Embed Size (px)
Citation preview
This article was downloaded by: [The University of Manchester Library]On: 18 December 2014, At: 07:02Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK
ErgonomicsPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/terg20
REHEARSAL AND RECALL IN IMMEDIATE MEMORYA. F. SANDEKS aa Institute for Perception , RVO-TNO, Soesterberg, The NetherlandsPublished online: 26 Apr 2007.
To cite this article: A. F. SANDEKS (1961) REHEARSAL AND RECALL IN IMMEDIATE MEMORY , Ergonomics, 4:1, 25-34, DOI:10.1080/00140136108930504
To link to this article: http://dx.doi.org/10.1080/00140136108930504
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in thepublications on our platform. However, Taylor & Francis, our agents, and our licensors make no representationsor warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Anyopinions and views expressed in this publication are the opinions and views of the authors, and are not theviews of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should beindependently verified with primary sources of information. Taylor and Francis shall not be liable for any losses,actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoevercaused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
gEHEARSAL AND gECALL IN IMMEDIATE MEMORY*
By A. F. SANDERS
Institute for Perception, nVO-TNO, Soesterberg, The Netherlands
Experiments on the influence of rehearsal on the retention and recn.ll of digitcombinations are described, from the results of which it appears that a rehearsalperiod facilitates recall by producing a transition from immediate to permanentmemory. It fur-ther- seems that some parts of t.ho material presented reachpermanency very quickly, without much rehearsal, and also that mere t.runsmissionof material by tho organism without intention to memor-ize it, nevertheless has someeffect on material in the long-term memory store. It is argued that, during arehearsal period, the main activity of t.hc subject is not to repeat the material automatically, but to assimilate it with the conditional probabilities of past events.
§ 1. INTRODUCTION
DURING the last decade important research has been carried out on immediatememory and its functional meaning in human behaviour. Experiments haveshown that it is useful to distinguish a short-term mechanism from a morepermanent storage of information. The former may be thought of as closelyconnected with the perceptual system, while the latter forms a general background of experience and expectancy for the organism (Broadbent 1!l58).The material entering the immediate memory store is liable to a very rapiddecay (Brown 1958, Conrad 1957), but can be maintained by continuousrehearsal. Relatively little work has been done, however, to determine whatthe effects of rehearsal might be on the transition of the material from theshort- to the long-term store, or in more general terms, on learning. As Irion(195!l) states: "If rehearsal, as a form of perseveration, has any effect uponpost-rehearsal performance, it should be possible to treat it as an independentvariable and to determine how it relates to other variables in the determinationof learning performance. Far too little work has been done on this importanttopic." The general feeling, however, seems to be that rehearsal has a negligibleeffect on learning. For example, Brown states: "Everyday experiencesuggests that the effect of rehearsal may be to counteract decay of theimmediate memory trace more than to strengthen it"; and Woodworthand Schlosberg (l !l54) write: "If you look up a telephone number asPlaza 46315, you can repeat it easily a moment later because it falls withinyour primary memory span. If you wish to retain it for later use, you wouldgain very little by saying it over and over."
The experiments reported in this paper .were designed to throw some lighton this matter in relation to the following questions
1. What is the effect on the stability of retention of inserting a rehearsalperiod between memorizing and recall 1
2. What are the characteristics of the transition from immediate topermanent memory 1
*This paper is bused on one read to the Ergonomics Rcseurcb Sociot.y on 2U "larch, 1960..
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
26 A. F. Sanders
§ 2. EXPERIMENT I
A combination of eight digits (DC) was presented to the subject for 5 sec,in which time he had to memorize it. This was followed by a rehearsal periodand a recall (R2a, RIb). Two conditions were compared:
A : Another recall (RIa) was interpolated immediately after memorizing.H : Only rehearsal occurred between memorizing and recall. The lengths
of these rehearsal periods were respectively 35 and 40 sec.The two conditions may be illustrated thus:
A : DC RIa 35 sec R 2aB : DC 40 sec RIb
We should expect R 2a to be poorer than RIb, since the material in the shortterm store is supposed to deteriorate during the interpolated recall (RIa) incondition A.
An answer to the first question, concerning the influence of rehearsal onthe stability of retention, is provided by a variant of this experiment shownin the following plan:
C: DC 12 sec RIC 23 sec R 2cB : DC 40 sec RIb
The interpolated recall (RIC) in condition C is preceded by a rehearsal period,instead of coming immediately after DC as in condition A. The B conditionserves as a control. If rehearsal merely counteracts decay of the immediatememory trace, we should expect again a deterioration at R 2c, as comparedwith R,b' If, however, rehearsal also improves the stability of the material,the first rehearsal period of 12 sec should bring R 2c and RIb nearer equality.
The digits were projected onto a screen and a mechanical device assuredthe constancy of the presentation time. Before each presentation and recallthe subject was given a warning signal. The total time for one DC, within thedescribed procedure took approximately 50 sec. Forty-four Air Force ratingsserved as subjects. Thirty-two were tested in the A and B conditions in aconnterbalanced order. Twelve subjects were tested in the C and B conditionsin the same way. All were instructed to memorize the digits in groups oftwo, e.g, forty-one instead of four one. Each subject was given ten differentdigit combinations in each of the two conditions. The score consisted of thenumber of mistakes made in reproduction, and every digit that was notreproduced in its right place counted as a mistake. For example, completefailure to answer a whole digit combination would have counted as eightmistakes. The mean numbers of mistakes per subject that were made areshown in the following plan :
A : DC RIa 15·9 R 2a 26·3
B: DC RIb 20·1
C: DC RIC 17·8 R 2c 17·9
B : DC RIb 15·8
An analysis of variance of these data showed that in the first pair of conditions (A, B), there were significant differences between subjects (F = 6,70,P < 0'01), between trials (F = 10·62, P < 0'01) and between R 2a and RIb(F = 16,51, P < 0·01). A I-test showed R 2a significantly different from RIa.
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
Rehearsal and Recall in Immediate Memory 27
Another analysis of variance failed to reveal a significant difference betweenRIa and RIb. For the second pair of conditions (B, 0) there were no significantdifferences between recalls (R , b, RIc and R 2c) when the same statisticalprocedures were employed.
From these results it appears that the interpolation of a recall immediatelyafter memorizing, has a detrimental effect on a later recall, compared withan interpolated rehearsal period. The effect seems, however, to be overcomeif a rehearsal period precedes the interpolated recall as happened in theC condition. This appears to imply that the memorized material has beenstrengthened during the rehearsal period.
It should be mentioned in passing that the individual scores for the firstpair of conditions (A, B) proved to be highly related to verbal intelligencetest performance, the less intelligent group performing worse at the memorytests than the more intelligent (t = 4,38, P < 0,01).
§ 3. EXPERIMENT II
In this experiment the effect of interpolated recall was stndied moresystematically. Again combinations of eight digits were presented for 5 sec.After presentation a recall was required every 4 sec until 10 reprodnctions(R I to RIo) had been given. Twenty-five new Air Force ratings were tested,to each of whom five digit combinations were given. They were instructedalways to reproduce eight digits, even if they were not sure they were exactly
average % of changes
10
9
8
7
6
S
3
2
N = 2S x 3 x 8 items
o~='--""""'----'--'-----"-""""~_"'--""""'_
10 9 8 7 6 S 3 2
number of recalls
Eigure 1. Mean percentages of changes between each recall and the next. (Experiment II.)
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
28 A. F. Sanders
those of the original combinations. The scoring was done in terms of theamonnt of change betwcen each recall and the next. Each digit not in thesame place as in the immediately previous recall, was regarded as a change.The other details of method and procedure were the same as those ofExperiment 1.
It was expected that the rehearsal between recalls-and the recallsthemselves-would strengthen the memory traces and result in progressivelyfewer changes between each recall and the next.
The mean percentages of changes are shown in Fig. 1. Applying analysisof variance to these results significant differences were found between recalls(1'=3'15, p <0·01) and between subjects (1'=2·01), p <0'01), while t-testsshowed that significant differences between recalls stopped after the fourth recall.
There appears from these results to be a gradual tendency towards greaterstability of the material with each successive rehearsal period and recall.That this cannot be solely due to the recalls as opposed to the rehearsal isindicated by the results of the ensuing two experiments.
§ 4. Ex L'~;RIMENT III
From the experiments reported so far we can conclude that, with rehearsal,some transition occurs from short- to long-term memory, but we cannot saywhat is the extent of the process or what are its characteristics. We can,however, obtain a measure of the degree of transition by means of the interfcrencc technique of Brown (1958). In this, information is presented, andimmediately afterwards the subject is given another task, such as readingprose or digits aloud. The difference from traditional' retroactive inhibition'techniques lies in the fact that no new memorizing is required. After completingthe second task, thc subject is asked to recall the original information. Brownfound that as the quantity of material neared the limit of the memory span,the proportion retained became less.
Now it is possible to carry out such an experiment allowing rehearsalbefore interference. In fact Brown did this with rehearsal times up to 4 secand found there was a significant improvement in performance. The presentexperiment was designed to extend his work to longer rehearsal and interference times.
Digit combinations were presented in the same way as in Experiments Iand II. Subjects were allowed to rehearse them for some time after presentation, and then performed an additional task, which consisted of a quick readingaloud of digits from a Kraepelin test sheet. Just before starting the reading thesubjects recalled the digit combination and this recall served as a criterion fora final recall after the reading had been completed. Before the test begansubjects were asked to read out the KraepeJin digits at their fastest rate, andthis was used as a criterion they were required to reach during the interferencetask. This was done to prevent rehearsal during the interference task. Sixtysubjects were tested. Since intelligence had appeared in Experiment I to bean important factor in the memorizing process, students were used as subjectsin order to prevent the occurrence of very low scores. The subjects weredivided into seven groups. In three groups rehearsal was varied and interference helel constant: in the others rehearsal was held. constant and
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
Rehearsal and Recall in Immediate Memory 29
interference varied. The times of both rehearsal and interference were asindicated in the following plan:
Experimental Times in Number-ofGroup Oonclit.ions seconds subjects
l. Rehearsal varied : 40, 28, 20, 12, 8 10Interference constant: 120
2. Rehearsal varied : 28, 20, 12, 8 8Interference constant, : 90
:J. Rehearsal varied: 40, 28, 20, 12, 8 10Interference constant, : 1.5
4. Rehearsal constant: 40 8Interference varied 120, 60, 30, 15
5. Rehearsal constant: 20 8Interference varied: 120, 60, 30, 15
6. Roheursel constant: 12 8Interference varied : 120, 60, 30, 15
t. Rehoursnl constant: 8 8Interference varied: 120, eo, 30, 1.5
All tests within a group were performed by the same subjects in a Latin squarearrangement. For each test, three digit combinations with the appropriateprocedure were given, scoring was the same as in Experiment 1.
mean amount
of mistakes
1.
~17·0
15·0
1.·0
13'0
12·0
11'0
10·0
9·0
8·0
1'0
6'0
5·0
4'0
],0
2·0
t ·0 a'
0
........
.'" .
,,,I
-"
,iIiI
I,~
;.i
2' interference1.5' interference1S~ InterferenceBrown's data
(N=10i(N~ e,(N=IO)
40 36 32 28 24 20 16 12 8 4 0 rehearsal time
In seconds
Figure 2. Mean numbers of mistakes made under conditions with length of rehearsal time variedand duration of interference task held constant. (Experiment III.)
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
30 A. F. Sanders
The results of groups I, 2 and 3, in which rehearsal times were varied whileinterference was held constant, are shown in Fig. 2. Analysis of variance onthese results showed a highly significant difference between results withdifferent times (F = 10'37, P < 0·01 at 120 sec interference; F = 8'59, P < 0·01at 90 sec interference and F = 13·7, P < 0·01 at 15 sec interference). Brown'sdata of 4 sec rehearsal with 7 sec interference have been transformed andincluded in this graph where, it will be seen, they fit very well.
The results of the remaining groups, in which rehearsal time was heldconstant and interference varied, are shown in Fig. 3. Analysis of varianceshowed signifieant differences between results with different interference timesonly when rehearsal times were 20 see (F =8'87, P < 0'01) and 12 see (F = 10·7,P < 0,01). There were no significant differenees with 40 see and 8 see rehearsaltimes.
It is clear from Fig. 2 that recall is better after prolonged rehearsal. Wefound in Experiment I that substantial benefit occurred after a rehearsal timeof 12 sec, but this does not imply that the transition to the long-term store hasbeen conipleted after 12 sec. Further rehearsal continues to improve recall ofthe digits. It is reasonable to suppose that we can get an idea of the actualstability of the material by examining the extent to which interference exertsan effect. Fig. 2, however, does not give satisfactory evidence on this point,
mean amountof mistakes
2~ _ complete~ blank
17·0
,5'0
14·0
13·0
(N=8)
(N=8l(N~8
(N=8
•
8" rehearsal_. _._._ 11" rehearsal__ _ 20· rehearsal___ 40" rehearsal
•
c
•
·.--e-- --l='--~--·=:"·::·:=:'"
Ii
j
I;;
;;
I
•,.
oI
//
///
_0
3·0 "',-
2·0: •------,r-------J
8·0
1.0
6'0
5'0 :
12·0
11.0
10·0
9·0
o W 30- 60'
Figure 3. Moun numbers of mistakes made under conditions with length of rehearsal time heldconstant and duration of interference task var-ied. (Experiment 111.)
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
Rehearsal and Recall in Immediate Memory 31
because the different interference times were studied with different groups ofsubjects. The results for different interference times shown in Fig. 3 were,however, all obtained from the same subjects. These results seem to suggestthree points about the transition of material from the short to the long termstore, and also shed some light on the characteristics of the latter:
(a) When interference starts we can reasonably assume that the untransferred material still in the short term store, decays immediately. This ishypothetically indicated in Fig. 3 by the dotted lines drawn from15 sec to 0 sec interference. The fact that plateaux exist in the graphspresumably indicates that the transferred digits have a fair permanenl'Y.This might serve as an additional argument for the functional distinctionbetween the short- and the long-term stores, as retention in this studyseemed to be of the all or none variety.
(b) With prolonged rehearsal the material is not only transferred but is alsoadditionally strengthened in the permanent memory. Thus with40 sec rehearsal time the trace is strong enough to stand 120 sec ofinterference. It cannot, however, do so with rehearsal times of only20 or 12 sec.
(c) At first sight it is plausible to suppose that the decay in the long-termstore reaches a maximum value with 12 and 8 sec rehearsal times.It appears however that it has not produced a complete blank. Sincethree digit combinations were presented in each test, a blank wouldhave produced 3 x 8 = 24 mistakes. Allowing for chance, this makesapproximately 22 mistakes. Yet even with 8 sec rehearsal time,the maximum number of mistakes averages no more than ten. Theexplanation of this is presumably that some parts of the material areplaced very quickly and very firmly into the permanent store. Wemay reasonably suppose that these correspond most closely to theconditional probabilities of past events-or to express the point inmentalistic terms, those parts where some rule is evident to the subjectare found easy to remember. Other parts of the material are moreliable to decay, as is clear from the increase of mistakes with prolongedinterference at 20 and 12 sec rehearsal times.
§ 5. EXPERIMENT IV
If we suppose that certain parts of the immediate memory load are veryquickly transferred to the long-term store, it might be hypothesized that theformer has an influence on earlier stored material in the latter, especiallywhen the stability of the material in the permanent memory is not very high.The present experiment was carried out to test this hypothesis. The experimental scheme was thus :
40 sec2. DC I Rn c I LC RI•C Rn c I--------------40 sec3. DC I RnOl (15 sec reading) R nol-----------
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
a2 A. :F. Sanders
4. DC I 12 sec Rn Ol DC 2 Rn c 2 E n c l
5. DC, 12 sec UVCI LC Rw R DCI
li. DC I 12 sec UDCI (15 sec reading) E VCl
All conditions again started with an eight-digit combination (DC,) presentedfor 5 sec. A rehearsal period of 40 sec for the first three and of 12 sec for theremaining conditions followed. The interference consisted of memorizingeither a new digit combination CDC.), or a letter combination (LC). In twoconditions (3 and 6) subjects were asked to read digits from the Kraepelin sheetas a control. .Immediately after the interference task a final recall took place.Where interference had consisted of a new immediate memory load, recallwas first required of the new material. The interfering digit combinationsconsisted again of eight digits and the letter combinations of six letters. Thesubjects were instructed to recall this material as well as possible. Otherdetails were identical with the previous experiments. Twelve new studentsperformed all six conditions in a Latin Square order. In each condition twoDC,s were presented. The scoring was the same as in Experiment III. Justbefore the interference task tile subjects were asked to recall the DC, and thisserved as the criterion for the final reproduction.
It was expected that the new memorized material (DC., LC) would exercisea negative influence on the reproduction of the earlier material (DC,) especiallywhen rehearsal had been relatively short, as in the 12 sec conditions. \'\Te werealso interested in comparing the results of the digit and letter conditions, inview of the similarity of the former to DC I .
mean amountof mistakes
16
~10
8
6
4
2
o
..---
..---
..---
r-r-
n coNDITIONS
DCl~DC,DCl~LCDC,~lS"READINGDC,-,-,-DC,DC,-'-'-LC DC,-'-'-IS" READING
Figure 4. Mean numbers of mistakes mode following different types of interference task andrehcaranl times. {Exper-iment. IV.)
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
Rehearsal and Recall in Immediate Memory
The mean numbers of mistakes made under the various conditions are:shown in Fig. 4. An analysis of variance showed significant differences betweenthe conditions (F = 16,09, P < 0,01). Appropriate t-tests revealed that recallsafter rehearsal times of 40 sec were generally better than after those of 12 sec.There was no evidence for significant differences within the 40 sec rehearsaltime conditions, but in the 12 sec conditions performance after memorizingletters was better than after memorizing digits (t = 4,44, P < 0,01), and alsobetter than after reading digits (t = 2,46, P < 0,05). The difference betweenperformances after memorizing and reading digits was not significant.
The results for the 12 sec rehearsal times suggest that loading of theimmediate memory directly affects the long term store. Memorizing newmaterial which belonged to the same class as the material originally memorizedhad a clearly negative effect on retention of the original, and this' retroactiveinhibition' must presumably have taken place in the long-term store. Thefact that the mere transmission of information that resembled the material tobe recalled (i.e. reading Kraepelin digits) had a similar effect points to theconclusion that, whether or not memorizing is required, the permanent storeplays a role in, and is influenced by, every intake of information especiallywhen the material resembles very closely that which is already being carriedin memory. The disappearance of the effects with 40 sec rehearsal time wouldbe understandable if the material gained stability in the course of rehearsal.
§ 5. GENERAL DISCUSSION
The results of the experiments taken together indicate strongly thatrehearsal has a strengthening and stabilizing effect on retention of the kind ofmaterial used here. It seems doubtful, however, whether this improvementresults from rehearsal in the sense of a mere automatic repetition, or indeedwhether this kind of process is very important at all during rehearsal periods.Brown for instance, reported that several subjects tried to find interpretationsfor the letters he presented (e.g. National Debt for N.D.). This kind oftendencywas noticed during our experiments and subjects also tried to impose rhythmsand artificial rules on the material. In fact, the main activity of the organismduring a rehearsal period seems to be the assimilation of the material by meansof interpretation, imposition of rhythms, finding of rules, etc., to the conditionalprobabilities of past events in the long-term store. Only during the periodimmediately after memorizing a digit combination does rehearsal approachautomatic repetition. We may, perhaps, suggest that as soon as some partsof the material are transferred, the immediate memory load is lessened andthe repetition necessary to counteract decay of immediate memory becomessecondary to repetition aimed at long-term assimilation. In support of thisview it may be noted that subjects tended to slow down their rehearsal ratesafter they had rehearsed for some time.
As we have seen, the mere transmission of information may have an effecton the long-term store. The effect is probably transitory in most cases thoughit may interfere with material which has been newly stored. Its effects appearto be less on material which has been rehearsed for a period and thus,presumably, strengthened and consolidated. It seems from observation thatthe long-term store may perhaps be subject to satiation in experiments like
ERG. C
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
Rehearsal and Recall in Immediate Memoru
this, so that retention of later digit combinations is hindered by theperseveration of earlier ones.
The results of these experiments do not agree with those of Conrad (1960),who found no strengthening effect of a 5 or 10 sec rehearsal period. Therewere however several differences between the conditions of his experimentsand the present ones which might explain the discrepancy. For example, hisrehearsal periods were very short compared with our experiments, and thepresentation was auditory instead of visual. Also his subjects were notinstructed to group the digits into pairs and they were perhaps somewhatless intelligent than ours-both factors which might have tended to a shorterimmediate memory span,
The field seems ready for further experiments. Apart from testing thehypotheses made in this discussion, the question arises of how the mode oftransition to the permanent memory relates to what is found in traditionallearning experiments and phenomena of inhibition in that field.
I am indebted to Drs. E. W. J. Zwaan for help with testing subjects.
Loa experiences decritos dans le present article etudient. I'Influcnce de In. repetition sur losouvenir et 10 rappel de combinaisons de chiffrea. II s'evere qu 'uno pcriodc de repetition facilite10 rappel en creunt uno transition entre In. memoire immediate at In memoire differce. II semble,on outre, qu'unc certnine partie du materiel presente attcint tres rapidement un degre de permnnOIlCO, suns bORUCOHp do repetition; de memo, In. simple transmission de materiel par I'orguniame,BUllS intent.ion do lc memorisoi-, influence a un certain degrc le materiel dejA enmagasine dans lamemoirc lliffcrce. On demcnt.ro que pendant In periode de repetition, l'ect.ivite principale dusujut no consistc PUg a repetcr automatiquement Ie materiel, mais a l'assimiler tout en tenantcompto des probabilit6s conditionnellos effcrentea (lUX evenementa du passe.
Ea worden Expcrimente libel' dna Foethulten und Wicdererinnorn von Zahlen-Kombinationenbcechr-icben, aUB denen hervorgcht., duss des \Vicdcrerinncrn durch einen Ucbergang aus dcmKurz-Godttchtnis in ein blcibcndes Gcduchcnie crlcichert.. B~ scheint ferner-, dues cin Toil desdurgebotonon Materials sehr reach, ohne viclc Uobung, blcibend fcstgchalten wird, und dossuuch die blosse Uebernuhmo von l\laterial in den Orgunismus. ohnc Absicht cines Erinnerns, desMaterial in langfristigen Gcdncbtnis-Vorrat bccinfiusst. .Es wird urgumentiert, dess die Hauptsachc boi del' Uebung nich t die nutomat.ische Wicdcrholung cines Mater-ials ist, sondern dcssenVerbindnng mif den bedington \Vahrscheinlichkciten vergangener Eroignisse.
REFERENCES
Bn.OADUgN~·, D. E., IOfi8, Perception and Oommuniccuion (London: Pergamon Press).Bnowx, J., 1958, Some tests of decay theory of immediate memory. Qrmrf.. J. expo Psychol., 10,
12-31.CONRAD, R.; J958, Decay theory of immediate memory and paced recall. Canad. J. Peucnot.,
12, 1-6j 19no, Very brief delay of immediate recall. Quart. J. expo Psychol., 12, 45-47.rUlON, A. 1...., 105H, Rotc leut-ning. In Psychology: A Study of a Science, Editor: Sigm. Koch.
Vol. II (London: McGmw.Hill).\VOODWORTH, H.. S., and SCHLOSBERG, H., 1954, Experimental Psychology. (London: Methuen
& Co. Ltd.)
Dow
nloa
ded
by [
The
Uni
vers
ity o
f M
anch
este
r L
ibra
ry]
at 0
7:02
18
Dec
embe
r 20
14
