Spatial memory by blind and sighted children

8/18/2019 Spatial memory by blind and sighted children.

1/14


2/14

Br. J. Psychol.

(1975), 66, 4,

p p.

449-459 4 4 0

Printed in

reat

Britain

SPATIAL MEMORY BY BLIND AND SIGHTED CHILDREN

SUSANNA MILLAR

Department of Experimental

Psychology

University of

xford

Non-verbal recall of haptically presented spatial positions by three age groups of blind and

sighteil children was tested under conditions varying cueing, recall type and stimulus position

in a within-subject design. Sighted stMuii was not only significant, but interacted significantly

with recall typ e, and further w ith stimulus position, consistent with sequential hap tic by blind

and quasi-simu ltaneous visuai processing by sighted children. Age was eignifleant, bu t its only

significant interaction was a relatively small one with cueing conditions and stimulus position,

suggesting that the oldest group, regardless of sightedness, used verbal strategies in pre-cued

conditions. The findings support th e hypothesis th at visual and hap tic modalities of representa-

tion have demonstrably different effects on processing and efficiency in spatial recall, but

couiiterindicate the hypothesis tha t these relate differentially to age. Resu lts also suggest t ha t

a combination of cue utilization and verbal strategies is a significant, but relatively minor, factor

in improvements in spatial recall.

The nature and function of imagery in memory for spatial position, the extent to

which children use it, and its relation to other factors which affect memory are by

no mean s clear. A t least thre e views are possible; Piag et Inhe lder (1971, 1973)

regard visual imagery as largely irrelevant, and stress the level of abstraction from

actual action by the subject in the development of spatial as of other types of

memory. This may mean either that modalities of representation have no differential

effects and serve as 'props' for mental operations in the same way, or that action

and movement with regard to objects and positions, and thus haptic factors, have

special relations to memory functions, A second view is that visual imagery has

specifically facilitating effects in spa tial mem ory. H utte nlo ch er Presson (1973)

suggest th at children, like adults (A ttneave Benson, 1969; Rock , 1974; Shephard

;

Feng, 1972; Shephard Metzler, 1971), visually image spatial layouts and mentally

track positions on it. This suggests that visual imagery has a special role in spatial

memory and possibly also in improvements with age, A third possible view is that

spatial memory and development depend on verbal strategies.

Theoi'etically, tbese views involve questions about 'memory modalities' or, more

simply, whether differential mo dality effects can be dem onstra ted in spatial m emo ry,

and, if so, how these relate to cognitive skills; for instance, the organization of

material for recall, and the utilization of prior information; and to age. Memory for

haptic (touch and movement) inputs, about which relatively little is kllow^l as yet,

is of particular interest, because it provides a test case for the nature and effects of

representations, but also because it is of practical importance in teaching the blind

who need to rely on spatial memory far more than the sighted.

The present study was therefore designed to test hypotheses arising from these

differing views. Although visual imagery is frequently reported and suggested as an

exp lana tion , it is more difficult t o find experimentally dem on strated differential

effects on spatial memory in children. The methodological problems and difficulties

of interpretation inherent in Piagetian and Paivio-type methods have been dis-


3/14


4/14

Spatial memory 451

delay (Gilson Baddeley , 1969) or with unfam iliar non-verba l m ater ial (Millar, 1974).

In the author's current study on mental spatial rotation the sighted seemed to rely

on visual strategies from the point of stoppage, while the blind were differentially

penalized on the more complex directions and lost track already after two turns,

consistent with sequential coding and the findingB that haptic memory is less

subject to active strategies, except with familiar material. Thus, if the blind rely on

haptic and movement rather than on verbal coding for spatial recall, a differential

effect on the primacy position for the older blind should not be found, while an inter-

action of recall type with sighted status would be predicted.

Support for the hypothesis that modality of representation is a factor in develop-

ment with age would be shown by appropriate age interactions. Rate of forgetting

as such does not seem to be an important factor in memory development {Belmont

Butterfield, 1969). I t is of interes t, however, to re late th e role, if any , of the m oda lity

of spatial representation to factors involving mental work in memory: the reorganiza-

tion of material as in backward recall, and the utilization of prior information for

encoding.

This study therefore tested spatial recall by three age groups of blind and sighted

children, varying cueing conditions, type of recall and stimulus position to test the

hypotheses that (1) the modality of representation has differential effects on efficiency

and on strategies of spatial recall; (2) this is a factor in improvements with age; and

(3) improvements with age in spatial recall depend upon verbal strategies, organiza-

tion, or cue utilization.

METHOD

A flve-way factorial design with age, sighted status, pre- and post-cueing, recall type, and

stin]ulu8 position (repeated measures on the last three factors) was used for non-verbal recall of

Iiaptically presented spatial positions as follows.

Subjects

There were 72 subjects: 36 congenitally or very early (20 mo nths or less) totally or severely

Ijlind (minimal light, but no pattern perception) from two residential schools for the blind; and

• S

sighted children from two local authority schools within tlie catchment area, and similar in

range of ability (low average to very superior) and socio-economic status of the parents. Tiie

sighted were matchod to the blind on sex, age (within threo calendar months), forward digit

span, and backward digit span (ascertained in the manner of Binet tests). IQ scores, although

necessarily on different tests, available for the blind and some of the sighted, and ability ratings

by teachers wore taken into account in matching.

There were 12 sighted and 12 blind subjects in each of three age groups - oldest gi-oups:

seven girls, five boys; mean age 10:4 (from 9:1 to 11:8); mean forward span, 6-0 (from 5 to 8);

mean backward span, 3-2 (from 2 to 4); middle groups: four girls, eight boys; mean age, 8:2

(from 7:0 to 8 :10 ); mean forward sp an, 5-4 (from 4 to 7 ); mean backw ard apan, 2-9 (from 0

to 5); youngest groups: six girls, six boys; mean age, 6:11 (from 6:2 to 7:5); mean forward

span, 5-0 (from 4 to 7); mean backward span, 2-2 (from 0 to 4).

aterials

The material consisted of a

35

x 4 cm slot cut in a 20-25 cm Perspex square to form a five-sided

open-shaped sp atial plan w itli four right-angled tur ns . This was presented in one of four positions

(north, south, east, west) in randomized order across conditions and trials for each subject.

For presentation, the subject traced along the slot with a pencil from the start to prepared


5/14

452 SU S N N MI L L R

the subject s haiid at tliis point. It was removed after contact and tlie subject continuod tracin

the slot to the eud. The recall task was to trace the slot with the penoil, either

fi-ora

he start o

from the end of the slot, to the position on the slot where tho stop had hoen encountered on th

presentation trial. After recall the stop was replaced for the next presentation.

Subjects usod only their preferred hand to hold the pencil. They were not permitted to palpa

the display except by tracing tlirough the slot with tho pexxcil in the experimental condition

For siglitcd subjects the display was presented in a curtained unit which prevented viewing, b

permitted free hand and arm movements.

Task

The task was presented as a game of teacliing

Bozo ~ a rubber animal fastened to the end

the pencil to remember the position of a traffic -light post, so that he would stop at that positio

on the road next time.

Tho starting point for eacli presentation was the start of the slot nearest to the subject an

his preferred hand, in one of the four positions. Tliia point

w£is

called the subjects house . T

end-point of the slot was called Bozo s kennel , and the slot was the road between the subjec

house and Bozo s kennel . The stop formed by the stick and suction base was called the traffi

light post .

The subject was told that he always had to take Bozo on two journeys {by tracing the sl

with the pencil to which Bozo was fastened) between his house and Bozo s kennel. The fir

journey would always be from the subject s house (starting-point) to Bozo s kennel (en

point), but they would be stopped by colliding with a traffic light-post somewhere on the roa

I he post would then be removed so that they could go on to Bozo s kennel. The main point wa

to remember the point on the road where they had felt the post. On the second joiu-ney th

traffic-light post would not be there. Nevertheless, on the second journey Bozo would have

stop dead at the point where the traffic-light post hatl been on the first journey ( to avoid traff

liazards ). Presentation consisted in tlie first journey from house to kennel. The recall task w

the second journey .

Typt of recall

Type of recall was either forward or bacltward.

Forward

recall: the subject was instruct

that the second joumoy between the house (starting-point) and tho kennel (end-point) wou

start again at the same point

—

the subject s house . He would have to take Bozo along

road from the house until he came to the point where he and Bozo had felt the trafSc-light po

on the first journey. This meant that he would have to trace through the slot with tlie pencil

which Bozo was fastened to that poiat and not any farther. The traffic post would no longer b

there, but he had to make sure that Bozo stopped at exactly the same point on the road whe

the traffic-light post had been on the first journey. Backtvard

recall:

the subject was instruc

that the second journey between the house and the kemiel would start from the keixnel

the end of the road. He would have to take Bozo back along the road (trace through the slo

etc.) from the kennel (end-point of the slot) to the point on the road where they had felt th

traffic-light post on the first journey. The traffic post would no longer be there, but he had t

make sure that Bozo would stop at exactly that point.

Recall was imniediato (approx. 3 sec.) after the end of presentation, and was signalled b

same, from the house or back, from kennel , as appropriate for forward and backward reca

To ensure a rapid start to the recall, the experimenter helped the subject to insert the top of th

pencil at the start, or end of the slot, as appropriate,

xt ing

conditions

Tliere were two cueing conditions, Tliey difFered solely on whether instructions for the type

recall were given prior to presentation (first journey) or only prior to recall (second journey

Pre cued conditions: the subject was informed prior to a block of forward recall runs that rec

(the second journey) would always start at the same point - his house - on the second journe

as on the first. For the block of backward recall runs, the subject was informed prior to the bloc

of runs that recall (second journeys) would always be back from the kennel (end-point of slot


6/14

Spatial memory

453

Fig. 1. The five-sided spatial display slot here shown in the W es t position. Inpu t d uring

presentation is from tracing the slot from H (house) to K (kennel). Sj represents the first (near),

S( represent the recent (far) stimulus position. Forward recall is from H to either ̂or Sg. Back-

ward recall is from K to either Sj or Sj.

also sta rt from th is house or Rem em ber, the second journey will sta rt from the kennel at th e

end ) .

Presentation was then signalled immediately by st ar t . Post cued

conditions

the subject

was told prior to the whole block of post-cued runs that the second journeys could either be

forward from the house, starting at tlie same point as the first journey or backward from the

kennel, starting back from tho end-point, but that he would not know before each of the first

journeys w hether the second one would be from the house or from the kennel. He would, however,

be told which of these it was to be, at the start of each of the second journeys by the signal

sam e, from the house or baek, from kenne l . Presentation was signalled by s ta rt .

Subjects had no difficulty in understanding the tas k, and thoroughly enjoyed th e ga m e .

Stirmdus position

Two positions on the d isplay, one near the st arting point (7-5 om) and one far from it (27 6 om ),

were tested. Since tracing the display was sequential, the near stimulus represented an initial,

and the far stimulus a recent, stimulus in relation to the touch and movement inputs of the

display. Tlie display slot, sta rt ( house ) and end ( kenn el ) of presentation , and t he n ear (first)

and far (recent) stimulus positions are shown in Fig. 1.

Procedure

Subjects were tested singly in a quiet room of their school. The game waa explained to the m ,

and they received two demonstration trials (not counted), followed by 16 main trials (four in

each of the eonditions: pre-cued forward, pre-cued backward, post-cued forward, post-cued

backward). The order of cueing conditions, blocked in threes (post-cueing both recall types in

counterbalanced order, pre-cued forward; pre-cued backward recall), recall type and stimulus

position were counterbalanced across subjects.


7/14

454

S U S A NN A M I L L A R

RESULTS AND DISCUSSION

Errors were counted in mm deviations from the stimulus position. Mean absolu

errors,

shown

in

Table

1,

were subjected

to a

five-way ANOVA with

age

{younges

middle, oldest), sighted sta tus blind, sighted), cueing conditions pre -po st),

type

recall forward, backward) and stimulus position near, far) as variable s repeate

measures

on the

last three factors).

Table 1.

Mean

bsolute errors

mm)

for

blind

nd

sighted children under

different cueing, rec ll nd stimulus position conditions

Pre-cueing Po«t-cueing

stimulus position .. .

Mean age

Blind 6:11

8:2

10:4

Heaa

Sighted 6:11

8:2

10:4

Mean

Forward recall

Near

69

58

51

59

86

110

31

76

F a r

70

67

40

5 9

52

51

21

4 1

Backward recall

Near

100

119

90

103

78

57

44

60

P ar

59

37

42

46

41

40

21

34

Forward recall

Near

97

62

85

81

101

47

60

7 3

F a r

58

76

54

63

70

40

22

44

Backward

Near

103

119

96

106

66

41

41

4 9

recall

F a r

55

54

54

54

69

62

11

47

8 I

7

6

50

Blind

Sighted

Forward Backward

Type of recall


8/14

Spatial memory

465

U O r

90

8

7

6

50

4

Forward recall

blind

Backward recall

blind

Forward recall

̂ • sighted

• Backward recall

sighted

far

Stimulus position

Fig. 3. Mean absolute errors (mm) by blind and sighted children for near and

far stimuli in forward and backward recall.

Sighted status was a significant effect {F = 12-04; d.f. = 1, 66 ; P 0-001) and

interacted significantly with recall type {F = 10-0; d.f. = 1, 66 ; P 0-005), sup-

porting th e first hypothesis. The interactio n, shown in Fig. 2, shows no t only tha t the

blind prod uced larger errors, bu t th at t he m ain difference was obviously on back w ard

recall, Avith significantly

{P

0-01) worse backward than forward recall for the blind,

while for the sighted the relation was reversed. This supports the assumption that

the blind used sequential strategies and the sighted did not.

The relation of sighted status to recall type waa further elucidated by the fact that

this interacted also with stimulus position {F = 11-40; d.f. = 1, 66; P 0-005).

Stim ulus position was indeed a highly significant m ain effect { =

69 97;d.f.

= 1,66;

P

0*0001). This meant that the near (primacy) position was very much more

difficult. The in teraction of sighted sta tus , recall type a nd stimulus position

is

graphed

in

Fig.

3. This shows th at for the b lind the difference between recall of th e two po sitions

was clearly greater for backward recall, with greatly increased errors on the near


9/14

456 SusAJ^NA

M I L L A R

recall from stimulus presentation to recall point were greatest for the near positio

and least for the far position for backward recall, and between these for forward recal

this implies that time and distance determined errors by the blind. Tliis sugges

a sequential haptio rather than a verbal rehearsal strategy. For the sighted, th

pic ture is very different. The rece nt position was unaffected by recall t> pe, b

responses for the near (primacy) position were actually easier in backward tha

forward recall, i.e. in the opposite direction from w hat would be predicted with verb

strat-cgies. Reorganization of sequential verbally encoded positions on the displa

would produce larger rather than smaller error for backward recall, and verb

rehearsa l would no t penalize responses to th e near position in forward recall. A com

pletely simultaneous strategy could not have been used either. The results sugge

that one sighted visually evoked or envisaged the display from tbe end-point (

recall) regardless of cueing conditions, rather than during presentation. This wou

penalize responses more on the near position in forward recall. A similar strategj^

reconstruc ting fi om the po int of stoppag e w as used by the sighted in the au tho r

current study on mental rotation of a haptically presented line.

Age {F = 8 -23; d.f. = 2, 66; P < O-OOl) was significant, b u t did no t significantl

interact with sighted s tatus in any combination, cou nterindi eating th e hypothes

that modality effects were a determining factor in improvements with age.

The only significant interaction of age \vith other factors was with a combinatio

of cueing conditions and stimulus position {F = 3-25; d.f. = 2, 66; P < 0-05

graphed in Fig. 4. This shows that pre-cueing significantly P


10/14

p ti l

m emory

7

I-

9

7

o 60

SO

4

3

N^

Youngest group

post-cued

^ «

Middle group post-cued

Youngest group

pre-cued

Middle group pre-cued

' N _ \ ^ Oldest group

^. post-cued

- '* Oldest group

pre-cued

far

Stimulus position

Fig. 4. Mean absolute errors (mm) for the youngest {mean age, 6:11), middle (mean age, 8:2)

and oldest {mean age, 10:4) groups for near and far stimuli in pre- and post-euod coiiditions.

point, possibly because they mainly recoded visually at recall rather than during

presen tation. The findings supported th e hypothesis th a t m odalities of represe ntation

differentially affect efficiency and processing in spatial recall.

The findings also supported the view that visual imagery has a special role in

spatial recall by children. In so far as Piaget's theory implies the opposite, or suggests

th at visual imagery provides 'p ro ps ' in the same w ay as other types of represe ntation,

this was counterindicated. The absence of any interactions between sighted status

and age, also counterindicated the implication in Piaget's theory that haptic and

movement representations have a special role in age changes. At the same time, this

finding does support his view that visual imagery as such does not relate to age. It

suggests indeed that age changes were independent of visual and haptic modality-

specific processing.

The fact that recall type did not relate to age may explain this. For visual -

simultaneous or evoked imagery - reorganization and therefore worse backw ard

recall and age changes in this \vith increased use of active organizational strategies


11/14

4 5 8 SUSAJ^NA MlLLAH

wa rd recall was predicted for sequential hap tic processing an d w as found for th e blin

The fact th a t tliis did no t in terac t w ith age either

is

consistent w ith previous findi

that haptic memory is less subject to active strategies unless the material is ove

learned (Gilson Badde ley, 1969; Millar, 1974, 1975a,

b

Sullivan Turvey, 197

A reasonable explanation of the absence of relations between modality effects an

age is therefore th a t visual ima gery and processing does not requ ire, and ha ptic repr

sentations and processing do not readily lend themselves to, active reorganization

Finally, the fact that the only variables which interacted significantly wilh ag

were a combination of cueing conditions and stimulus position which implied th

utilization of prior information elicited verbal strategies by both blind and sighte

subjects in th e oldest gro up , m ust be considered. Since th e effect, a ltho ugh significan

was relatively sm all, and ne ither cueing conditions nor stimu lus position alone relate

to age, this suggests (i) that age changes depended on combinations of factors rath

than on any single cognitive or verbal one, and (ii) that cue utilization and verb

coding played a significant, but relatively small part in age changes in spati

memory.

The practical implications for the blhid are twofold. First, the deliberate use

verbal labelling and rehearsal, particularly at the time spatial positions are fir

encountered, seems to aid recall of the more difficult positions, but probably requii'

specific teaching and encouragement in younger children who do not seem to do th

spontaneously. Secondly, since the effect was small and the blind apparently main

relied on memory for touch and movement, repeated actual familiarization wi

spatial layouts which allows over-leaniing of touch and movement cues is probab

essential.

In summary, the findings suggest a model of spatial memory in children whic

includes modahties of representation as important factors, determining type

processing, but not changes with age. It was argued that this may be because visu

an d ha ptic, or modality-specific represe ntations and the typ e of processing these elic

do not easily lend them selves to active controls. The finding.s suggest further th

such a model may need to include changes to a combination of verbal strategies an

utilization of prior information for encoding as a significant, but relatively min

factor in age changes in spatial memory. The main improvements in spatial memor

with age seemed to take place within the modalities of representation without bein

determined by the differences between them.

The research was aupported by a grant from the Social Scionoe Reaoarch Council. Tho auth

wislies to thank tlie staff and the ctiildren of the seliools wlio took part in the experiment f

their unfailing help and cooperation.


12/14

Spatial memory

459

R E F E R E N C E S

R. C. S H I F P R I N , R . M . (1968). Human memory: a proposed Bystem and its control

processes. In K. W. Spence J. T. Spence (eds.). The Psychology of Learning and Motivation,

vol. 2. New York: Academic Press.

F . B E N S O N , B . (1969). Spatial coding of tac tua l stimulation. J. exp. Psychol. 81,

216-222.

G. (1971). An experimental study of visual imagination and memory. Cogn. Psychol. 2,

290-299.

Y . M . BuTTERFiELD, E. C. (1969). The relations of ahort-tenn memory to develop-

ment and intelligence. In L. P. Lipsi tt H. W. Reese (eds.), Advances in Child Development

and Behaviour. London: Academic Press.

H . A . (1967). The effect of labels in short term memory for colours with nursery

school children. Psychon. Sci. 7, 149-150.

C, J . H E T J V E L , K . V, (1974). Development of geometric imagery in five- to eight-

year olds. Genet. Psychol. Monogr. 89, 89-143.

L . R . (1967). The suppression of visualizing hy reading.

Q. J. exp. Psychol.

19, 289-299.

R . (1971). Thechronology of the development of covort speech in children. Dev. Psychol.

5 ,

398-405.

GiLsoN, E. Q. B A D D E L E Y , A. D. (1969). Tactile short-term memory. Q. J. exp. Psychol. 21

180-184.

B . O ' C O N N O R , N . (1971). Spatial coding in normal, autistic and blind children.

Percept, niot. Skills 33, 127-132.

NLOCnEB, J. P R E S S O N , C. C. (1973), Mental rotation and the perspective problem.

Cogn. PsycJwl. 4, 277-299.

S. (1974). Tactile short-term memoi-y by blind and sighted children. Br. J. Psychol. 65,

2 5 3 - 2 6 3 .

M I L L A R , S. (1975a). EfFects of phonological and tactual similarity on serial object recall by blind

and sighted children. Cortex (in press).

MiLLiiR, S. (19756). Effects of tactual and phonological similarity on the recall of Braille letters

by blind children. Br. J. Psychol. 66, 1 9 3 - 2 0 1 .

X E I S S E E , V. K E R R , N . (1973). Spatial and mnemonic proper ties of visual images. Cogn.

Psychol. 5, 138-150.

O ' C O N N O R ,

N . H E R M E L I N , B .

(1972). Seeing and hearing in spa.ce and t ime . Percept. Paychophys.

46-48.

PiAGET, J. I N H B L D E R , B . (1971). Mental Imagery in the Child London: Routledge Keg€in

Paul.

PiAQET, J. I N H E L D E R , B . (1973). Memory and IrUelligence. London: Rontledge Kegan Paul.

R O C K , I. (1974). Tlie perception of disoriented figures. Sci. Amer. 230 78-85.

S H E P H A R D , R . N . F E N G , C. (1972). A chronometric study of paperfolding.

Cogn. Psychol.

3,

2 2 8 - 2 4 3 .

SHErHARD, R. N.

M E T Z L E R ,

J . (1971). Mental rotat ion of three-dimensional objects. Science,

N.Y. 171 701-703.

StJLLiv.,iN, E. V. TcTRVEY, M. T. (1972). Sliort-term retention of tac tile stimulation. Q. J. exp.

Psychol. 24,

2 5 3 - 2 6 1 .

{Manuscript received 3 October 1974; revised manuscript

received 10 February 1975)


13/14


14/14

Documents

Spatial memory by blind and sighted children