Memory and rehearsal characteristics of profoundly deaf children

JOURNAL

OF EXPERIMENTAL

CHILD

PSYCHOLOGY

38, 415-428 (1984)

Memory and Rehearsal Characteristics of Profoundly Deaf Children JAMES M. BEBKO York

University

This study examined whether a deficiency in spontaneous strategy use accounts for deaf children’s verbal short-term memory performance. Various colors were presented for 3 s each, followed by a 15-s recall delay. The delay was either unfilled, or subjects were induced to rehearse or were prevented from rehearsing. Sixty-four deaf students from oral and total communication settings, 5 to 15 years of age, were tested. The spontaneous rehearsal of both deaf samples seemed to emerge later than the hearing sample’s, and it was both inefficiently implemented and less effective in mediating recall than hearing children’s However, when rehearsal was prevented or was induced in all samples, the deaf recalled as well as, or better than, the hearing. Implications discussed include the need to compare both spontaneous and controlled strategy use in developmental memory studies, and the need to provide additional training for deaf children in the strategies of remembering, as opposed to the content material. o 1984 Academic PES. IX.

Many studies have shown that deaf children’s performance on verbally oriented short-term memory tasks is lower than that of hearing children (Chen, 1974; Conrad, 1973; Heinen, Cobb, & Pollard, 1976; Wallace & Corballis, 1973). However, on tasks involving significant motor or visual components, the deaf have been found to recall as well, or even better than, hearing children (O’Connor & Hermelin, 1976; Olsson & Furth, 1966; Siple, Fischer, & Bellugi, 1977). This pattern of findings indicates that deaf children are most impaired in situations in which hearing subjects This research was supported by the Ontario Mental Health Foundation. I wish to express particular appreciation to Stephen Springer of the MacKay Centre for Deaf and Crippled Children in Montreal, Quebec, and Ellen Forrest for assistance throughout the studies; and to John Belmont and an anonymous reviewer for comments on earlier drafts. The cooperation of the following in facilitating testing of the children was also greatly appreciated: Ms. Louise Crawford, Hospital for Sick Children, Toronto; Mr. Herb Goldie, Metro School for the Deaf, Toronto; Mr. R. G. Dodds, Assistant Superintendent, E. C. Drury School, Milton, Ontario; and the parents and children themselves. This manuscript was derived in part from a doctoral dissertation completed under the supervision of Michael Friendly, York University. Requests for reprints should be sent to James M. Bebko, now at the Department of Psychology, York University, 4700 Keele Street, North York, Ontario M3J lP3, Canada. 415 0022-0965/84 $3.00 Copyright 0 1984 by Academic Press, Inc. All rights of reproduction in any form reserved.

416

JAMES

M.

BEBKO

typically use verbal mediation, such as rehearsal, which in turn suggests production and/or mediational deficiencies (e.g., Bebko, 1979; Flavell, Beach, & Chinsky, 1966). The goal of the present study was to investigate the types and the efficiency of rehearsal in deaf children’s performance on verbally oriented memory tasks. The production deficiency hypothesis has been proposed to account for a similar recall deficit among hearing children. Younger children’s (e.g., 5 and 6 years) recall of significantly fewer items than older children on serial recall tasks has been attributed to a production deficiency in the early ages (e.g., Flavell et al., 1966; Keeney, Cannizzo, & Flavell, 1967). That is, while younger samples have the stimulus labels in their repertoire, they do not produce them for rehearsal. When these nonproducing children were induced to rehearse the sequences, their recall increased very significantly (Bebko, 1979; Keeney et al., 1967; Kingsley & Hagen, 1969). Bebko’s (1979) data also suggested a small memory span difference between the producing and the nonproducing groups. Several studies have reported some limited rehearsal to be used by deaf samples. The nature of the rehearsal tended to be associated with the educational method the child experienced: manual rehearsal for those in total communication programs, fingerspelling by Rochester method students, and verbal by orally trained students (e.g., Beck, Beck, & Gironella, 1977; Wallace & Corballis, 1973). However, little data are available with deaf samples regarding either the efficiency of those strategies with regard to recall, or the pattern of emergence of spontaneous rehearsal across ages. In the existing literature limited age ranges were studied (e.g., 6-8 years by Liben & Drury, 1977; over 11 years by Wallace & Corballis, 1973); data were not broken down by age groups (Beck et al., 1977); mixed groups of profoundly deaf and hard of hearing students were tested (Beck et al., 1977); or rehearsal was possibly induced as a procedural artifact (Wallace & Corballis, 1973). The intent of the present study was to examine the emergence of rehearsal in deaf children on verbally oriented memory tasks. Several questions were of interest: (a) the form(s) of cumulative rehearsal used (if any) on a serial recall task; (b) whether rehearsal emerges in synchrony with hearing children’s vs a “production lag” of several years; (c) the efficiency of the rehearsal as measured by retail performance; (d) the number of items being actively rehearsed compared to hearing samples (i.e., “rehearsal span”); (e) whether the limited performance of deaf samples in previous studies is an artifact of the verbal orientation of those studies. METHOD

Subjects Two samples of deaf children were tested. The oral sample (educated via oral/aural methods) comprised 29 students, ages .5 years 6 months

MEMORY

AND

REHEARSAL

OF

DEAF

CHILDREN

417

to 14 years 1 month, mean 10 years 4 months. The total communication sample (educated via oral/aural + fingerspelling + manual signs + speechreading methods) comprised 34 students, ages 6 years 7 months to 1.5 years 2 months, mean 11 years 10 months. Every student met the following criteria: age 5 to 15 years; normal or above normal measured intelligence (standardized test scores 3 85 in school files); no known emotional, behavioral, or additional medical dysfunctions; hearing loss of 2 85 dB in the better ear; hearing loss endogenous (Myklebust, 1966) or, alternatively, of unknown etiology (i.e., deafness not known to be secondary to disease or other nonhereditary factors). Materials

The stimulus cards were 15 x 20-cm pieces of colored paper mounted on 20 x 30-cm neutral grey background cards, one color per card. Duration of stimulus exposure was controlled by the experimenter in response to an electronic timer that clicked at preset intervals. Subjects responded by arranging 5 x 5 x OS-cm response blocks that corresponded in color to the stimulus cards. The eight stimulus colors were those readily recognized and labeled by 5-year-old children (Bebko, 1979): red, blue, green, yellow, purple, white, orange, and pink. In addition, a black and a brown response block were provided as distracters. Stimuli were randomly ordered within sequences; no color occurred more than once per sequence. Procedure

All subjects were tested individually. The basic procedure was adapted from Bebko (1979). To avoid lack of understanding of verbal or written instructions (Liben, 1979; Moores, 1978; Piaget, 1966), all instructions were given by demonstrations and by gesture (not signs). After subjects labeled each stimulus color (verbally, by sign, or by writing), several trials were demonstrated. Two practice trials followed these demonstrations. In all trials stimuli were shown sequentially for 3 s each, then were turned face down on a single stack. Following the last stimulus, an unfilled delay of 15 s was timed; response blocks were then uncovered as a signal to respond. A correct response consisted of arranging the response blocks from the subject’s left to right in the order the stimuli had been shown. There was no time limit, and subjects were congratulated regardless of the correctness of the response. Base condition. Following the practice trials, all subjects were given two sets of trials, each consisting of one trial of 3, 4, and 5 colors, the trials arranged in ascending length. Thus, a total of six trials were given in this condition to each subject. As in previous studies, subjects were carefully observed during stimulus presentation and the delay period for any evidence of spontaneous cu-

418

JAMES M. BEBKO

mulative rehearsal. Such behaviors included rehearsing (for example, “blue, red, green; blue, red, green; blue, red, green”) verbally or by sign, repeated rhythmic head or body movements with some color names, etc. Subjects were accordingly classified as “Producers” if they spontaneously produced a cumulative rehearsal strategy during presentation or delay on a minimum of two trials, or as “Nonproducers” otherwise. Also the modality of rehearsal, whether verbal or manual, was recorded. Where classification was unclear students were asked (verbally for oral children; manually and verbally for total communication children) how they remembered the colors. Classification as Producer, then, was based on rehearsal having been observed or reported. Thus, potential errorsi.e., classifying a Producer as a Nonproducer (or vice versa)-would work against the hypothesis of Producers’ performance being superior to Nonproducers’. Following the base trials, Producers and Nonproducers were assigned alternately to either the interpolated task condition or the teaching condition. Therefore, each subject participated in two conditions: the base condition (all subjects), and either the teaching condition or the interpolated task condition (half of each group). Interpolated task condition. In this condition, the 15-s delay period prior to recall was filled with a counting task to prevent rehearsal. After the last stimulus was presented, a two-digit number was shown and the student counted on paper as rapidly as possible from that number, forward by threes for the older students (8 years and older), backward by ones for the younger students. Following recall the student counted the numbers he/she had written and a conspicuous recording was made to encourage more rapid counting on subsequent trials. Six trials were given in total. On the first trial, three colors were shown; subsequently, a staircase method was used. If recall was correct in the preceding trial, the next stimulus sequence was increased by one color; if recall was incorrect, the next sequence was decreased by one color, to a minimum of two colors. Teaching condition. The way the Producers in each sample rehearsed (manually, verbally, etc.) determined how the others in that sample were taught to rehearse. It was demonstrated to each child to label each color as shown, and to rehearse immediately that color, plus all the preceding colors on that trial twice, in overt cumulative fashion. A time limit of 15 s was imposed for presentation of the new stimulus and completion of the rehearsal. If the child was successful, another color was shown, and the new sequence was rehearsed twice, cumulatively. This continued until the child failed to rehearse the sequence twice through successfully (or until a limit of eight stimuli had been presented). No recall with response blocks was required. When the child was finally unsuccessful, in order to minimize the

MEMORY

AND REHEARSAL

OF DEAF CHILDREN

419

possibility that it was simply a verbal or manual miscue, e.g., “tonguetwisting” or “slip of the hand, ” which caused the mistake, for each new stimulus card, if required, one verbal or manual cue of the correct sequence was given. If the child was then successful in repeating this sequence twice, the trial continued; if not, the trial was terminated. In this way a more accurate measure of what might be termed the subject’s “rehearsal span” was determined, since corrections should not affect underlying capacity. Six of these trials were administered. RESULTS

Ages of Transition Table 1 shows the number of Producers and Nonproducers compared to hearing children in Bebko (1979). Since most of the Producers rehearsed overtly, classification was a relatively simple task. With regard to the first question, the style of the children’s rehearsal tended to reflect their educational milieu. When children in the total communication programs rehearsed, their rehearsal consisted of the manual signs corresponding to the color names, or the signs and accompanying TABLE NUMBER

OF CHILDREN AND

Grade

1

CLASSIFIED

NONPRODUCERS

Age

AS PRODUCERS BY AGE

Nonproducers

Producers

Hearing“ K 1 2 3 4 Total

5-6 6-7 7-8 8-9 9-10

17 27 7 0 0 51

2 9 18 29 26 84

Total

Oral deaf 3 8 0 4 0 15

2 2 3 5 2 14

Total

Total communication <8 2 8-9.11 4 10-11.11 6 12-13.11 6 14+ 1 19

-3 8-9.11 10-11.11 12-13.11 14+

a From Bebko (1979).

deaf 1 0 1 9 4 15

420

JAMES

M.

BEBKO

verbalizations, often unintelligible. When the children in the oral programs rehearsed, their rehearsal was generally verbal in nature, consisting of varying approximations of the color names or the vowel phoneme contained in the name (e.g., the /i/ of green; the /u/ of blue). Several additional idiosyncratic strategies were observed (counting the colors on the fingertips, placing a stimulus in an imaginary location on the table top and touching that location during rehearsal) or reported (encoding the colors by their relative brightness; e.g., pink > red > purple). The second question centered on the age at which spontaneous rehearsal emerges in deaf children. Adopting as a criterion the earliest age at which 50% or more of the children of that age rehearse spontaneously (Bebko, 1979; Flavell et al., 1966), Table 1 shows rehearsal emerging at age 78 years in hearing children, but at lo-11 years for the oral deaf group, and 12-13 years for the total communication group. The difference between the deaf samples must be interpreted cautiously, as there were only three subjects at the critical age range for the oral group, ages lo-11 years. Base Condition

The mean recall levels by groups and conditions appear in Table 2. Two dependent variables were examined: mean percentage trials of correct recall and mean maximum sequence length correct. The percentage recall measure represents the proportion of trials in which subjects recalled all stimuli in their correct serial positions. The maximum sequence length (MSL) is the longest sequence recalled correctly. A default MSL of two was assigned to any child who failed to recall even the shortest sequences of colors (three). To assess the effectiveness of the spontaneous rehearsal of the deaf (question 3), the two recall measures for the base condition were analyzed separately in two 3 (group: hearing, oral, total communication) x 2 (producer classification: Producers, Nonproducers) x 2 (second condition: teaching, interpolated task) analyses of variance. Second condition was included as a check for the randomness of assignment of subjects to the teaching and interpolated task conditions; there were no significant differences in any of the tests involving second condition. Percentage correct recall. The analysis of the base percentage correct data, arcsine transformed, revealed significant group and producer main effects (F(2, 103) = 9.76, p < .OOOl; F(1, 103) = 133.54, p < .OOOl, respectively). The producer difference demonstrates the significant positive effect of rehearsal. Duncan multiple range comparisons among groups revealed that the hearing sample’s recall was significantly greater than that of the two deaf groups (p < .05), who did not differ. The Group x Producer interaction was also significant (F(2, 103) = 4.21, p < .02). A comparison of means indicated that while Nonproducers did not differ, hearing Producers recalled more (p < .05) than either the

MEMORY

AND

REHEARSAL TABLE

PROPORTION AND

TRIALS

MEAN

(MSL)

OF DEAF

421

2

OF CORRECT

MAXIMUM

CHILDREN

RECALL

(IN %I

SEQUENCELENGTHS

RECALLED BY THREE OF PRODUCERS AND

SAMPLES

NONPRODUCERS ACROSS CONDITIONS

Nonproducers %

MSL

Producers %

MSL

Base Teach Interpolated Sample means

Hearing” 2.7 14.4 5.9 1.9 3.3

80.6 -

4.8 6.8 3.1 4.9

Base Teach Interpolated Sample means

Oral deaf 3.1 23.3 7.1 2.9 4.1

54.8 -

4.2 8.0 3.0 4.9

Total communication Base 14.0 2.8 Teach 6.2 Interpolated 2.6 Sample means 3.6

deaf 57.8 -

4.1 8.0 3.5 4.9

a From Bebko (1979).

oral or total communication Producers; no differences were found between deaf samples. While children who rehearsed recalled correctly on more trials than those who did not, the magnitude of the advantage was greater for hearing Producers than for deaf Producers. A count of “borderline Producers” -those who rehearsed spontaneously on exactly two trials-indicated that 21% of the deaf Producers (6 of 28) met the minimum criterion, while only 7% of the hearing Producers (3 of 43) did so. Thus one reason for the deaf Producers’ poorer recall seems to have been unreliable strategy implementation. Conditional probabilities of correct recall following rehearsal, calculated for each sample, were hearing = .84; oral = .49; total communication = S4. The differences between the hearing and the deaf samples’ probabilities (both z 3 4.19, p < .OOl) indicate that the rehearsal that was used was not as strong a predictor of correct recall for the deaf as for the hearing. Maximum sequence length. The analysis of the base MSL yielded significant group and producer main effects (F(2, 103) = 4.67, p = .Ol; F(1, 103) = 107.26, p < .OOl, respectively). The group x producer

422

JAMES M. BEBKO

interaction only approached significance (F(2, 103) = 2.78, p = .07). Individual comparisons by group indicated hearing children’s recall to be slightly better than either deaf group’s. Serial position curves. Conditional error probabilities by serial position (e.g., Huttenlocher & Burke, 1976; McCrary & Hunter, 1953) were compared for the three groups in the base condition in Fig. 1. (A strict recall criterion was used, where only items recalled in correct position were credited.) To reduce noise, data were collapsed across sequence lengths, yielding three data points. The middle point represented the mean for all nonfirst and nonlast positions combined, thus yielding a chance expectation equivalent to the first and last positions. Linear and quadratic trend components for individual curves were examined in two separate 3 (group) x 2 (producer) analyses of variance. There were no differences in linear trend (all p > .20); the producer main effect attained a high degree of significance in the test of the quadratic component (F(2, 113) = 28.62, p < .OOOl). The latter is a result of Producers making fewer errors in the initial position and more in the middle positions, consistent with the assumed effect of rehearsal (e.g., Hagen & Kail, 1973; Hoving, Konick, & Wallace, 1975). Teaching and Interpolated

Task Conditions

Maximum sequence length. The overall means for the teaching and interpolated task conditions appear in Table 2. Due to the staircase method the default sequence length in the interpolated task condition was one. (Percentage correct measures were inappropriate to the procedures of both conditions.) The MSL data were analysed as in the base condition. The highly significant main effect for second condition (F(l) 103) = 354.74, p < .OOOl), was a function of the anticipated large difference in recall performance when rehearsal is induced vs prevented. However, of greater interest is the lack of a significant interaction between each

P 8’60 (a) e

PRODUCERS

(b) NON-PRODUCERS

‘O

a 6o E 50 LL 0 40 g .30 k:: a.

0

FIRST SERIAL

MID

LAST

POSlTlON

SERIAL

POSITION

Proportion of total errors per serial position collapsed across sequence for Producers (a) and Nonproducers (b) in base condition. FIG.

1.

MEMORY

AND REHEARSAL

OF DEAF CHILDREN

423

of the other terms and second condition. These nonsignificant interactions indicate that a residual recall advantage by the Producers over the Nonproducers (F(1, 103) = 31.59, p < .OOOl) remains, whether rehearsal is induced in all groups, or is prevented via interference. Comparison among the Table 2 values for each group reveal an approximate one-item difference, consistent with the finding in Bebko (1979). All other tests were not significant. The nonsignificant group main effect suggests that sequence length differences between hearing and deaf samples in the teaching and interpolated task conditions are inconsequential. However, potential differences in the teaching condition may be masked by ceiling performance, particularly among deaf Producers. Regression curves (teaching condition). To examine rehearsal span more directly (question 4) the proportion of trials of correct recall per sequence length for the two deaf samples are compared with the hearing sample’s data in Fig. 2. From ceiling performance at the shorter lengths, there is a generally increasing recall advantage for the deaf children with increases in sequence length. Regression parameters and means for each curve were analyzed via an analysis of variance to determine whether the lines were homogeneous or distinct (e.g., Snedecor & Cochran, 1969). This clarified the nature of the ceiling effects in Table 2. Producers’ curves were significantly above Nonproducers’ (F(1, 24) = 103.79, p < .OOOl) and were flatter in slope (F(1, 24) = 6.04, p < .02); the slope differences may have been in part due to the ceiling effects at shorter lengths. Both deaf groups significantly outperformed the hearing group (p < .OOl), their advantage increasing with sequence length, resulting in flatter slopes (both p < .02). Interactions were not significant.

PRODUCERS NON-PROD -HEARING*---a MORAL D----C I TOTAL b---a COMMUN.

LENGTH

OF

SEOUENCE

FIG. 2. Recall of two samples of deaf Producers and Nonproducers according to sequence length in teaching condition. (Hearing data from Bebko, 1979.)

424

JAMES M. BEBKO

Serial position curves (interpolated task condition). The conditional error probabilities x position for the interpolated task condition are shown in Fig. 3. The proportion of errors made across positions is remarkably similar for Producers and Nonproducers across samples. Analysis of the linear and quadratic trends revealed no differences among any of the main effects or their interactions (all F < 1.80, p > . 10). The performance of both Producers and Nonproducers is nearly identical to that of the Nonproducers in the base condition, as primacy effects have been reduced. The interpolated task was clearly effective in preventing rehearsal. DISCUSSION

Similar to these data, Liben and Drury (1977) had found limited evidence (5 or 6 of 20 children) of spontaneous cumulative rehearsal in 6- to 8year-old children. Based on the number of trials where rehearsal was observed, at least half of these were “borderline” rehearsers; further, no correlation was found between use of strategy and correct recall. The wider age ranges of the two samples in this study clarify further the nature of rehearsal in the deaf. First, students in oral programs tended to rehearse verbally; those in total communication environments tended to rehearse using manual signs. Moreover, there was an equivalence in the effectiveness of these forms of rehearsal. The second question concerned the pattern of emergence of the deaf children’s strategies. A pattern similar to hearing children’s was observed. Furthermore as with the hearing children, Nonproducers were characterized by a production deficiency which accounted, in large part, for their poorer recall relative to their rehearsing counterparts. While the patterns were similar, an apparent lag of several years in NON- PROD

PRODUCERS .6Or

0’

-HEARING@--a

MID

FIRST SERIAL

FIG.

3. Proportion

LAST

POSITION

of total errors per serial position in interpolated task condition.

MEMORY

AND

REHEARSAL

OF DEAF

CHILDREN

425

the rate of emergence of spontaneous rehearsal for the deaf was found. While not an unexpected outcome, different lags were suggested for the two deaf samples. A number of hypotheses have been proposed to explain operatory lags in deaf children (e.g., Furth, 1964; Moores, 1978; Springer, 1977). The delays observed in the present study may be a function of linguistic and educational experience. While early identification is becoming more frequent, diagnosis and language intervention for the deaf has typically not occurred until approximately 3 years of age (Meadow, 1978). This initial delay and the continued impairment in subsequent language training (Liben, 1978) may account for a large part of the 3-year lag observed here. A further lag for the total communication sample might be similarly expected due to frequent initial attempts at oral programs prior to effective total communication training. This suggests an interesting area for further study, as the present sample size in the critical age range was limited. Interestingly, one student tested was a deaf child of deaf parents, whose training in total communication had begun at a very early age. He was the only Producer in the total communication sample who was less than 10 years old, his age (6 years 7 months) corresponding closely to the mean age of transition to spontaneous rehearsal for the hearing sample. While only a single subject, his data is consistent with a linguistic experience interpretation. The third and fourth questions, respectively, concerned the efficiency of the deaf samples’ rehearsal strategies in terms of recall, and an assessment of the memory spans of the deaf children. In the case of spontaneous rehearsal (base condition), the deaf Producers recalled more and longer sequences than the Nonproducers, and were more accurate in the early serial positions (consistent with Kingsley & Hagen, 1969, and others). Liben and Drury (1977) also found primacy effects for deaf and hearing children, but most were not rehearsing. However, the nature of their task involved a larger spatial component which has been shown to contribute to primacy (Siegel, Allik, & Herman, 1976); with the greater temporal component in this study, rehearsal plays a more central role. These rehearsal effects indicate an equivalence of function in the rehearsal strategies of the deaf and the hearing, similar to Hoemann, Andrews, and DeRosa’s (1974) inferred equivalence for categorical encoding and Liben’s (1979) for clustering and organization. However, as in earlier studies, deaf samples recalled less when strategies were used spontaneously. This contrasts with equal or superior performance when rehearsal was controlled in the teaching and interpolated task conditions. Indeed, although consistent statistical differences were not found, deaf subjects outperformed hearing in all but the spontaneous situation. These data suggest that the rehearsal and memory spans of deaf children are at least equivalent to hearing children. The poorer performance in the base condition appears to have been

426

JAMES M. BEBKO

due to unreliable and inefficient spontaneous rehearsal. The higher proportion of deaf “borderline” Producers supports the conclusion of Belmont, Karchmer, and PiIkonis (1976) who attributed the deaf’s recall deficiencies to a “disinclination to use effective information-processing strategies” spontaneously (p. 46). But unreliable strategy implementation was clearly not the only difference between the samples. The conditional probabilities of correct recall when rehearsal was observed (54 for the deaf; .84 for the hearing) suggest an additional problem in the efficiency of the spontaneous rehearsal. However, the specific nature of the difficulty (e.g., abandoning rehearsal too soon prior to recall, altering the string of labels being rehearsed) remains an open question. A number of implications may be drawn which relate to previous studies (question 5). Those which have shown deficient recall in the deaf should be reassessed (e.g., Hermelin & O’Connor, 1975; Myklebust, 1966; Wallace and Corballis, 1973); most used stimuli (e.g., digits, letters) and tasks which are verbally oriented in nature. The poorer recall by the deaf in these studies is probably a function of inefficient strategy usage. Some researchers have attempted to bias stimuli appropriately for both hearing and deaf samples by using visually presented manual signs for the deaf subjects (Bellugi, Klima, & Siple, 1975; Liben & Drury, 1977). However, the results of the present study illustrate the need for a comparative assessment of the contribution of strategy use both in spontaneous and in well-controlled instructed situations. Educationally, several implications may be noted. Deaf children appear to be able to retain information as effectively as (possibly better than) hearing children once the information has been efficiently encoded. Therefore, additional attention should be focused on the processes of remembering (i.e., the how to remember) as opposed to the content (the what to remember). In support, MacKay (1979) reported significant generalization of a similar procedure to new materials, and further, that more than 50% of the trained samples maintained the strategy 10 days later. It appears, then, that additional training in strategy implementation would yield significant long-term benefits for the child. Finally, if the linguistic experience account of the lag in the emergence of spontaneous rehearsal of deaf populations is supported subsequently, then the importance of early effective language intervention is strongly reinforced. The results of this study suggest that for memory processing skills, the modality of that training would be less important for the child than the timing of the intervention. REFERENCES Bebko, J. M. (1979). Can recall differences among children be attributed to rehearsal effects? Canadian Journal of Psychology, 33, 96-105.

MEMORY

AND REHEARSAL

OF DEAF CHILDREN

427

Beck, K., Beck, C., & Gironella, 0. (1977). Rehearsal and recall strategies of deaf and hearing individuals. American Annals of the Deaf, 122, 544-552. Bellugi, U., Klima, E., & Siple, P. (1975). Remembering in signs. Cognition, 3, 93-125. Belmont, J. M., Karchmer, M. A., & Pilkonis, P. A. (1976). Instructed rehearsal strategies’ influence on deaf memory processing. Journal of Speech and Hearing Research, 19, 36-47.

Chen, K. (1974). Free recall learning of deaf and hearing subjects. Journal Psychology,

of General

91, 155-156.

Conrad, R. (1973). Some correlates of speech coding in the short-term memory of the deaf. Journal of Speech and Hearing Research, 16, 375-384. Flavell. J. H., Beach, D. R., & Chinsky, J. M. (1966). Spontaneous verbal rehearsal in a memory task as a function of age. Chi/d Development, 37, 283-299. Furth, H. G. (1964). Conservation of weight in deaf and hearing children. Child Development, 35, 143-150.

Hagen, J. W., & Kail, R. V. (1973). Facilitation Child

Development,

and distraction in short-term memory.

44, 831-836.

Heinen, J. R., Cobb, L., & Pollard, J. W. (1976). Word imagery modalities and learning in the deaf and hearing. Journal of Psychology, 93, 191-195. Hermelin, B. M., & O’Connor, N. (1975). The recall of digits by normal, deaf and autistic children. British Journal of Psychology. 66, 203-209. Hoemann, H., Andrews, C., & DeRosa, D. (1974). Categorical encoding in short-term memory by deaf and hearing children. Journal of Speech and Hearing Research, 17, 426-431.

Hoving, K. L., Konick, D. S., & Wallace, J. (1975). Memory storage and retrieval within and across modalities in children. Journal of Experimental Child Psycho/ogy, 19, 440447.

Huttenlocher,

J., & Burke, D. (1976). Why does memory span increase with age? Cognitive 8, l-31. Keeney, T. J., Cannizzo, S. R., & Flavell, J. H. (1967). Spontaneous and induced verbal rehearsal in a recall task. Child Development, 38, 953-966. Kingsley, P. R., & Hagen, J. W. (1969). Induced versus spontaneous rehearsal in shottterm memory in nursery school children. Developmental Psychology, 1, 40-46. Liben, L. S. (1978). The development of deaf children: An overview of issues. In L. S. Liben (Ed.), Deaf children: Developmental perspectives, New York: Academic Press. Liben, L. S. (1979). Free recall by deaf and hearing children: Semantic clustering and recall in trained and untrained groups. Journal of Experimental Child Psychology, 27, 105-l 19. Liben, L., & Drury, A. (1977). Short-term memory in deaf and hearing children in relation to stimulus characteristics. Journal of Experimental Child Psychology, 24, 60-73. MacKay, E. R. (1979). Generalization of instructed rehearsal strategies in deaf children. Unpublished master’s thesis, University of Manitoba, Winnipeg. McCrary, J., & Hunter, W. (1953). Serial position curves in verbal learning. Science, 117, 131-134. Meadow, K. P. (1978). The “natural history” of a research project: An illustration of methodological issues in research with deaf children. In L. S. Liben (Ed.), Deaf children: Developmental perspectives. New York: Academic Press. Moores, D. (1978). Currect research and theory with the deaf Educational implications. In L. Liben (Ed.), Deaf children: Developmental perspectives. New York: Academic Press. Myklebust, H. R. (1966). The psychology of deafness: Sensory deprivation, learning and adjustment (2nd ed.). New York: Grune & Stratton. O’Connor, N., & Hermelin, B. M. (1976). Backward and forward recall by deaf and hearing children. Quarterly Journal of Experimental Psychology, 28, 83-92. Psychology,

JAMES M. BEBKO

428

Olsson, J. E., & Furth, H. C. (1966). Visual memory span in the deaf. American Journal of Psychology, 79, 480-484. Piaget, J. (1966). Surdi-Muite et conservations operatoires. Erudes d’Ep&emologie Genetique, 20, 61-64. Siegel, A. W., Allik, J. P., & Herman, J. F. (1976). The primacy effect in young children: Verbal fact or spatial artifact. Child Development, 47, 242-247. Siple. P., Fischer, S. D., & Bellugi, U. (1977). Memory for nonsemantic attributes of American Sign Language signs and English words. Journal of Verbal Learning and Verbal Behavior, 16, 561-574. Snedecor, G. W., & Cochran, W. G. (1969). Statistical methods, Ames, IA: Iowa State Univ. Press. Springer, S. (1977). A study of the performance of deaf and hearing subjects on Piagetian and Neo-Piagetian tasks. Unpublished PhD Dissertation, York University, Downsview, Ontario. Wallace, G., & Corballis, M. C. (1973). Short-term memory and coding strategies in the deaf. Journal of Experimental Psychology. 99, 344-348. RECEIVED:

March

8,

1983;

REVISED:

February 13, 1984