The Journal of Genetic Psychology Research and Theory on Human Development

ISSN: 0022-1325 (Print) 1940-0896 (Online) Journal homepage: http://www.tandfonline.com/loi/vgnt20

Comparison of Normal and Learning Disabled Children on a Nonverbal Short-Term Memory Serial Position Task Lee Swanson To cite this article: Lee Swanson (1978) Comparison of Normal and Learning Disabled Children on a Nonverbal Short-Term Memory Serial Position Task, The Journal of Genetic Psychology, 133:1, 119-127, DOI: 10.1080/00221325.1978.10533364 To link to this article: http://dx.doi.org/10.1080/00221325.1978.10533364

Published online: 04 Sep 2012.

Submit your article to this journal

Article views: 3

View related articles

Citing articles: 1 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=vgnt20 Download by: [York University Libraries]

Date: 06 November 2015, At: 23:04

The Journal of Genetic Psychology, 1978, 133, 119-127.

COMPARISON OF NORMAL AND LEARNING DISABLED CHILDREN ON A NONVERBAL SHORT-TERM MEMORY 'SERIAL POSITION TASK*'

Downloaded by [York University Libraries] at 23:04 06 November 2015

School of Special Education and Rehabilitation, University of Northern Colorado

LEE SWANSON SUMMARY A serial recall task was used to compare performance of 15 normal and 15 learning disabled elementary school children matched on CA, ZQ, and sex with two and three dimensional representations of nonverbal eight-point shapes. Two a priori assumptions were not supported: (a) no differences in recall were found between groups and (b)no differences in recall were found for either group using two or three dimensional stimuli. Three dimensional stimuli did facilitate visual rehearsal at the primacy position for both groups. Learning disabled children's performance was consistent with the mediation deficiency hypothesis found with normal children. A. INTRODUCTION The purpose of this investigation was to compare recall performance of normal and learning disabled (LD)children on a nonverbal serial recall task. The a priori assumption that visual short-term memory (STM) differs between normal and learning disabled children has not been experimentally investigated, although several clinical findings have suggested that learning disabled children have a deficit recall performance on visual STM tasks due to atypical mediational deficiencies (e.g., 12, 17, 18,22), compared with normal children. A secondary purpose of this study was to investigate the notions of Piaget and Inhelder (19) and Gibson (11) that three dimensional stimuli

* Received in the Editorial Office, hovincetown, Massachusetts, on March 17, 1977. Copyright, 1978, by The Journal Press. * The research reported herein was supported by the Research and Fellowship Office at the Universityof New Mexico. Appreciationis due to the AlbuquerquePublic School System for their cooperation. Requests for reprints should be sent to the author at the address shown at the end of thii article. 119

Downloaded by [York University Libraries] at 23:04 06 November 2015

120

JOURNAL OF GENETIC PSYCHOLOGY

facilitate STM recognition. A common assumption of many remedial approaches for the supposed “visual STM lag” is that performance of LD children is hampered by two dimensional objects; three dimensional objects are held to facilitate correct perceptions which then further facilitate recall performance (e.g., 6, 20, 22). Generally, the development of nonverbal visual STM has not been studied extensively in young children. Current mediation literature has been monopolized for learning groups (normal and LD) by the effects of language. Recall response patterns with normal children using familiar stimulus material have usually yielded primary-recency effects and middle serial position bias (1, 38). The bowed shape serial position pattern found in verbal mediation studies has indicated that the first (primacy) and last (recency) serial positions have a higher probability of being correctly recalled than the middle positions. The shape of a child’s forgetting function is due to a failure to rehearse (e.g. 9, 16). Ellis (9) found that verbal rehearsal plays its most important role in primacy performance. Attempts to account for this have centered around a mediation deficiency hypothesis. The mediation deficiency hypothesis states that there is a stage in development during which verbal responses are present but do not serve as mediators (e.g., 7, 10, 15). This hypothesis has been used in studies (e.g., 5 , 7, 21) to explain nonverbal mediation on a STM memory task with normal children. Corsini, Pick, and Flavell (5) have suggested that visual STM deficits found in children are not just related to verbal rehearsal, but to visual coding. To test the mediation deficiency hypothesis and to explore the developmental effect of verbal labeling on children’s STM, studies have been done with the procedure outlined by Atkinson, Hansen, and Bernbach (1). These studies have included an overt label, as well as a nonlabel condition (3, 10, 16). In a study by Bernbach (3), an assumption was made that pictures without overt labels would lead to more of a forgetting function than pictures with labels. Results showed that the nonlabeling group had no primacy effect. This study did not try to add perceptual cues (e.g., three dimensionality) outside of the labeling aspect to determine if a primacy effect could be achieved without a verbal mediation means. Other investigators have also failed to get a primacy effect with young children (1, 4, 8). A later analysis by Keely (14) of the data from the preceding three studies found the presence of a primacy effect by the use of a statistical measure adapted from signal detection theory. The present investigation was designed to eliminate verbal mediation as much as possible from visual stimuli in order to determine learning group differences in visual STM. The study was also designed to determine whether nonverbal data were recalled with the use of similar response patterns to those

Downloaded by [York University Libraries] at 23:04 06 November 2015

LEE SWANSON

121

in verbal mediation studies. Specific hypotheses guiding this investigation were that (a)normal learning children would perform significantly better on both two and three dimensional tasks than the learning disabled children; and (b) three dimensional recognition would be significantly better than two dimensional for both groups. Dimensional salience effects were investigated for group and serial positions. Salience was based on the degree to which a two or three dimensional item was correctly recalled by the child. B. 1.

METHOD Subjects

Normal and LD children in two groups of 15 each, matched on CA, ZQ,and sex, served as Ss in the experiment. Children were selected from regular or special education classes at the same elementary school in the Albuquerque Public School system. Both groups had a mean C A of 9.1 (SO .26). The mean Otis-Lennon computed ZQs for both groups were 103.5 (SO 6.34). Classification of the learning disabled children was based upon the New Mexico State Standards and the Appraisal and Review Committee in the Albuquerque public schools which define a learning disabled student as “one who exhibits one of more deficits in the essential learning process requiring Special Education, which may be characterized by various combinations of deficits in perception, conceptualization, language, memory, control, and attention, impulse or motor function.” None of the children in this experiment was on medication. All children in both groups were randomly assigned for participation in the two treatment conditions. Seven of the children in each group received the two dimensional treatment first. All children participated in both treatment groups with a two week time interval between the sessions. Order effects were not found in the treatment conditions.

Materials The stimuli were six nonsense random shapes selected from Vanderplas and Garvin’s (23) eight-point assortment. These shapes (Numbers 19, 20, 2 2 , 24, 25, 26) were chosen because of their low “association” and “content” values. Each of the shapes used was drawn in black ink on a 10 x 10 cm white card with the same dimensions as the normed shapes. Also, a three dimensional representation of each random shape was constructed out of black Styrofoam. Both pictures and objects were made in duplicate. The objects and pictures were of the same dimension, color, and surface texture. The depth of the three dimensional representation was 4.5 cm. Wooden boxes 10 cm square were 2.

122

JOURNAL OF GENETIC PSYCHOLOGY

used to cover the three dimensional objects in the serial recall task after they had been presented.

Downloaded by [York University Libraries] at 23:04 06 November 2015

3.

Procedure

A serial recall task procedure, similar to that used in Atkinson et a1.k (1) study was used in this investigation. The main difference in this investigation was that coded unfamiliar items were used and no pretest was given to determine labels. The absence of a pretest was based on the assumption that pretesting children to determine spontaneous labels would provide a set for verbal mediations, even though the verbal codability of the items would be low for both groups. Individual children were first shown all of the items at one time for two seconds. This was done to give the child a rough approximation of the shape of the stimuli and to protect him from being overwhelmed when the rate of information input exceeded his capacity for information processing, thus causing a low hit rate in the early trials (3, 14). Cards or objects were collected immediately after this initial presentation. Following this, the child was presented with a series of six items (two or three dimensional) shown one at a time. Each item was exposed for two seconds. Children were not allowed to touch items. Two dimensional pictures were put in a face down array after exposure and the three dimensional objects were covered when not being shown. Once the array had been presented, a duplicate item (probe) was then shown and the S was asked to point out the corresponding item in the presentation series. Fifteen trials were presented in each session. Each trial consisted of six stimuli selected randomly for each serial position with the stipulation that eachposition would be correct no less than two times or more than three, and no shape would be correct more than twice or less than once over the 15 trials. Stimulus was presented from the child’s right to left so that spatial and presentation positions were confounded. After the last trial presentation, children were asked how they remembered the items. All content and association responses were recorded as was done in Vanderplas and Garvin’s (23) study. Since the children participated in both treatment groups, each session was separated by a two week interval. The length of each session was 30 minutes. 4.

Statistical Analysis

Recent studies (e.g., 2, 13, 14) have discussed and applied signal detection theory to the study of short-term memory. According to signal detection theory, two values may be obtained to estimate memory strength: a discrimi-

123

Downloaded by [York University Libraries] at 23:04 06 November 2015

LEE SWANSON

nation index and response bias. The discrimination index (d') can be calculated from the hit rate and false alarm rate (13). The hit rate was defined in this experiment as the proportion of correct serial positions that were recognized by the child; the false alarm rate was the proportion of times a serial position was identified when the item was not actually presented. The Cj index was used as a measure of response bias (2). The Cj index was determined from thez score of the false alarm rate.

C. RESULTS The d' values were calculated at each serial position under both dimensionality conditions for each child's performance. A 2 X 2 X 6 repeated analysis of variance was performed on these scores. The mean d' values were calculated and are illustrated in Figure 1.

t f

3.50-

Ef!!

3.00

Q

u

'6 c 2.50 u

c

2.00

u e

1.50

c

;c

f

1.00 .50

.oo -.so 2

1

3

4

5

Serial Position FIGURE

MEANd' FIRSTCHOICERESPONSE

1

AS A FUNCTION OF LEARNING GROUP, DIMENSIONALITY, AND SERIAL POSITION

LD = learning disabled.

6

124

JOURNAL OF GENETIC PSYCHOLOGY

Downloaded by [York University Libraries] at 23:04 06 November 2015

Nonsignificant main effects were found for normal vs. LD (F = .016, df = 1/28,p > .05), andtwovs. threedimensionality(F = .028,df = 1/28,p > .05). Figure 1 shows a similar overall recall performance for both groups and dimensionalities. There was no significant interaction between dimensionality and group, which indicated that salience of two or three dimensional representations was not affected by learning ability. The main effects of serial position were significant (F = 3.65, df = 5/140,p < .OS),as was theinteraction of dimensionalities x serial position (F = 2.38, df = 5/140, p < .05). Mean comparisons, t (140), of primacy and recency performance at the .05 level of significance were done to investigate apriori concerns of the investigation and interaction effects. Primacy was defined in this investigation as significantly better recall performance at serial position 1 compared with the child’s recall performance on serial positions 2 and 3. Recency was defined as significantly better recall performance at serial position 6 compared with serial positions 4 and 5. A t-test within dimensionalities showed significant effects (p < .05) for three dimensional primacy and two and three dimensional recency performance. A most unusual finding of these data (as shown in Figure 1 ) is the lack of a primacy effect in the two dimensional condition for both learning groups. t-test values indicated significantly better recall within the two dimensional treatment for both groups on serial positions 2 and 3 compared with serial position 1 . Between groups and dimensionalities, significantly better recall was found for the three dimensional primacy condition compared with the two dimensional serial position 1 score. Significant (p < .05) recency effects were found within both dimensionalities and groups except for normal groups performance on the three dimensional condition. No significant differences were found between groups and dimensionalities on recall performance for the recency position. To obtain more direct evidence regarding differences in criterion level over serial positions, a measure of response bias, C j (2), based on the false alarm rate was determined. In Table 1 , the lower score represents the greater bias for selecting a particular serial position. As shown in this table, the greatest response bias for both learning group (normal vs. LD) and dimensionality (two vs. three dimensional) was in serial positions 3 and 4. The lack of primacy effect in the two dimensional condition cannot be explained by a marked bias for the second serial position, which Keely (14) found in other studies masked the primacy effect. Both group and dimensionality treatments had a bias for the middle positions, a finding which is similar to other studies where familiar verbal materials were used (e.g., 1 , 10, 16). S s in the experiment could not be identified as verbal rehearsers utilizing the

125

LEE SWANSON TABLE 1

MEANRESPONSE Bus VALUES(Cj) Grouo

Downloaded by [York University Libraries] at 23:04 06 November 2015

3-D (Objects) Learning disabled

Normal 2-D (Pictures) Learning disabled

Normal

1

Serial positions 2 3

4

5

6

1.80 1.92

1.19 1.44

.88 .97

1.08 1.36

1.49 1.58

1.68 1.70

1.33 1.79

1.54 1.18

.96 1.16

1.26 1.02

1.37 1.49

1.53 1.85

approach of Keeney, Cannizzo, and Flavell (15). Direct observations indicated that none of the Ss was giving overt verbalizations. At the end of the 15 trials, Ss were assessed as being verbal rehearsers if they could provide association or content responses to three items. Again, none of them met this criteria. D.

DISCUSSION

One of the most obvious findings of the preceding experiment is that there was 110 group difference on nonverbal serial recall STM performance. Similar recall strategies (e.g., middle response bias, primacy and recency effects) were found for both groups, suggesting also similar mediational strategies. According to other nonverbal STM findings of Flavell and associates, learning disabled children and normal learning children followed similar developmental mediational patterns in nonverbal STM memory. Mediation in this study refers to some intermediate variable which is involved in storing and retrieving input (10). The nature of the stimulus, whether two or three dimensional, surprisingly did not have an overall significant effect on the recall performances in either group using nonverbal material. This is contrary to Piaget and Inhelder (19) and Gibson’s (11) notions and remedial assumptions on the salience of three dimensional material. While recall of nonverbal materials followed similar recall patterns as those studies using familiar material, stimulus dimensions did not facilitate recall. Why the additional cue of depth on these nonverbal shapes was not a facilitating factor for recall would only be speculation at this point. However, the absence of a primacy effect in the two dimensional treatment was a unique finding. Since no verbal rehearsal was used, as indicated by the lack of primacy effect (e.g., 9), the children’s poor performance may have been due to an inability to attach a visual code or to rehearse visually the first serial position. The three dimensional treatment did yield a

Downloaded by [York University Libraries] at 23:04 06 November 2015

126

JOURNAL O F GENETIC PSYCHOLOGY

primacy effect, indicating that a redundancy (salience) of visual information was necessary in order to provide more cues to facilitate visual rehearsal. The results of the present experiment suggested the possibility of a nonverbal visual rehearsal strategy at primary position on three dimensional treatment. It is important to note, however, that no overall effects or other serial positions were found. In summary, nonverbal STM recall in this study was not dependent on learning ability but upon the effective use of the available mediator. The present results suggested that LD children’s nonverbal memory can function effectively in an STM task similar to nonverbal STM response patterns of normal children. Both learning groups in the present group appeared to be subject to similar mediational inefficiencies. Since little experimental research has been done in terms of nonverbal deficiencies, more research has been done in terms of nonverbal deficiencies, more research is needed to determine what mediators are available to the children and what perceptual cues are needed to improve mediation prediction. REFERENCES 1.

2. 3. 4. 5. 6. 7. 8. 9.

ATKINSON,R. C., HANSEN,D. N . & BERNBACH, H. A. Short-term memory with young children. Psychon. Sci., 1964, 1, 255-256. BANKS,W. P. Signaldetection theoryand human memory. Psychol. Bull., 1970,74,81-99. BERNBACH, H . A. The effect of labels on short-term memory for colors with nursery school children. Phychon. S c i . , 1967, 7, 149-150. CALFEE,R. C., HETHERINGTON,E . M . , & WALTZER, P. Short-term memory in children as a function of display size. Psychon. S c i . , 1966, 4, 153-154. CORSINI,D. A,, PICK, A. D . , & FLAVELL,J. H. Production of deficiency of nonverbal mediators in young children. Child Devel., 1968, 39, 53-58. CRUICKSHANK, W. Perception and Cerebral Palsy Studies in Figure-background Relationships. Syracuse, New York: Syracuse Univ. Press, 1965. DAEHLER,M . W., HOROWITZ,A. B . , WYNNS,F. C., & FLAVELL,J. H. Verbal and nonverbal rehearsal in children’s recall. ChiM Devel., 1969, 40, 443-452. DONALDSON, M . & STRANG,H. Primacy effect in short-term memory in young children. Psychon. Sci., 1969, 16, 59-60. ELLIS, N . R. Memory processes in retardates and normals. In N . R. Ellis (Ed.), International Review of Research in Mental Retardation (VoL 4 ) . New York: Academic Press, 1970.

Pp. 1-32.

FLAVELL,J. H . Developmental studiesof mediated memory. In H . W. Reese and L. Lipsitt (Eds.), Advances in Child Development and Behavior (Vol. 5 ) . New York: Academic Press, 1970. 11. GIBSON,E. Principles of Perceptual Learning and Development. New York: AppletonCentury Crofts, 1969. 12. HERIOT, J. T. Attention, short-term memory as learning requisites. J. Learn. Disobil., 1974, 7, 37-38. 13. HOCHHAUS, L. A table for the calculation of d‘ and B. Psychol. Bull., 1972, 77,375-376.

10.

LEE SWANSON 14. 15.

Downloaded by [York University Libraries] at 23:04 06 November 2015

16. 17. 18. 19. 20. 21. 22. 23.

127

KEELY,K.Age and task effects in short-term memory of children. Percept. & Psychophys., 1971, 9, 480-482. KEENEY,T. J., CANNIZZO, S. R., & FLAVELL,J. H. Spontaneous and induced verbal rehearsal in a recall task. Child Devel., 1967, 38, 953-966. KELLAS,G., MCCAULEY, C., 81 MCFARLAND,J. Developmental aspects of storage and retrieval. J . Exper. ChiM Psychol., 1975, 19, 51-62. KIRK, S. Ethnic differenceson psycholinguistic abilities. Except. Child. 1972,38,112-118. KOPPITZ,E. Visual Aural Digit Span Test performance of boys with emotional and learning problems. J . Clin. Psychol., 1973, 59,463466. B. The Child’s Perception of Space. New York: Humanities PLGET, J., & INHELDER, Press, 1956. ROACH,E., & KEPHART,N. The Purdue Perceptual Motor Survey. Columbus, Ohio: Merrill, 1966. RYAN,S. M., HEGION,A. C., & FLAVELL,J. H. Nonverbal mnemonic mediation in pre-school children. Child Devel., 1970, 40, 539-550. J. Unexpected reading failure. Amer. J . Ortho-Psychiat., 1972, 33, 395-414. SYMMES, VANDERPLAS, J. M., 81 GARVIN,E.A. The association value of random shapes. J. Ex*. Psychol., 1959, 57, 147-154.

School of Special Education and Rehabilitation University of Northern Colorado Greeley, Colorado 80639

Comparison of normal and learning disabled children on a nonverbal short-term memory serial position task.

The Journal of Genetic Psychology Research and Theory on Human Development ISSN: 0022-1325 (Print) 1940-0896 (Online) Journal homepage: http://www.ta...
537KB Sizes 0 Downloads 0 Views