Working Memory and Articulation Rate in Children With Spastic Diplegic Cerebral Palsy

Neuropsychology 1994. Vol. 8, No. 2, 180-186 Copyrighl 1994 hv I he American Psychological Association. Inc Working Memory and Articulation Rate in ...
Author: Scarlett Howard
2 downloads 3 Views 736KB Size
Neuropsychology 1994. Vol. 8, No. 2, 180-186

Copyrighl 1994 hv I he American Psychological Association. Inc

Working Memory and Articulation Rate in Children With Spastic Diplegic Cerebral Palsy Desiree A. White, Suzanne Craft, Sandra Hale, and T. S. Park A. D. Baddeley, N. Thomson, and M. Buchanan (1975) suggested that articulatory rehearsal rate determines the amount of verbal material that can be maintained in working memory. In the current study, 12 children with spastic diplegic cerebral palsy (SDCP) and 38 normal children were tested on measures of articulation rate and memory span for one-, two- and three-syllable words. Across all conditions, articulation rate for the SDCP group was significantly slower than for the normal group; nonetheless, memory span was equivalent for both groups. This finding implies that covert rehearsal proceeded normally for the SDCP group, in spite of decrements in speech rate. Thus, the relationship between overt and covert rehearsal rates differs for children with SDCP compared with normal children. Findings from the current study further suggest that normal speech rates are not necessary for development of normal covert rehearsal rates.

Working memory, the system by which information is held on-line during execution of a task, has been aptly described as an interface between the domains of memory, attention, and perception (Baddeley, 1992a, 1992b). Not surprisingly, a process involving integration of information from this variety of cognitive domains has also been implicated in a number of complex cognitive operations, such as reasoning (Just & Carpenter, 1992), comprehension of complex speech (Vallar & Baddeley, 1984), vocabulary acquisition (Garthecole & Baddeley, 1989), and reading (Jorm, 1983). Knowledge of the component processes underlying working memory is critical to the understanding of the workings of higher order cognitive operations. On the basis of both experimental studies and neuropsychological research on functional dissociations within the larger domain of memory, Baddeley and his colleagues developed the concept of a multicomponent working memory system (for reviews, see Baddeley, 1983). Baddeley and Hitch's (1974) original model has undergone various revisions (e.g., Salame & Baddeley, 1982; Vallar & Baddeley, 1984). As currently conceptualized, the core of the working memory system is the central executive processor of limited capacity that coordinates storage and processing operations

for at least two subsystems. One, the visuospatial sketch pad, is postulated to be responsible for short-term retention and manipulation of visual images. The other, the articulatory loop (also called the phonological loop), is postulated to be responsible for management of verbal material and is of particular relevance to the current study. The articulatory loop is thought to function as a loop of tape capable of storing approximately 2 s of verbal material (Baddeley, Thomson, & Buchanan, 1975). The rapidly decaying phonological trace that is registered in the articulatory loop may be refreshed and maintained through articulatory rehearsal. Thus, the length of time required to pronounce verbal material is the crucial parameter determining the volume that may be correctly stored and recalled in serial order (Baddeley et al., 1975; Schweickert & Boruff, 1986). That is, one may successfully maintain as much information as can be spoken in approximately 2 s. Baddeley and colleagues' model of the articulatory loop successfully accounts for a number of phenomena found in studies of working memory. One, the word-length effect, is thought to be a direct effect of the articulatory rehearsal process. This effect was demonstrated in studies showing that memory span is greater for words of shorter than longer spoken duration (Baddeley et al., 1975; Schweickert & Boruff, 1986). In contrast, both the number of chunks composing individual items (e.g., syllables or phonemes) and the semantic characteristics of memory material are much less important than duration in determining volume of recall. Studies conducted with children also support these findings and point to age-related increases in processing speed, in this case articulation rate, rather than increases in the capacity of working memory to account for developmental improvements in serial recall (Hitch & Halliday, 1983; Hulme, Thomson, Muir, & Lawrence, 1984; Nicolson, 1981). For example, Hulme et al. (1984) showed that from the age of 4 years to adulthood memory span is an ageinvariant linear function of articulation rate. That is, as articulation becomes faster with age, a proportional increase

Desiree A. White, Suzanne Craft, and Sandra Hale, Department of Psychology, Washington University; T. S. Park, Department of Neurology and Neurological Surgery, St. Louis Children's Hospital, Washington University School of Medicine. This study was conducted in partial fulfillment of the requirements for Desiree A. White's doctoral degree. The study was supported by grants from the Missouri Planning Council for Developmental Disabilities and from the McDonnell Center for the Study of Higher Brain Function. We would like to thank Madelaine R. Ortman for her assistance with scheduling subjects and Sister Beverly Reck of the St. Mary Magdalene School for allowing her students to participate. Correspondence concerning this article should be addressed to Desiree A. White, who is now at the Department of Psychiatry, University of California, San Diego, 2760 Fifth Avenue, Second Floor, San Diego, California 92103.



is observed in memory span, such that the relationship between articulatory processing speed and volume of verbal material recalled remains constant. Findings from studies of the articulatory loop have obvious implications for populations with impairments in articulation in that such impairment may disrupt short-term memory function. The present study examined the relationship between short-term phonological memory and articulation rate in children with spastic diplegic cerebral palsy (SDCP). These children frequently exhibit subtle impairments in articulation (Blasco, 1989) that might be expected to affect working memory performance. Cerebral palsy is a nonprogressive developmental disorder of movement and posture that is typically the result of cerebral insult related to premature birth and perinatal asphyxia (Blasco, 1989). Subtype classifications of cerebral palsy vary depending on the region and extent of brain injury. The most common neuropathology associated with the spastic diplegic subtype is periventricular and subcortical leukomalacia (Binder & Eng 1989; Blasco, 1989; Park, Phillips, & Torner, 1989), a bilateral white matter lesion with damage to the medial aspect of the corona radiata as it sweeps around the lateral ventricles and into the internal capsule. Additional effects include reduced corpus callosum size and ventricular enlargement. Damage to subcortical gray matter is not typical (Park et al., 1989). The pattern of brain injury in SDCP produces a characteristic complex of cognitive and motor deficits. Spasticity is a defining symptom of the disorder, primarily affecting the lower extremities. Mild impairments in cognitive function are frequently reported, as well as disruptions in speech, such as dysarthria. This latter finding raises the possibility that disrupted articulation in SDCP may interfere with the development of working memory processes. To investigate this possibility, we addressed three questions. First, are articulation deficits or short-term phonological memory impairments exhibited by children with SDCP? Second, if speech production is slowed in SDCP, how do reductions in articulatory speed relate to memory span performance? Third, is the relationship between memory span and articulation rate identical for the neurologically impaired SDCP group and a neurologically intact group with no deficit in articulation? Method

Subjects Twelve children diagnosed with SDCP were recruited from the Division of Pediatric Neurosurgery at St. Louis Children's Hospital. Only subjects with the spastic diplegic subtype of cerebral palsy were included, permitting better control of several factors, such as region and type of brain injury and cognitive and physical impairments. Clinical reading of magnetic resonance scans, which were available for 10 of the SDCP children, revealed a typical lesion pattern for 8 subjects, consisting of periventricular leukomalacia and ventricular enlargement. Two of the 10 scans showed no observable abnormalities. In addition, no SDCP child showed evidence of lesions to basal ganglia structures, a clinical feature deliberately documented as a contra-indication for a surgical pro-


cedure for which these children were being considered. Other clinical features of the SDCP group are presented in Table 1. Performances of children with SDCP were compared with those of 38 normal control (NC) children recruited from St. Louis community schools. These children had no reported history of neurological impairment or learning disability. Children in both diagnostic groups ranged in age from 4 to 11 years old. Mean ages for SDCP and NC groups were 6.92 and 7.16 years, respectively. Years of education ranged from 0 to 5 years. Mean years of education for SDCP and NC groups were 1.3 and 1.8, respectively. Fifty percent of the children in each group were male, and 50% were female. Results of t tests revealed no significant betweengroup differences on any of these demographic variables (t < 1 in all cases). In addition, all children had estimated IQs greater than 80, based on their scores from the Picture Vocabulary and Spatial Relations subtests of the Woodcock-Johnson Psycho-Educational Battery-Revised (1989).

Procedure Memory span task. Three pools of stimuli were selected, which consisted of one-, two-, or three-syllable words (see Appendix). Nine stimuli of a given word length constituted each pool, for a total of 27 words utilized in construction of memory span sequences. All stimuli were nouns chosen from Gilhooly and Logic's (1980) corpus. Each word within a given pool was matched as closely as possible to one word in each of the other two pools on the following factors: age of acquisition, familiarity, concreteness, and imageability. For each of the three word-length conditions, stimuli were drawn from the appropriate nine-word pool to construct lists that ranged from two to seven words in length. Two sequences were constructed for each list length, yielding a total of 12 sequences for each word-length condition. The one-syllable sequences were

Table 1

Clinical Features of the SDCP Group Feature

No. of subjects

Pregnancy complications None Bleeding Placenta previa Anemia Delivery complications None Caesarean section Neonatal complications None Apnea Intraventricular hemorrhage Seizures Hydrocephalus MRI findings" Periventricular/subcortical white matter lesions Reduced corpus callosum Ventricular enlargement 7 Note. Mean gestational age = 31.33 weeks. Mean birthweight = 4.12 Ib (range, 1.7-6.8 Ib). SDCP = spastic diplegic cerebral palsy; MRI = magnetic resonance imaging. a MRIs for two children were unavailable. No lesion was evident on MRI for two additional children.



compiled first, by random selection without replacement within a single list (e.g., boot-gun, bag-boot, earth-hall-land, bag-landhall, etc.). Each word was removed from the pool after being selected for inclusion in six lists. Adjustments were then made to control for the percentage of word repetition between lists of the same length and to ensure that no word occupied the same position in adjacent lists. To control for possible recall effects of the matching factors between word-length conditions, we compiled two- and three-syllable word lists by placing words in the same positions held by the matching one-syllable words. This procedure was used to construct two forms of the memory span task, using the same word pools. Analysis of variance revealed no significant effect of form on word span. All children received four practice trials using two- and threeword sequences composed of one-syliable nouns not used as test stimuli. Subjects were instructed to listen as the experimenter read each list aloud and to repeat the words in the order presented. Subjects were next familiarized with test words to ensure that recall errors during memory span trials were not due to misperceptions during presentation. This was achieved by having the children repeat each stimulus from a given word poo! after it was read aloud by the examiner, with a total of two presentations for each stimulus. Test trials were then administered to each child via auditory presentation. To maintain a constant presentation rate, the examiner read the words as they appeared on a computer screen (unseen by subjects) at a rate of one per second. Subjects attempted serial recall of each list immediately following presentation. Two trials were administered at each list length, beginning with two-word sequences, and thereafter increased until the subject failed to correctly recall both lists of a particular length. Memory span was calculated as the number of words in the longest list successfully recalled if both trials at that list length were correct. If only one trial was correct, then memory span was calculated as the number of words in the longest list minus one-half. Articulation rate task. Three words for rapid repetition were randomly chosen from the one-syllable word pool. Two- and three-syllable words were then chosen using the matching procedure described for the memory span stimuli. The following words were selected: lap, sink, earth (one-syllable); bubble, pillow, shower (two-syllable); and alphabet, animal, magazine (threesyllable). Subjects received two practice trials with words from the memory span practice trials. Rate of articulation was determined during test trials by having each subject repeat single words, as rapidly as possible, for a total of 10 repetitions per trial. This procedure was repeated twice for the three words chosen from each stimulus pool, yielding a total of 60 stimulus repetitions per word-length condition. Each repetition series was spoken into a microphone interfaced with a portable computer that recorded the time required to complete the series. Recording was triggered by voice onset and ended when the examiner pressed a computer key at voice offset. Articulation rate was determined for each word-length condition by converting the mean pronunciation duration across repetitions for a single word-length condition to number of words articulated per second.

word-length conditions were presented between subjects. Analysis of variance revealed no significant effect of order on word span or articulation rate. Children received toy sticker rewards following completion of each experimental block.

Results Word Span A Word Span X Word Length X Group repeated measures analysis of variance was performed to examine possible group differences with regard to memory span and to examine the effects of word length on memory performance. This analysis revealed a significant main effect of word length on word span, F(2, 96) = 39.14, p < .0001. No effect of diagnostic group and no Word Length X Group interaction were revealed, suggesting that for both groups word length had a similar effect on memory span. That is, as word length increased, memory span decreased to a comparable degree for SDCP and NC subjects. As shown in Table 2, the mean word span scores for the two groups were quite similar, with the SDCP group actually scoring slightly higher, though not significantly so, across all word-length conditions. Articulation Rate Analysis of variance was similarly performed for Articulation Rate X Word Length X Group. Significant main effects of group, F(\, 48) = 17.74, p < .0001, and word length, F(2, 96) = 80.37, p < .0001, were observed. As shown in Table 3, for both groups, the number of words pronounced per second decreased as words increased in syllable number. In addition, the SDCP group pronounced fewer words per second, indicating slower articulation than the NC group across all word-length conditions. A significant Word Length X Group interaction was also observed, F(2, 96) = 3.61, p < .05. This interaction indicates that the effect of word length on articulation rate differed between the two diagnostic groups. That is, the magnitude of the word-length effect on articulation rate was less for SDCP subjects than for NC subjects. Relationship Between Word Span and Articulation Rate To explore the relationship between memory span and speed of pronunciation, we plotted mean group values for Table 2 Mean Word Span Scores for One-, Two-, and ThreeSyllable Words for SPCD and Normal Control Children SPCD

Task Administration Presentation of memory span and articulation rate trials was conducted in three blocks. A single block was composed of span and articulation rate trials for a single word-length condition. A Latin Square design was used to determine the order in which

Word length One syllable Two syllables Three syllables Note. SPCD = spastic


Normal control



4.38 0.91 4.17 1.04 3.92 3.74 0.67 3.50 3.38 diplegic cerebral palsy.

SD 0.76 0.71 0.68



Table 4 Regression Values Obtained in Plotting Word Span as a Function of Articulation Rate for SDCP and NC Children

Table 3 Mean Articulation Rate Scores for One-, Two-, and Three-Syllable Words for SDCP and Normal Control Children Word length



Normal control




One syllable 1.62 0.80 2.35 0.65 Two syllables 1.23 0.42 0.40 1.91 1.32 Three syllables 0.93 0.29 0.31 Note. SDCP children pronounced fewer words per second in all three word-length conditions (p < .0001). SDCP = spastic diplegic cerebral palsy.

word span as a function of articulation rate and calculated separate regression lines for each group. As may be seen in Figure 1, the regression lines for the two diagnostic groups are quite different with respect to slope. The slopes, intercepts, and squared multiple correlations for each group are reported in Table 4. The difference in the functions apparent from inspection of Figure 1 was verified in a test for separate regressions, which revealed that the slopes of the lines were significantly different, /(2) = 3.63, p < .05. However, no significant difference was found with regard to intercept (t < 1), providing further evidence for the equivalency of working memory function between the SDCP and NC groups. As stated previously, there was no difference in word span between the two groups. The difference in the obtained regression lines is therefore entirely due to the contrasting articulation rates. Discussion The present study examined the relationship between working memory and articulation rate in children with

Articulation Rate (WPS) Figure I. Word span plotted as a function of articulation rate in each word-length condition for children with spastic diplegic cerebral palsy (SDCP) and normal controls (NC). The lines represent the best-fitting linear functions.



2.34 .9976 SDCP NC .9799 2.35 Note. Slopes of the regression lines differ significantly (p < .05). SDCP = spastic diplegic cerebral palsy; NC = normal control.

SDCP. It was anticipated that this group would exhibit subtle impairments in articulatory speed compared with normally speaking controls, which might in turn be related to a relative reduction in memory span. Findings revealed deficits in articulatory speed for the SDCP children, but a concomitant deficit in working memory was not demonstrated. Short-Term Phonological Memory Function in SDCP In the present study, SDCP children showed intact shortterm memory for auditorally presented verbal material. In addition, the word spans of SDCP and NC groups were equally affected by variations in word length. This finding suggests that both groups used rehearsal as a retention strategy (Baddeley et al., 1975). Similarly, increases in word length resulted in decreases in articulation rate for both groups. The major difference between the performances of the two groups was in articulation speed. The SDCP group demonstrated deficits in pronunciation speed across all word-length conditions, with a mean articulation rate approximately two-thirds that of the NC group. Thus, in spite of the otherwise remarkable similarities in performance, the critical relationship between memory span and articulation rate differed significantly for the diagnostic groups. The robust relationship between articulation rate and memory span that has been found in many previous studies has led some investigators to propose that the function describing this relationship is unitary and group invariant across a broad age range (Hulme et al., 1984; Nicolson, 1981). In addition, some authors have suggested that articulation rate may play a direct causal role in determining the volume of material that may be maintained for recall. For example, a study conducted with speech-disordered children revealed both impairments in memory span and a reduced word-length effect (Raine, Hulme, Chadderton, & Bailey, 1991). Raine et al. interpreted their findings as indicating that speech rate is directly related to short-term memory capacity and that overt articulation is an adequate measure of the covert rehearsal process. Conversely, some investigators have obtained results suggesting that rate of articulation is unrelated to working memory capacity. As in the current study, Bishop and Robson (1989) found no evidence for working memory impairment in normally speaking and speech-impaired (dysarthric and anarthric) children with mixed subtypes of ce-



rebral palsy. Unlike the current investigation, however, Bishop and Robson also found no relationship between articulation rate and memory span. The present study replicated previous findings of a powerful association between speech rate and memory span. The functions describing this relationship, however, were strikingly different between SDCP and NC children, indicating that this relationship is not a unitary function across all groups as suggested by previous studies (Hulme et al., 1984; Nicolson, 1981). The present findings further suggest that the determinant of working memory capacity for auditorally presented verbal information is most likely covert rehearsal. Baddeley and Wilson (1985) described covert rehearsal as "inner speech" that occurs in the absence of overt vocalization. Although covert rehearsal rate was not directly measured in the current study, the equivalency in word spans for SDCP and NC children strongly suggests that rates of covert rehearsal were equivalent for both diagnostic groups. Thus, the relationship between articulation rate and covert rehearsal speed differed for children with SDCP compared with normal children. Previous studies have also documented differences between covert and overt rehearsal rates and memory span. Standing and Curtis (1989) found subvocal rehearsal, presumably a measure of covert rehearsal, to be more strongly associated with memory span performance than were indexes of overt rehearsal. In addition, adult brain injury patients with dysarthria or anarthria, but otherwise relatively intact language function, have demonstrated unimpaired digit span and typical word-length effects (Baddeley & Wilson, 1985). This finding suggests that covert rehearsal proceeded normally in these patients in spite of difficulty or absence of speech. Baddeley and Wilson therefore proposed that covert rehearsal is reliant on central programming mechanisms for articulation rather than peripheral articulatory processes, which play a greater role in overt rehearsal. Given that central programming mechanisms are involved in both covert and overt rehearsal (with overt rehearsal entailing additional subsequent input from peripheral articulatory mechanisms), it would not be surprising if the two forms of rehearsal were related. This relationship would then be reflected in a correlation between overt rehearsal rate and word span, as was documented for both groups in the present study. The differences observed in the slopes of the articulation rate-word span functions, however, may be due to variation in peripheral articulatory mechanisms. Thus, peripheral articulatory and central programming mechanisms do not maintain a constant relationship across diagnostic groups, resulting in different relationships between covert and overt rehearsal processes in some groups with impaired articulation. An alternative interpretation of the current findings is that by virtue of their disability children with SDCP have developed a different memory strategy from that of normal children (e.g., employment of semantic processing strategies). Though this possibility cannot be entirely ruled out, the strong relationship between word length and word span, which was observed for both groups, suggests utilization of similar rehearsal processes. Thus, a more parsimonious ex-

planation for differences between the functions for the two groups is that both relied on covert rehearsal. The present study demonstrates that the relationship between articulatory speed and memory span is not invariant across diagnostic groups, and it suggests that covert rehearsal may play an important role in working memory. An additional implication of the present results is derived from the fact that children with SDCP have had slowed articulation since birth. Thus, findings of intact memory span and a typical word-length effect on memory span in this population strongly suggest that normal speech rates are not necessary for development of normal rates of covert rehearsal. Neuroanatomical Substrates of Working Memory Neilson and O'Dwyer (1981, 1984) proposed that the articulatory impairment associated with cerebral palsy is due to disruption of pathways from subcortical motor programming structures to cortical motor regions. More specifically, this disruption is the result of damage at the level of the internal capsule, as neural signals for speech programming are relayed from basal ganglia to thalamus, then ascend from ventral tier thalamic nuclei through internal capsule en route to cortical motor areas. Findings of slowed speech output for SDCP subjects are not therefore surprising given that the lesions of internal capsule and corona radiata which are typical in SDCP (Park et al., 1989) have been related to speech deficits (Neilson & O'Dwyer, 1984). However, neuroanatomical substrates for motor programming, such as the basal ganglia, are rarely compromised in children with SDCP (Park et al., 1989). In the present study, examination of magnetic resonance scans for SDCP subjects revealed no evidence of damage to basal ganglia structures, though the typical pattern of damage to subcortical white matter was observed in the majority of subjects. These findings therefore suggest that motor programming rather than motor production may play the critical role in covert rehearsal, permitting maintenance of information in the articulatory loop for both intact and certain neurologically compromised populations. Other neuroanatomical mechanisms may also contribute to preserved working memory in SDCP. For example, impairment in short-term phonological memory is frequently associated with damage to the left inferior parietotemporal region of the brain (Warrington, James, & Maciejewski, 1986; Warrington, Logue, & Pratt, 1971). When cortical damage occurs in SDCP, the lesions are typically located in more superior and anterior brain regions, sparing the area thought to subserve short-term phonological memory function. Thus, both parietotemporal cortex and basal ganglia may be components of a system subserving short-term retention and manipulation of verbal information, a system that is spared in SDCP. A final mechanism that could have permitted sparing of working memory in children with SDCP is neural reorganization. Given that brain damage resulting in SDCP occurs quite early in development, it is possible that the plasticity of the young brain may have permitted use of intact ana-

WORKING MEMORY IN CEREBRAL PALSY tomical structures to compensate for damage to short-term memory areas. Woods and Teuber (1973) demonstrated relative sparing of left hemisphere verbal function at the expense of right hemisphere nonverbal functions following developmental brain injury. Perhaps short-term phonological memory has a developmental priority similar or related to that of language. Further study of subjects with welldefined brain lesions will be necessary to definitively elucidate the neuroanatomical substrates of working memory.

Implications for Conceptualization of the Articulatory Loop Findings from the present study generally support the conceptualization of the articulatory loop developed by Baddeley and colleagues. A strong relationship between overt articulation speed and performance on tests of memory span has been repeatedly shown, and this was also the case for subjects with speech deficits associated with SDCP in the current study. However, the present results indicate that caution must be used in estimating covert rehearsal rate from overt articulatory speed. For groups with no neurological damage, articulation rate seems to provide a good estimate of the volume of information that may be maintained and manipulated within the articulatory loop. But in groups with neurological damage resulting in articulatory slowing, the relationship between overt and covert rehearsal rates may differ from normal populations. In conclusion, the results of the present study suggest that memory span is not a unitary function of articulation rate across all diagnostic groups. It is likely, however, that the relationship between covert rehearsal rate and memory span may be described by a single function across various groups. Findings of similar word spans for our two diagnostic groups support this view, implying that material was rehearsed at equivalent rates. Thus, internal processing speed, or covert rehearsal rate, rather than articulation rate appears to play the crucial role in determining the capacity of working memory.

References Baddeley, A. D. (1983). Working memory. Philosophical Transactions of the Royal Society of London, Series B, 302, 311-324. Baddeley, A. (1992a). Working memory. Science, 255, 556-559. Baddeley, A. (1992b). Working memory: The interface between memory and cognition. Journal of Cognitive Neuroscience, 4(3), 281-288. Baddeley, A. D., & Hitch, G. (1974). Working memory. In G. H. Bower (Ed.), The psychology of learning and motivation (pp. 47-89). San Diego, CA: Academic Press. Baddeley, A. D., Thomson, N., & Buchanan, M. (1975). Word length and the structure of short-term memory. Journal of Verbal Learning and Verbal Behavior, 14, 575-589. Baddeley, A., & Wilson, B. (1985). Phonological coding and short-term memory in patients without speech. Journal of Memory and Language, 24, 490-502. Binder, H., & Eng, G. D. (1989). Rehabilitation management of children with spastic diplegic cerebral palsy. Archives of Physical Medical Rehabilitation, 70, 482-489. Bishop, D. V. M., & Robson, J. (1989). Unimpaired short-term


memory and rhyme judgement in congenitally speechless individuals: Implications for the notion of "articulatory coding." Quarterly Journal of Experimental Psychology, 41A, 123-140. Blasco, A. P. (1989). Cerebral palsy: Clinical diagnosis and natural history. Neurosurgery: State of the Art Reviews, 4, 371-378. Garthecole, S. E., & Baddeley, A. D. (1989). Evaluation of the role of phonological STM in the development of vocabulary in children: A longitudinal study. Journal of Memory and Language, 28, 200-213. Gilhooly, K. J., & Logie, R. H. (1980). Age-of-acquisition, imagery, concreteness, familiarity, and ambiguity measures for 1,944 words. Behavior Research Methods and Instrumentation, 12, 395-427. Hitch, G. J., & Halliday, M. S. (1983). Working memory in children. Proceedings of the Royal Society of London, Series B, 302, 325-340. Hulme, C., Thomson, N., Muir, C., & Lawrence, A. (1984). Speech rate and the development of short-term memory span. Journal of Experimental Child Psychology, 38, 241-253. Jorm, A. F. (1983). Specific reading retardation and working memory: A review. British Journal of Psychology, 74, 311-342. Just, M. A., & Carpenter, P. A. (1992). A capacity theory of comprehension: Individual differences in working memory. Psychological Review. 99, 122-149. Neilson, P. D., & O'Dwyer, N. J. (1981). Pathophysiology of dysarthria in cerebral palsy. Journal of Neurology, Neurosurgery, and Psychiatry, 44, 1013-1019. Neilson, P. D., & O'Dwyer, N. J. (1984). Reproducibility and variability of speech muscle activity in athetoid dysarthria of cerebral palsy. Journal of Speech and Hearing Research, 27, 502-517. Nicolson, R. (1981). The relationship between memory span and processing speed. In M. P. Friedman, J. P. Das, & N. O'Connor (Eds.), Intelligence and learning (pp. 179-183). New York: Plenum. Park, T. S., Phillips, L. H., & Torner, J. C. (1989). Magnetic resonance imaging in selective dorsal rhizotomy for spastic cerebral palsy. Neurosurgery: State of the Art Reviews, 4, 485495. Raine, A., Hulme, C., Chadderton, H., & Bailey, P. (1991). Verbal short-term memory span in speech-disordered children: Implications for articulatory coding in short term memory. Child Development, 62, 415-423. Salame, P., & Baddeley, A. (1982). Disruption of short term memory by unattended speech: Implications for the structure of working memory. Journal of Verbal Learning and Verbal Behavior, 21, 150-164. Schweickert, R., & Boruff, B. (1986). Short-term memory capacity: Magic number or magic spell? Journal of Experimental Psychology: Learning, Memory, and Cognition, 12, 419-425. Standing, L., & Curtis, L. (1989). Subvocalization rate versus other predictors of the memory span. Psvchological Reports, 65, 487-495. Vallar, G., & Baddeley, A. D. (1984). Fractionation of working memory: Neuropsychological evidence for a phonological shortterm store. Journal of Verbal Learning and Verbal Behavior, 23, 151-161. Warrington, E. K., James, M., & Maciejewski, C. (1986). The WAIS as a lateralizing and localizing diagnostic instrument: A study of 656 patients with unilateral cerebral lesions. Neuropsychologia, 24, 223-239. Warrington, E. K., Logue, V., & Pratt, R. T. C. (1971). The anatomical localization of selective impairment of auditory verbal short-term memory. Neuropsychologia, 9, 377-387.



Woodcock, R. W., & Johnson, M. B. (1989). Woodcock-Johnson Psycho-Educational Battery-Revised. Allen, TX: DLM Teaching Resources.

Woods, B. T., & Teuber, H. L. (1973). Early onset of complementary specialization of cerebral hemispheres in man. Transactions of the American Neurology Association, 98, 113-117.

Appendix Word Pools Used in Compiling Lists for Word Span and Articulation Rate Tasks One-syllable word pool

Two-syllable word pool

Bag Porch Sink Hall Land Boot Earth Lap Gun Note. Words across a row are matched imageability.

Three-syllable word pool

Window Potato Cherry Umbrella Pillow Animal Summer Family Teacher Vegetable Holiday Music Shower Magazine Alphabet Bubble Elephant Lion for age of acquisition, familiarity, concreteness, and

Received February 26, 1993 Revision received June 29, 1993 Accepted June 30, 1993

Suggest Documents