Identi f ication , and P i c ture-Word Matching . The fact that all subj ects r e corded significant gains as measured by a comparison o f pre- versus pos t- tes t change s cores , leads . to spe culation as to ·wh ich underlying p ro cesses ac count for these cons is tent improvements in l earning .
The results of this s tudy are in accordance with the f indings o f Parmenter e t al . ( 19 79 ) who demons trated that mildly re tarded adole s cents learned a shor t l i s t o f words equally w e l l under two condi tions o f ins truc t ion . S imilarly , Vergason ( 19 66) f ound that both automated and trad i t ional methods of ins truct ion r esu l t ed in s imilar learning out comes as measured by a tes t of r e tention adminis tered one day af ter comp l e tion of training . The main p o int that mus t b e considered in this s tudy and preced ing research is the extent to which treatment cond i tions were differ ent f rom one another . I t may be that a number o f common e l ements were operating under b o th conditions despite more obvious variat ions in the methods o f ins tru c t io n .
as us ed b y Vergas on ( 1 9 6 4 ) and Holz ( 19 76) , was compared w i th an error l ess discriminat ion t echnique in whi ch no pic tures were used such that the s t imulus word to b e l earned s erves as the s o l e focus o f the l earner ' s a t t ent ion (Walsh and Lamb erts , 1 9 79 ; Vandever and S tubbs , 1 9 7 7 ) . If , as S amuels ( 1 9 70) has argued , the e f f e c t of pi ctures on acquis i tion of a s ocial s igh t vocabulary is minimal , then one is left to ques t ion wh e ther o r no t subj e c ts were shi f t ing at tent ion across dimens ions f rom picture to word . Indeed , the word ' s relat ion to a referent picture may have
b een perceived as almos t coinciden t al by subj ects participat ing in the paired ass o c iate mode . Th is as sump tion is in l ine wi th Duel l ' s ( 19 6 8 ) obs ervation that a des i red sh i f t in s t imulus control i s more likely t o o ccur when s ubj ects a r e forced to n o t i c e a cue wh ile respond ing . I n practical t erms , failure to attend to extra s t imulus dimens ions (pictures ) may mean that subj ects in the paired associate mode wer e , in e f f ect ,
focus ing mos t of their at t ention on the printed material and re ceiving a s imilar f orm of training to that offered in the errorless discrimin ation mode of ins truc t ion . Thus , the es s ential difference b e tween th e two modes may have b een s imply the order and number of mul t iple cho i c es that were presented f o r each of th e four trials in a given training s e quence .
Subj e c t ive impress ions s ugges t ed that while trainees taugh t by the
paired ass ocia te method attended to p i ctorial ma terial at the b eginning of each sequence , they rapidly sh i f ted to a s traight word focus and t ook l ess not i c e of the p i c ture during the training phase . The fact that no fur ther reference was made , e i th e r by the ins tructor or computer , to the pic ture f o llowing introduc t ion o f the word , l eads to the observation that the extra s timulus material was optional ( no t ess ential ) for success ful completion of each of the word i t ems and trials . By imp lication , the photographs may have played a p as s ive rol e in the learning pro cess s u ch that the two mo�_af ins t ru c t ion operated in a s imilar manner to one ano ther .
The res u l ts o f the experiment c learly demons trate that the two groups and mod es of ins tru c t ion were equally effective f o r t eaching a word recognit ion tas k . I n view o f the nons ignificant interaction eff ects ,
i t s eems l ikely that a common s e t of factors , independent of the treat ment variab les , cont r ibuted to the changes in learning . Firs t ly , the use of individual ins truction in a high ly interactive s e tt ing resul ted in high l evels of a t tention and mot ivation . The effect of b eing s ingled
out for t raining was apparent in the subj ects ' b ehaviour . They usually anticipated their turn on th e computer or w i th individual ins t ruct ion and eager ly came to the sess ions . Secondly , the relative lack of ex traneous s timul i in th e training rooms ( e . g . the ab sence of wi ndows , ot her trainees and indus trial no ises) prob ab ly aided concentrat ion on the learning task . In terms of the Zeaman and Hous e ( 1 9 6 3) th eory o f at tention i t can r easo nab ly b e ass umed that the focus on a s t ructured set of d imens ions ( e . g . target words and r e f erent pictures ) as s is ted the learner to sort out relevant learning cues . The logical s equence of ins tru c t ion provided a concrete learning s e t in which the s ubj ect could org anis e input materials .
Th i r d ly , th ere are a number of principles of learning that are operating under b o th condi tions of ins truction . Thes e factors are a func t ion o f programmed ins truct ion in general rather than computer as s is ted ins tru c t ion in particular , and have previously b een shown to ass i s t the l earning p rocess ( Clarke and Clarke , 1 9 7 3 ; Brown , 1 9 7 5) :
1 . Ac t ive partic ipation
2 . S e l f-pacing ins truction 3 . Overlearning of materials
4 . Reduc tion of emo t ional dependence 5 . Immediate r einf orcemen t .
Ac t ive participation was an es s ential part of the l earning process .
Subj ects were encouraged to make their own d e cis ions and to work inde pendently . This was c ons is tent with Skinner ' s ( 1 9 68) view that every a t t emp t should b e made to ensure the s tudent does no t become a pas sive recipien t of new information . Rather , oppor t uni ties for exploration
and self direction need to b e o f f ered with in a s tructured l earning
medium . I n b o th the computer as s is ted ins truction and individual t ui t ion groups , s ubj ects h ad full con tro l of th e s p e ed o f pres entation . The longes t response t ime measured f o r comp l e t ion o f a module ( 5 s es s ions ) was 20 3 minutes whi l e th e sho rtes t time inte rval was 24 minu tes . The
time generally decreas ed markedly fos all s ub j e c ts as they b ec ame f amiliar with the procedures : However , i t was no ted that sub j ec t s
rec eiving computer s s i
�
ted ins t ruction req���ed s ligh t ly more time to complete their training s e s s i ons than pers ons who were given individual ins truc t i on . This t r end could be a ttributed to the sequence of time delays programmed into the compu ter .The empirical nature o f this s tudy predisposed prime attention to the mor e obj ec t ive and measurab l e aspects of b ehaviour and learning .
The shor t coming o f this approach was that many impor tant s i tuat ional fac tors and descrip t ive features of the s ub j ects ' respons e to the teach ing programmes tended to be overlooked . A variety of no ti ceab l e changes i n b ehaviour that were s ub s equen t ly no ted by th e t eachers and experimenter were n o t amenab le to measurement in the pre- and pos t- t es t s i tuations . Thes e qual i tative aspec ts o f b ehaviour were b as e d upon obs ervations and anecdo tal record ings comp l eted by the examiner at the time o f each s es s ion .
Inf ormal obs e rvations s ugges t that c omputer based ins truc tion can be of particular v alue to subj ects who are overly dis trac tab le or in clined to engage th e teacher in o ff- task verbal interact ions . One
par ticipant , previo us ly cons idered incapab l e of b ene f i t ing from individual ins truct ion due to aggressive b ehaviour ( e . g . abusive language , inapprop riate physical con tact) was ob s erved to f o l low ins truc tions and presumab ly learn through interaction with the compute r . Ano ther s ubj e c t who d emon s trat ed gro t esque personal manners ( e . g . l oud burping no ises , gazing into space , and verbal imitation o f background noises) could operate th e
computer sys t em and eagerly at tend to the requirements o f th e l earning task . Many other exampl es could be ci ted to support the view that computer assis ted ins truc t ion , while no t a t t emp ting to modify dif f i cult b ehaviours , may o f fer a sui tab l e learning environment for persons who
are less ab l e to a t t end and interact in an interpersonal t each ing s ituat ion .
A maj or obs tacle in providing prac t i cal social education for mentally retarded adults is the inappropriate us e of child o riented mat er i als . Commonly , the teaching of ear ly learning concep ts involves
the use of obj e c t s ( e . g . b l o cks , f l ashcards , children ' s s tories ) that are no t in keeping with regular adult interes ts . I t was no tab l e that s everal of the t eachers who adminis t ered the individual ins t ruc t ion programmes , c ommented af terwards that they felt the t eaching of word recognition by t r aditional p i c torial and f lashcard methods was a rather und ignified · and r egress ive approach not suited for us e wi th adul t
subj e c ts . By way o f contras t , the same materials pres ented under
computer control give the outward appear ance o f an adul t learning medium that incorporates element s o f new technology and ski l lful operation by the user .
I t is generally the case tha t mentally retarded adu l ts are exposed
to sheltered l i fes tyl es and have a l imi ted range of oppor tunities for
i t was p o s s ible to o f fer an ind ependent learning experience such that s ubj e c t s interacted with the computer at the i r own pace and were re quired to take ini tiative and r espons ib i l i ty for solving problems . Indeed , many subj ects preferr ed to work ind ependently and would wai t for th e experimen ter to leave the room before proceding wi th th eir l esson .
S ubj e c tive impres s ions by teachers s ugges ted that par t i cipants receiving individual tu i tion were aware of failure experiences wh en
these o c curred and that p ercep t ions of error by both the s tudent and t eacher would impede performance on sub s equent i tems . Des p i t e the us e
o f neutral feedback ( e . g . "Okay " , "Try again") there was a need to
provide verbal encouragement to help over come these fai lure experiences . Convers ely , i t was noted that t rainees who were given computer ass is ted ins truc t ion responded actively to the feedb ack provided throughout the t raining sequences . I t was no t i ced , for example , that subj e c t s would o f t en verbalise the flashing messages ( e . g . *** IS RIGHT *** , ***TRY AGAIN ***) and nod their heads in respons e to ' Happy Face ' i llus trations display ed on the video s creen . Repetition o f the i t ems when errors were made d id not appear to result in negative perceptions of failure though s ubj e c ts were aware that they had reco rded an incorr e c t response . Such obs ervat ions are in agreement with a s tudy by Holz ( 1 9 7 6) who no ted th e willingness of men tally re tarded ch ildren to proceed with computer as s is t ed ins truction despi t e f requent r ep e t i t ion of i tems when errors o c curred . In sum, it would s e em that the relatively impersonal nature of th e computer can p rovide d irect and meaningful f ee dback wi thou t triggering negative percep t ions o f failure that tend t o have a n advers e e f f e c t on performance in more conventional forms of i ndividual ins truct ion wh ich involve direct interpersonal f eedback .
Contrary to expectati on , it was found that lis t order e f fe c ts were s ignif icant f o r the Word Recogni tion Tes t of Module One . An examination of the s ubgroup means indicated that s ubj ects who were given lis t o rder one (module 1 followed by module 2) tended to achieve margina l ly higher pos t- tes t s cores than was the case for subj ects receiving l i s t order
two (module 2 followed by module 1 ) . Several different interpretations
may b e advanced to accoun t for this finding . It may be that the tool
names contained in Module One formed a more dis cernib le and concep tually ,
meaningful s t imulus clas s ( e . g . HAMMER , SPANNER) than the varied infor mat ion items o f Module Two ( e . g . BREAD , RAILWAY ) . I f this was the cas e ,
elements facili tated input organisat ion and generally res u l t ed i n im p roved performance by sub j ects given lis t order one . Suppor t for this assump tion i s available f rom previous res earch by Bi lsky and Evans ( 1 9 70 ) . These exper imenters found that when mentally re tarded ado les cents were f irs t trained on a clus t ered l is t of words , their performance on sub s equent l is t s of random i tems improved to a s ignif icant extent . The educat ional implicat ions o f the present results are that i t may be pos s ible to facili tate performance on educational tasks s u ch as word recogni t ion by remed iating speci f i c deficienc ies in input o rganisation