Journal of Child Psychology and Psychiatry 46:3 (2005), pp 304–316

doi: 10.1111/j.1469-7610.2004.00352.x

Serial order reconstruction in Down syndrome: evidence for a selective deficit in verbal short-term memory Jon Brock and Christopher Jarrold University of Bristol, UK

Background: Individuals with Down syndrome consistently perform less well than appropriately matched comparison groups on tests of verbal short-term memory, despite performing relatively well on non-verbal short-term memory tasks. However, it is not clear whether these findings constitute evidence for a selective deficit in verbal short-term memory, or whether they instead reflect the influence of noncentral factors such as speech difficulties or poor number knowledge. Methods: Twenty-six individuals with Down syndrome and 32 typically developing children were tested on a digit reconstruction task in which participants were presented with auditory digit sequences and responded by pressing the corresponding digits on a touch-screen in the correct serial order. Background measures were performance on a closely matched visuo-spatial reconstruction task, reaction time on a simple digit identification task, receptive vocabulary age and non-verbal ability (Raven’s matrices). Participants were also tested on a conventional digit recall task. Results: All four background measures accounted for significant individual variation in digit reconstruction performance, but there remained a significant effect of group that reflected relatively poor performance of individuals with Down syndrome. Hierarchical regression showed that group membership accounted for unique variation in both digit reconstruction and recall performance, even after all group differences on background measures had been accounted for. Conclusions: The results provide strong evidence that Down syndrome is associated with a selective deficit in verbal short-term memory, and a deficit in verbal serial order memory in particular. Implications for the language difficulties associated with Down syndrome are discussed. Keywords: Down syndrome, verbal and non-verbal short-term memory, serial order memory, speech and language difficulties. Abbreviations: BPVS: British Picture Vocabulary Scale II; DS: Down syndrome; TD: typically developing; RCPM: Raven’s Coloured Progressive Matrices; RT: reaction time.

Down syndrome affects around 5 in 10,000 of the general population (Steele & Stratford, 1995), and accounts for almost 40% of all cases of moderate to severe mental retardation (Pennington, Moon, Edgin, Stedron, & Nadel, 2003). Although Down syndrome is often seen as resulting in general across-the-board mental retardation, there is consistent evidence that language abilities, and expressive language abilities in particular, are even poorer than might be expected given overall abilities or IQ (see, e.g., Chapman, 1997; Fowler, 1990). A particularly striking feature of Down syndrome is the poor performance of individuals with the syndrome on tests of verbal short-term memory where participants are required to repeat a list of digits or words in correct serial order. For example, participants with Down syndrome perform more poorly on digit recall tasks than younger typically developing children matched on measures of receptive vocabulary (Jarrold & Baddeley, 1997; Jarrold, Baddeley, & Phillips, 2002; Kay-Raining Bird & Chapman, 1994; Mackenzie & Hulme, 1987; Marcell & Weeks, 1988; Marcell, Harvey, & Cothran, 1988) or overall mental age (McDade & Adler, 1980; Pennington et al., 2003). They have also been found to show inferior digit recall performance compared with learning-disabled controls matched on vocabulary knowledge (Jarrold

& Baddeley, 1997; Jarrold et al., 2002; Marcell & Weeks, 1988; Marcell et al., 1988; but see Mackenzie & Hulme, 1987), non-verbal ability (Numminen, Service, Ahonen, & Ruoppila, 2001), or chronological age and overall IQ (Marcell & Cohen, 1992; but see Varnhagen, Das, & Varnhagen, 1987). A number of studies have also compared performance on digit recall tasks with performance on the Corsi blocks test (Milner, 1971) – a test of visuospatial short-term memory in which the experimenter taps a number of wooden blocks in sequence and the participant is then required to tap the same blocks in the correct order. The majority of these studies have reported that, whereas individuals with Down syndrome perform more poorly than controls on digit recall tasks, on the Corsi blocks task, group differences are non-significant or the Down syndrome group actually perform significantly better than controls (Jarrold & Baddeley, 1997; Jarrold et al., 2002; Laws, 2002; Numminen et al., 2001; see also Jarrold, Baddeley, & Hewes, 1999a; Pennington et al., 2003). These results suggest that short-term memory deficits in Down syndrome are much more severe in the verbal than in the non-verbal domain. Such findings could have important implications for our understanding of the language difficulties associated with Down syndrome (see Jarrold, 2001).

Ó Association for Child Psychology and Psychiatry, 2004. Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA

Short-term memory in Down syndrome

According to the working memory model (Baddeley, 1986), performance on serial recall tasks depends on the functioning of a specialised system – the phonological loop – that has evolved in humans to play an important role in language acquisition (cf. Baddeley, Gathercole, & Papagno, 1998). A number of authors have therefore argued that Down syndrome is associated with a specific impairment of the phonological loop (e.g., Broadley, MacDonald, & Buckley, 1995; Hulme & MacKenzie, 1992; Jarrold & Baddeley, 1997; Kay-Raining Bird & Chapman, 1994; Vicari, Marrotta, & Carlesimo, 2004; Wang & Bellugi, 1994), and that this can potentially account for some of the language difficulties associated with the syndrome (Jarrold, Baddeley, & Phillips, 1999b; Laws, 1998; see also Chapman, 1995). An alternative view, however, is that serial recall performance does not depend on any specialised system but, rather, reflects the operation and activation of language- and, in particular, speech-processing mechanisms (e.g., Gathercole & Martin, 1996; Gupta & MacWhinney, 1997; Martin & Saffran, 1997). Poor verbal shortterm memory performance in Down syndrome may therefore be seen as a consequence rather than the cause of speech and language difficulties (Hulme & Roodenrys, 1995). Nevertheless, such a theoretical position implies that determining the causes of poor serial recall performance will provide an insight into the underlying mechanisms that lead to speech and language difficulties in Down syndrome. However, before any theoretical claims can be made regarding the causal relationship between language and verbal short-term memory difficulties in Down syndrome, it is first necessary to rule out a number of alternative explanations for poor task performance. For example, hearing difficulties are common in Down syndrome (see, e.g., Dahle & McCollister, 1986; Limongi, Carvallo, & Souza, 2000; Marcell & Cohen, 1992; Welsh & Elliot, 2001) and could potentially lead to problems on digit recall tasks when material is presented auditorily. In fact, a number of studies have failed to find any association between digit recall and hearing or speech discrimination abilities (Jarrold & Baddeley, 1997; Laws, 1998; Marcell & Cohen, 1992; Seung & Chapman, 2000). Moreover, presenting verbal material in a visual format (Broadley et al., 1995; Laws, MacDonald, & Buckley, 1996; Marcell & Armstrong, 1982; Marcell & Weeks, 1988; Varnhagen et al., 1987) or presenting digits visually at the same time as presenting them auditorily (Jarrold et al., 2002) has not been found to significantly improve the recall of individuals with Down syndrome. Clearly, severe hearing difficulties will inevitably lead to poor recall of auditorily presented digits, but there is little evidence that the kind of mild to moderate hearing loss that is common in Down syndrome has any measurable effect on digit recall performance. Two further potential confounds are that Down syndrome is associated with speech difficulties (see,

305

e.g., Dodd 1975; Dodd & Thompson, 2001; Gibson, 1978), and with poor number knowledge (Gelman & Cohen, 1988; Nye, Fluck, & Buckley, 2001). Speech difficulties may lead to slower output rates, and the resulting increase in trace decay for later items in the list (cf. Cowan et al., 1992; Dosher & Ma, 1998) could potentially be the cause of poor digit recall. Similarly, verbal short-term memory task performance depends on familiarity with the stimuli (e.g., Brener, 1940; Hulme, Maughan, & Brown, 1991), so poor number knowledge may lead to poor digit recall (Laws, 1998; cf. Tehan & Lalor, 2000; Thorn & Gathercole, 2001). Jarrold et al. (2002) addressed these two issues by using a serial order recognition task in which participants listened to two presentations of the same set of digits and were then required to determine whether the items were in the same order in the two lists. Speech output demands were thus minimised. Moreover, this task provides a test of order as opposed to item memory (cf. Healy, 1974), and because stimulus familiarity primarily influences memory for the items in the list rather than their order (Brock, McCormack, & Boucher, 2004; Gathercole, Pickering, Hall, & Peaker, 2001; Saint-Aubin & Poirier, 1999; Turner, Henry, & Smith, 2000), performance on this task should have been relatively unaffected by individuals’ familiarity with numbers. However, the results were somewhat inconclusive: on the one hand, individuals with Down syndrome performed significantly worse than vocabulary-matched controls on this task, and the use of this recognition task did not result in a significant improvement in the performance of individuals with Down syndrome relative to their performance on a conventional digit recall task. This suggests that speech output demands and number familiarity did not explain their poor digit recall. On the other hand, the Down syndrome group did not perform significantly worse on the digit recognition task than on an analogous ‘Corsi recognition’ task (i.e., there was no longer evidence that their short-term memory deficit was more severe in the verbal domain). As Jarrold et al. admitted, the absence of significant effects on these recognition tasks could simply reflect a reduction in the sensitivity of these tests to individual differences in short-term memory ability. Consequently, it is difficult to be sure of the extent to which speech difficulties and number familiarity contribute to poor verbal short-term memory performance in Down syndrome. To summarise, studies have consistently shown that the performance of individuals with Down syndrome on verbal short-term memory tasks in general, and digit recall tasks in particular, is poorer than that of controls matched on receptive vocabulary or measures of overall or non-verbal mental age, and that deficits in digit recall are generally greater than those for non-verbal short-term memory tasks. This pattern of performance does not appear to be a

306

Jon Brock and Christopher Jarrold

consequence of hearing difficulties, but it is unclear at present whether speech output difficulties or relatively poor number knowledge are factors that contribute to this level of poor performance. These issues were addressed in the current study by investigating the performance of individuals with Down syndrome on a conventional digit recall task and a novel digit reconstruction task. In the latter task, participants heard a list of numbers before seeing the digits on a touch-screen and having to touch them in the correct order. Only the relevant digits were presented on the screen, and the digits were removed from the display as they were touched. Thus, participants could only make errors by pressing the items in the incorrect order, and the task can be considered as a relatively pure test of order memory. Like the serial order recognition task employed by Jarrold et al. (2002), performance on this task should therefore be relatively unaffected by individuals’ familiarity with numbers. A further important advantage of this task over more conventional measures of verbal short-term memory is that it avoids the requirement to produce a spoken response and therefore removes the potential confounding influence of speech difficulties on performance. Performance on the digit reconstruction task was contrasted with that on a closely matched spatial reconstruction task, in which a number of locations on the touch-screen were highlighted sequentially and the participants were required to touch those locations in the correct order. Apart from superficial differences, this task differed from the digit reconstruction task only in terms of the nature of the tobe-remembered material. As such, performance on the spatial reconstruction task effectively provided a control for ‘non-central’ factors such as attention, motivation, task understanding, and motor speed and coordination that could affect performance on the digit reconstruction task (cf. Jarrold & Brock, 2004). Although the two reconstruction tasks were highly similar, there was still a potential confound in that only performance on the digit task could be affected by hearing problems or by difficulties in translating auditory tokens of digits into their visual representations. Moreover, the digit reconstruction task involves actively searching for the digits on the screen, whereas the spatial task does not have this requirement. Consequently, a third digit search task was employed, in which participants heard a single digit and were required to select the corresponding digit from among distracters on the touch-screen. By measuring accuracy of responses, it was possible to screen out participants with poor hearing or limited number knowledge. Search speed was investigated by comparing reaction times (RTs) when different number of distracters were presented. The time to locate the target digit is likely to increase with the number of distracters (cf. Treisman & Gelade, 1980),

so the search slope (i.e., the gradient of the slope obtained by regressing RT against display size) potentially provides an index of an individual’s visual search speed. The performance of individuals with Down syndrome on these tasks was contrasted with that of a group of younger typically developing children. As in previous studies comparing verbal and non-verbal short-term memory in Down syndrome, it was possible to test for a selective deficit in verbal shortterm memory by looking for an interaction between group and task. However, a potential difficulty with this approach is that, if individuals with Down syndrome perform worse than the comparison group on both tasks, any group by task interaction could be explained by differences in the sensitivity of the two tasks (cf. Strauss, 2001). Moreover, differences in effect size (i.e., differences between verbal and nonverbal short-term memory performance) may increase as a function of overall short-term memory performance (cf. Logie, Della Salla, Laiacona, Chalmers, & Wynn, 1996). The study was therefore designed to allow direct comparisons between performance on the two tasks. By regressing performance on one task against the other, it was possible to determine whether individuals with Down syndrome performed more poorly on the digit reconstruction task than predicted given their spatial reconstruction abilities. Indeed, it was also possible to control for RT on the digit search task and measures of verbal and non-verbal mental age in a similar manner. The study therefore provided a strict test of the claim that individuals with Down syndrome have a selective deficit in verbal short-term memory.

Method Participants The Down syndrome group were 32 individuals (12 males) with confirmed trisomy 21, ranging in age from 8;3 to 25;10 years. These participants were recruited through local Down syndrome support groups or schools for children with learning disabilities, or had taken part in previous studies with the Bristol research group. The typically developing group were 36 children (18 males), ranging in age from 4;11 to 9;6 years, who were recruited from a local primary school, and who had no documented history of speech and language difficulties or hearing problems. The aim of the experiment was to control for multiple potential predictors of task performance, so groups were not explicitly matched at the outset of the study on any single measure (cf. Jarrold & Brock, 2004).

Stimuli and apparatus Stimuli for the digit reconstruction task, the digit search task, and the digit recall task were the digits 1 to 9 recorded in a sound-proofed booth by an adult

Short-term memory in Down syndrome

native-English-speaking male. Stimuli were initially recorded as 16-bit, 32 kHz WAV files and were subsequently transformed using SoundEdit software so that the duration of each digit was exactly 500 msec. Computerised tasks were programmed using HyperCard, and were presented on a 15-inch ELO USB touchscreen controlled by a Macintosh iBook laptop computer. Auditory stimuli were presented via Labtec C110 R headphones.

Procedure Participants in the control group were all tested in a quiet room at school, as were approximately half of the individuals with Down syndrome. The remaining participants in the Down syndrome group were tested in a quiet room at home. Where possible, the three touchscreen tasks were administered within the same session, with each task taking approximately five minutes to complete (although individuals with longer spans obviously had more trials to complete). Half of the children in each group were administered the spatial reconstruction task first, followed by the digit search and digit reconstruction tasks. The remaining participants performed the digit search task, then the digit reconstruction and spatial reconstruction tasks. Thus, order of testing for the digit and spatial reconstruction tasks was counterbalanced, but all participants were familiarised with the auditory stimuli by performing the digit search task immediately before the digit reconstruction task. The digit recall task and background measures of verbal and non-verbal ability were also administered. However, practical constraints and the importance of maintaining participants’ interest levels meant that it was not always possible to conduct all testing within a single session.

Digit reconstruction task The task was introduced as the ‘rabbit game’. Participants were told that they would hear some numbers and would have to press those numbers on the screen in the correct order to show a rabbit where to jump. Each trial began with a number of green oval ‘pads’, 42 mm wide and 18 mm in height, appearing on the screen – the number of pads corresponding to the span length being tested. There were nine possible locations for the pads that were spread across the screen, and these were sampled approximately equally. On each trial, the participants heard a sequence of digits presented at a rate of one digit every 600 msec (500 msec for the digit followed by a silent pause of 100 msec). All digit lists were pre-specified, with the constraints that the nine digits were sampled approximately equally, and repetitions of digits within a trial and familiar sequences such as consecutive ascending or descending sequences were avoided. Immediately after the final digit had been presented, the corresponding digits appeared on the screen, one above each of the pads. The digits were 19 mm high and approximately 12 mm wide, and invisible circular touch-screen ‘buttons’, 51 mm in diameter, were defined such that each pad and the corresponding digit were centrally placed within the button. Participants were required to touch the pads in the correct order. An animated rabbit

307

jumped onto the first pad that was touched. Subsequent responses resulted in the rabbit jumping onto the corresponding pad as the old pad and digit disappeared. This prevented repetition errors and ensured that the correct number of responses was given for each list. The task began with two demonstration trials and two practice trials at span-length two. Participants then completed five trials at span-length two. Testing continued with lists of increasing length until all five trials at a particular span-length were incorrectly completed. All participants completed the trials in the same order, and performance was measured in terms of the total number of trials correct.

Spatial reconstruction task This task followed a similar format to the digit reconstruction task. The task was introduced as the ‘frog game’ and participants were told that their task was to help a green frog follow an orange frog. On each trial, an animated orange frog appeared and jumped from one pad to another. The frog landed on each pad for 500 msec, during which time the pad changed colour to a bright yellow, and it took 100 msec for each jump. The presentation rate therefore corresponded exactly to that in the digit reconstruction task. After the orange frog had landed on all the pads, a croak sound was played to indicate that the participant should begin his or her response. The participant was then required to touch the pads in the same order. The green frog followed their response in the same way that the rabbit followed the responses in the digit reconstruction task, with each pad disappearing as the frog jumped onto the next pad. The same pads were used as in the digit reconstruction task, and the two tasks were designed so that the correct response on a particular trial of the spatial order reconstruction task (in terms of the pads that were touched) was exactly the same as the correct response on the corresponding trial in the digit order reconstruction task.

Digit search task This task was introduced as the ‘bird game’. Participants were told that they would hear a number and would then have to press that number on the screen to show the bird where to land. On each trial, there were 3, 6, or 9 pads on the screen in the same set of locations as those employed in the two reconstruction tasks. A single digit was presented through the headphones and, immediately following the offset of the digit, a different digit appeared above each of the pads (one of the digits on the screen was the digit that had just been presented auditorily). Once the response had been made, an animated bird flew onto the corresponding pad, made a ‘chirrup’ sound and flew off again. Following a single demonstration trial and a single practice trial, all participants performed a total of 27 experimental trials with nine trials each with either 3, 6, or 9 numbers displayed on the screen. Trials were presented in a fixed random order (i.e., they were not blocked by display size). The nine digits and the nine possible locations were sampled equally for both targets and distracters for each of the three display sizes. Performance was measured in terms of the total number of correct responses, and reaction times were recorded from the

Jon Brock and Christopher Jarrold

offset of the auditory digit presentation to the onset of the response.

Digit recall task Participants were presented with a sequence of digits via headphones and were required to repeat the digits in the same order. As in the reconstruction tasks, a span procedure was used with five trials at each length, and testing stopped only when participants produced an incorrect response on all five trials at a particular length. However, unlike the reconstruction tasks, there were no practice trials, and the task began at spanlength one. Performance was again measured in terms of the total number of correct responses but, to enable direct comparison with the digit reconstruction task, performance on trials at span-length one were not included in the total score.

Intelligence measures The British Picture Vocabulary Scale II (BPVS; Dunn, Dunn, Whetton, & Burley, 1997), a measure of receptive vocabulary, and the Raven’s Coloured Progressive Matrices (RCPM; Raven, Raven, & Court, 1998), a measure of non-verbal or fluid intelligence, were both administered following standard procedures. Performance on the BPVS was measured in terms of ageequivalent score. However, the RCPM has not been standardised in such a way, and consequently raw scores were instead taken as the dependent measure.

Results Accuracy on the digit search task was used to screen out participants with hearing difficulties or poor number knowledge that might contribute towards poor digit reconstruction performance. Most participants scored either 26 or 27 out of a maximum possible score of 27. However, two participants with Down syndrome and three members of the comparison group scored 23 or less and were eliminated from subsequent analyses. All of the remaining participants scored either 26 or 27, with the exception of two typically developing children who scored 25. In both reconstruction tasks, even if participants were responding randomly, there was a 1 in 2 chance of a correct response on each trial at span-length two, and a 1 in 6 chance of a correct response on each trial at length three. It was important to ensure that, as far as possible, participants were not included who performed poorly on one or both of the tasks because they did not understand the tasks. Consequently, a further criterion for inclusion in the analyses below was that participants should score at least 4 in both the digit and spatial reconstruction tasks. On this basis, a further four participants with Down syndrome were excluded, all of whom had digit reconstruction scores of 3 or less, despite spatial reconstruction scores of at least 6. Reaction times for correct responses in the digit search task were screened by converting them to

z-scores on the basis of the total sample of correct responses across all remaining participants for each of the three display sizes. Responses for which the corresponding z-score was greater than 2 were treated as outliers and were eliminated. The proportion of outliers were similar across groups and display sizes and, after the elimination of outliers, each participant had at least 20 correct trials in total and at least 5 correct responses for each display size. For each participant, the average RT was calculated as an equally-weighted mean of RTs across the three different display sizes (i.e., the mean of the means for each display size). Search slopes (i.e., the gradient of the linear function relating reaction time to display size) were also calculated on an individual basis. Split half reliability measures (Spearman-Brown correction applied) were calculated on the basis of scores for the typically developing group because group differences could artificially inflate correlations when the two groups are combined. The digit recall (r ¼ .886), digit reconstruction (r ¼ .875), and spatial reconstruction (r ¼ .836) tasks all showed high levels of reliability, as did the average RT for the digit search task (r ¼ .901). Unfortunately, the reliability of the search slope measure was extremely low (r ¼ .285), preventing its use as a factor in the regression analyses reported below. However, it was possible to compare search slopes on a group-wise basis by way of an ANOVA using linear contrasts to examine the repeated measure of display size (as recommended by Carter, Krause, & Harbeson, 1986) (see Figure 1). There was a significant linear trend for display size [F (1,57) ¼ 184.45, p < .001], but no

2500

Digit search mean RT / msec

308

2000

DS TD

1500

1000 3

6

9

Display size Figure 1 The effect of display size on RT in the digit search task for individuals with Down syndrome (DS) and typically developing (TD) children. Error bars indicate +/) 1 standard error in the mean. Upper and lower lines show the linear trends fitted to the Down syndrome and typically developing group data respectively

Short-term memory in Down syndrome

309

Table 1 Background measures and overall performance. Significance values are for two-tailed t-tests. Equal variance was assumed for all measures with the exception of chronological age Down syndrome

Chronological age/months BPVS age/months RCPM score Digit recall score Digit reconstruction score Spatial reconstruction score Digit search RT/msec

Typically developing

M

SD

M

SD

t

p

211.1 85.5 16.9 8.81 7.42 12.92 1890

50.1 21.4 4.5 2.81 2.91 3.30 370

73.4 88.3 21.7 15.15 14.09 14.00 1810

16.4 19.3 5.1 3.93 4.25 2.88 380

13.08 ).52 )3.66 )6.94 )6.83 )1.34 .80

<.001 .608 .001 <.001 <.001 .218 .426

significant effect of group [F (1,57) ¼ .66, p ¼ .421], and no significant interaction between group and the linear trend for display size [F (1,57) ¼ .14, p ¼ .707]. As such, there was no evidence for any group differences in search speed. Mean chronological ages, vocabulary mental ages, and RCPM scores1 for the remaining 26 individuals with Down syndrome and 33 typically developing children are shown in Table 1, together with mean scores on the digit and spatial reconstruction tasks and the digit recall task, and average RT for the digit search task. The table also shows the results of a series of independent t-tests comparing the two groups on each dependent measure. There were no significant group differences in vocabulary age, performance on the spatial reconstruction task, or mean RT on the digit search task. However, the Down syndrome group were significantly older than controls, had significantly lower RCPM scores, and performed significantly worse than controls on the digit reconstruction and recall tasks. These data were analysed in three further ways. First, to enable direct comparisons with previous studies, performance on the digit and spatial reconstruction tasks was analysed by way of a mixed design ANOVA with group and task as between- and within-participants factors respectively. There were significant effects of task [F (1,57) ¼ 28.70, p < .001] and group [F (1,57) ¼ 27.49, p < .001], but these were qualified by a significant group by task interaction [F (1,57) ¼ 30.66, p < .001]. Two-tailed, paired samples t-tests revealed that the Down syndrome group performed significantly worse on the 1 Raven’s matrices scores were not available for two individuals with Down syndrome and one typically developing child. 2 Chronological age was not included in these analyses because there was very little overlap in age ranges for the two groups, so the use of age as a covariate is inappropriate (cf. Wright, 1997). Moreover, correlational analyses performed separately on each group showed that chronological age was not significantly correlated with performance on either the digit reconstruction task or the digit recall task when mental age (either BPVS or RCPM) was partialled out (one-tailed p’s > .05). Thus, mental age rather than chronological age was the relevant measure to include as an index of the extent to which general developmental differences might affect individuals’ performance.

digit reconstruction task than on the spatial reconstruction task [t (25) ¼ 7.19, p < .001], whereas the typically developing group showed no significant effect of task [t (32) ¼ ).14, p ¼ .892]. Second, in order to control for factors that may contribute to verbal short-term memory performance, a series of covariance analyses were performed.2 Figure 2 shows four graphs, each plotting the performance of participants in the two groups on the digit reconstruction task as a function of performance on either the spatial reconstruction task, mean RT on the digit search task, BPVS age, or RCPM score. In each case, the overwhelming majority of individuals with Down syndrome appear below the regression line fitted to the typically developing group, indicating that their digit reconstruction performance was worse than expected given their performance on that particular background measure. A corresponding series of ANCOVAs were performed with digit reconstruction score as dependent variable, group as a betweensubjects factor, and a single predictor variable as a covariate. These showed that, although the four predictor variables all accounted for significant variance in digit reconstruction (p’s < .01), the significant effect of group remained (p’s < .001) in each case. A similar pattern of results was also found when digit recall score was the dependent variable. Third, the data were subjected to principled hierarchical regression in order to determine the factors that were most important in predicting performance on the digit reconstruction and recall tasks.3 The top half of Table 2 shows the results of the first set of regression analyses in which either digit reconstruction or digit recall scores were the dependent variable. Performance on the spatial reconstruction task and RT on the digit search task were entered as independent variables in the first two steps because these were specifically designed as direct control measures for the digit reconstruction task. Both digit search RT and spatial reconstruction performance 3

The reported hierarchical regression analyses are based on participants for whom RCPM scores were available. These analyses were repeated with all participants included up to the point at which RCPM and BPVS were entered but there were no qualitative differences in the results.

310

Jon Brock and Christopher Jarrold

Digit reconstruction score

25

20

15

10

5

0 0

5

10

15

20

25

1000

Spatial reconstruction score

1500

2000

2500

3000

Digit search mean RT / msec

Digit reconstruction score

25

20

15

10

5

0 25

50

75

100

125

150

5

10

BPVS age / months

15

20

25

30

35

Ravens score

DS

TD

Figure 2 The relationship between performance on the digit reconstruction task and performance on the spatial reconstruction task (top left panel), RT on the digit search task (top right), BPVS mental age (bottom left), and RCPM score (bottom right). Regression lines show fits to data for typically developing children

Table 2 Results of hierarchical regression analysing the determinants of digit reconstruction and digit recall scores Digit reconstruction DR2 1. 2. 1. 2.

Spatial reconstruction Digit search Digit search Spatial reconstruction 3. BPVS & RCPM 4. Group 1. Group 2. Spatial reconstruction 3. Digit search 2. Digit search 3. Spatial reconstruction 4. BPVS & RCPM

.146 .099 .182 .063 .137 .274 .481 .064 .074 .115 .023 .037

Digit recall

p

DR2

p

.004 .011 .001 .040 .006 <.001 <.001 .008 .003 <.001 .084 .078

.110 .081 .146 .045 .125 .296 .494 .040 .058 .086 .012 .020

.013 .025 .004 .090 .014 <.001 <.001 .038 .009 .002 .221 .277

independently accounted for significant variance in digit reconstruction performance but, for digit recall, only digit search RT made a significant independent contribution. The proportion of variance explained by the shared contribution of digit search RT and spatial reconstruction score was .083 for digit reconstruction score and .065 for digit recall. Based on the benchmark values for the independent contributions, we can assume that both of these values are significant (cf. Chuah & Mayberry, 1999). BPVS age and RCPM score were entered simultaneously in the third step, and it was found that these more general measures of intelligence contributed significant extra variance in both digit reconstruction and recall, beyond that accounted for by the two control tasks. Group was entered as the final step and, crucially, it can be seen that the two groups differed in verbal short-term memory performance, even

Short-term memory in Down syndrome

when individual differences in the other measures had been accounted for. However, as group was not entered until the final stage of this initial set of regressions, it is possible that the predictive power of some of these background measures was driven by the extent to which they themselves shared variance with group membership. In particular, as the two groups differed significantly in RCPM score, adding this measure to the model may effectively add additional grouprelated variance to the prediction of the dependent measures. Consequently, in a second set of regression analyses (shown in the bottom half of Table 2), group was entered as a first step, with the other independent measures then entered as before. Again, RT on the digit search task accounted for significant variance in both digit reconstruction and recall, even after spatial reconstruction had been entered. In contrast, spatial reconstruction did not account for significant extra variance in either digit reconstruction or recall beyond that attributable to digit search RT. Similarly, measures of general intelligence, entered as the final step, did not account for unique variance in either digit reconstruction or recall.

Discussion Evidence for a selective deficit in verbal short-term memory in Down syndrome This study investigated the verbal short-term memory abilities of individuals with Down syndrome and typically developing children by comparing their performance on a digit reconstruction task with that observed on a closely matched spatial reconstruction task. Conventional ANOVA revealed a significant interaction between group and task, with the Down syndrome group performing poorly on the digit reconstruction task relative to their own performance on the spatial reconstruction task, and relative to the typically developing group’s performance on the digit reconstruction task. The existence of a selective deficit in verbal short-term memory performance in Down syndrome was confirmed by ANCOVA showing that group differences in digit reconstruction remained when performance on the spatial reconstruction task was covaried out. Furthermore, hierarchical regression showed that the Down syndrome group performed significantly worse than controls on the digit reconstruction task (and the digit recall task) even after variation associated with spatial reconstruction performance, as well as RT on the digit search task, and verbal and nonverbal mental age measures had been accounted for. These findings are consistent with the view that Down syndrome is associated with a selective deficit in verbal short-term memory. However, the advantage of the current study over previous studies is that it was designed in such a way as to eliminate a

311

number of potentially confounding factors that might lead to poor verbal short-term memory task performance in Down syndrome. The use of the touch-screen-based digit reconstruction task avoided the need for spoken responses and therefore eliminated the impact that speech difficulties in Down syndrome may have on performance. Moreover, it allowed the verbal and non-verbal tasks to be closely matched, so it is difficult to explain away group differences in terms of motivation, attention, or motor coordination. The ease with which participants could understand the task requirements should also have been comparable. However, to ensure that poor performance did not reflect a lack of task understanding, participants were only included if their scores on both tasks were well above chance levels. In the event, these inclusion criteria provided a conservative test of group differences as they resulted in the elimination of a small number of individuals with Down syndrome, each of whom performed much better on the spatial task than the digit task. A final advantage of the digit reconstruction task is that it was primarily a test of order memory and should therefore have been relatively immune to group differences in familiarity with numbers. Clearly, there remain a number of differences between the two reconstruction tasks, but these were addressed by the inclusion of the digit search task. This task allowed us to screen out participants who, as a result of either hearing difficulties or poor number knowledge, were not consistently able to match the auditory token of a digit with the corresponding visual representation. Thus, it appears unlikely that group differences in digit reconstruction performance can be explained in terms of such difficulties. The task also enabled investigation of visual search speed – another factor that could influence performance on the digit reconstruction task. It had been intended to use the individuals’ search slope as an index of search speed but, unfortunately, this measure proved unreliable. Nevertheless, when RTs were analysed on a group-wise basis, the two groups demonstrated extremely similar search slopes. It is therefore difficult to explain group differences in digit reconstruction performance in terms of slower visual search in the Down syndrome group.

Predictors of verbal short-term memory performance The regression approach adopted in the current study also enabled investigation of the associations between performance on the verbal short-term memory tasks and a number of background measures. Four measures – vocabulary mental age (BPVS), non-verbal ability (RCPM), performance on the spatial reconstruction task, and mean RT on the digit search task – were all found to predict

312

Jon Brock and Christopher Jarrold

performance on the digit reconstruction and recall tasks. Hierarchical regression analyses showed that the two intelligence measures predicted performance on the two verbal short-term memory tasks, even after performance on the spatial reconstruction task and digit search RT had been controlled for. However, this appears to reflect the fact that the individuals with Down syndrome had lower RCPM scores than the typically developing children, so the intelligence measures were effectively acting as a surrogate for group membership. When group membership was controlled for, the addition of the intelligence measures to the regression model did not account for significant extra variance in verbal short-term memory performance. This implies that the spatial reconstruction task and the digit search task provided adequate control of non-central determiners of task performance (such as attention and task understanding) (cf. Jarrold & Brock, 2004). In fact, it appears that the use of the digit search and spatial reconstruction tasks allowed for the control of at least three potentially independent sources of variance in digit reconstruction. First, the independent association between spatial reconstruction and digit reconstruction must reflect factors that are common to these two tasks but are not involved in the digit search task, such as the ability to execute a series of touch responses to the spatial locations on the screen. Similarly, the variance in digit reconstruction shared specifically with digit search RT must reflect processes common to these two tasks, but not required in spatial reconstruction. The digit search task was included primarily as a means of controlling for the effect of search speed on digit reconstruction. As noted above, search slopes were unreliable measures of search speed, but overall RT on the search task will also be influenced by search speed, and this could explain some of the shared variance. However, digit search RT also accounted for unique variance in digit recall (in which there was obviously no search component), so there must be a further source of shared variation. One possibility is that digit search RT is influenced by the ease with which long-term representations of digits can be accessed when the auditory tokens are presented, and this may also contribute to performance on digit reconstruction and recall tasks (cf. Dempster, 1981; Hitch & McAuley, 1991). Finally, the variance shared by the two control measures also made a reliable contribution to digit reconstruction score. This finding presumably reflects the influence of basic factors such as motivation or general intelligence that will constrain individual’s performance on all of the tasks. Although these results clearly indicate that digit reconstruction performance is multiply determined, and depends on a number of factors beyond those specifically associated with verbal short-term memory performance, it is important to remember that these potentially confounding factors were controlled

for in this study. There were no significant group differences in either digit search RT or spatial reconstruction score, and significant group differences in verbal short-term memory performance remained even when variance associated with these background measures had been accounted for. Thus, it seems highly unlikely that the deficits seen among individuals with Down syndrome on the two verbal short-term memory measures can be explained in terms of impaired sequencing and execution of motor responses, slower access to longterm representations of the digits, visual search difficulties, poor motivation, or generally low intelligence.

Evidence for a selective deficit in order memory? As noted earlier, the digit reconstruction task is primarily a test of order memory as opposed to item memory, so the poor performance of individuals with Down syndrome on this task indicates that they have impaired serial order memory. These findings are consistent with those of Jarrold et al. (2002), who reported that, compared with controls matched on receptive vocabulary, individuals with Down syndrome performed poorly on a serial order recognition task. Together, therefore, these two studies provide support for the claim that Down syndrome is associated with a selective deficit in sequencing or successive processing (cf. Das, 2002; Rosin, Swift, Bless, & Vetter, 1988; Snart, O’Grady, & Das, 1982). The main difference between the current findings and those of Jarrold et al. (2002) is that the current data clearly support a dissociation between verbal and visuo-spatial short-term memory performance in Down syndrome, although this may well reflect the greater sensitivity of the reconstruction tasks used here. Consequently, an important implication of the current data is that the impairment in representing order in short-term memory is particularly severe in the verbal domain, and does not simply reflect a general deficit that also leads to deficits in the reconstruction of the temporal order of spatial information (cf. Chuah & Maybery, 1999; Jones, Farrand, Stuart, & Morris, 1995). This is not to say, however, that verbal short-term memory difficulties in Down syndrome are necessarily limited to the representation of order information. A number of studies have shown that individuals with Down syndrome perform poorly on digit recall even when responses are scored as being correct irrespective of whether or not the items are in the correct order (Jarrold & Baddeley, 1997; Marcell et al., 1988). Moreover, in a recent study (Brock & Jarrold, in press), we compared performance on two recognition tasks – an order memory task similar to that used by Jarrold et al. (2002), and an item memory task in which participants were required to detect changes in one of the items in a list. Individuals with Down syndrome were impaired on both

Short-term memory in Down syndrome

order and item recognition tasks, but the extent of their deficit was greater for item than for order information. Although this may, at least in part, have been a consequence of difficulties in identifying the items, these findings clearly would not support a claim that there is a selective deficit in order as opposed to item memory. Any mechanistic account of verbal short-term memory deficits in Down syndrome therefore has to explain both order and item memory difficulties.

Mechanisms underlying the short-term memory deficit in Down syndrome Verbal short-term memory deficits in Down syndrome have previously been conceptualised in terms of an impairment of the phonological loop component of working memory (see Jarrold et al., 1999b). According to the phonological loop model, verbal material is held in a limited-capacity phonological store, but phonological representations decay rapidly unless they are refreshed by rehearsal (Baddeley, 1986). Evidence suggests that neither individuals with Down syndrome nor the individuals with whom they are usually compared engage in rehearsal (Jarrold, Baddeley, & Hewes, 2000). Consequently, the poor short-term memory performance of individuals with Down syndrome relative to these controls would appear to indicate some form of deficit in the phonological store – either a reduction in store capacity or an increase in decay rate (Jarrold et al., 1999b; Purser & Jarrold, 2004; cf. Gathercole & Baddeley, 1990). Both of these impairments would clearly lead to poor memory for the items in a list, and one would also expect order memory difficulties if the loss of item information made the list members indistinguishable. An alternative perspective on short-term memory difficulties is provided by recent computational models of serial order memory (e.g., Brown, Preece, & Hulme, 2000; Burgess & Hitch, 1999). According to such models, sequences are encoded by associating item nodes with different states of a timevarying context signal. The sequence can then be retrieved by resetting the temporal context signal to the beginning of the list and outputting items as they are cued by the signal. Poor order and item memory can be explained in terms of increased noise in the association processes, either at the encoding or retrieval stage. Order errors arise if neighbouring list items are cued more strongly than the correct item. However, the context signal is used as the cue for retrieving items, even when the task does not explicitly require order retention (Brown et al., 2000), so noise in the system can also lead to item errors. Intriguingly, similar models have been successfully applied to errors in speech production (Vousden, Brown, & Harley, 2000), and neuro-imaging studies indicate that the same cortical areas are involved in both serial recall and speech production (see Gupta

313

& MacWhinney, 1997). It is therefore possible that a common mechanism could account for poor verbal short-term memory and speech production difficulties in Down syndrome. Indeed, tentative support for this hypothesis comes from a study by Rosin et al. (1988), who noted that individuals with Down syndrome had greatest difficulty with speech production measures when the sequencing demands were high.

Future research Together with other recent studies (e.g., Brock & Jarrold, in press; Jarrold et al., 2000, 2002), the current study provides strong evidence that Down syndrome is associated with a fundamental deficit in verbal short-term memory that cannot be explained in terms of a general memory deficit, or extraneous factors such as hearing impairment, speech difficulties, or poor item knowledge. However, the precise cause of this deficit is not clear at present, and further research is required to distinguish between a number of candidate mechanisms. The current results also do not address the specificity of the verbal short-term memory deficit to Down syndrome. Previous studies have reported that individuals with Down syndrome have greater difficulty on verbal short-term memory tasks than do younger learning-disabled controls matched on vocabulary knowledge (Jarrold & Baddeley, 1997; Jarrold et al., 2000, 2002). However, a strict test of specificity would require that degree of learning disability (i.e., IQ) is controlled for as well as ability level. Marcell and colleagues (Marcell & Weeks, 1988; Marcell et al., 1988; Marcell & Cohen, 1992) have reported that individuals with Down syndrome do have poor verbal short-term memory relative to controls with comparable IQs but other researchers have failed to find such differences (Mackenzie & Hulme, 1987; Varnhagen et al., 1987), and none of these studies have provided the rigorous control of confounding factors exerted in the current study. We are therefore unable to make any strong claims regarding the specificity of verbal short-term memory deficits to Down syndrome. Indeed, if verbal short-term memory plays a causal role in language acquisition (cf. Baddeley et al., 1998), then one would predict that individuals with similar rates of language acquisition should also have comparable verbal short-term memory abilities (cf. Jarrold, Baddeley, Hewes, Leeke, & Phillips, 2004). Thus, we might expect to find very similar patterns of short-term memory performance in individuals with Down syndrome and those with similar delays in language abilities. In a sense, therefore, the issue of specificity is one with theoretical rather than practical implications. Having, we believe, determined that there genuinely is a fundamental verbal short-term memory deficit in Down syndrome, what we are really interested in is the consequences of such an impairment for language development. Earlier, we highlighted the

314

Jon Brock and Christopher Jarrold

potential link between verbal short-term memory and the mechanisms of speech production, but verbal short-term memory is also thought to play a crucial role in the learning of the phonological forms of new words (e.g., Baddeley, Papagno, & Vallar, 1988), and in sentence comprehension when word order is important (e.g., Vallar & Baddeley, 1984). However, at present, the causal links between verbal short-term memory and language deficits in Down syndrome are only vaguely specified, and it is unclear whether the profile of language difficulties actually corresponds to that predicted by these theoretical accounts. Consequently, an important challenge for future research is to make explicit the causal links between verbal short-term impairment and different areas of language difficulty in Down syndrome.

Acknowledgements This research was supported by a Charles J. Epstein Research Award from the National Down Syndrome Society of the United States. We are grateful to the pupils and staff of Ashton Gate, Fosseway, Kingsweston, and Ravenswood schools for their cooperation in this work.

Correspondence to Jon Brock, Department of Experimental Psychology, University of Bristol, 8 Woodland Road, Bristol BS8 1TN, UK; Email: [email protected]

References Baddeley, A.D. (1986). Working memory. Oxford, England: Oxford University Press. Baddeley, A.D., Gathercole, S., & Papagno, C. (1998). The phonological loop as a language learning device. Psychological Review, 105, 158–173. Baddeley, A., Papagno, C., & Vallar, G. (1988). When long-term learning depends on short-term storage. Journal of Memory and Language, 27, 586–596. Brener, R. (1940). An experimental investigation of memory span. Journal of Experimental Psychology, 26, 467–482. Broadley, I., MacDonald, J., & Buckley, S. (1995). Working memory in children with Down’s syndrome. Down’s Syndrome: Research and Practice, 3, 3–8. Brock, J., & Jarrold, C. (in press). Language influences on verbal short-term memory performance in Down syndrome: Item and order recognition. Journal of Speech, Language, and Hearing Research. Brock, J., McCormack, T., & Boucher, J. (2004). Lexical influences on phonological short-term memory in Williams syndrome. Manuscript submitted for publication. Brown, G.D.A., Preece, T., & Hulme, C. (2000). Oscillator-based memory for serial order. Psychological Review, 107, 127–181.

Burgess, N., & Hitch, G.J. (1999). Memory for serial order: A network model of the phonological loop and its timing. Psychological Review, 106, 551–581. Carter, R.C., Krause, M., & Harbeson, M.M. (1986). Beware the reliability of slope scores for individuals. Human Factors, 28, 673–683. Chapman, R.S. (1995). Language development in children and adolescents with Down syndrome. In P. Fletcher & B. MacWhinney (Eds.), Handbook of child language (pp. 641–663). Oxford: Blackwell. Chapman, R.S. (1997). Language development in children and adolescents with Down syndrome. Mental Retardation and Developmental Disabilities Research Reviews, 3, 307–312. Chuah, Y.M.L., & Maybery, M.T. (1999). Verbal and spatial short-term memory: Common sources of developmental change? Journal of Experimental Child Psychology, 73, 7–44. Cowan, N., Day, L., Saults, J.S., Keller, T.A., Johnson, T., & Flores, L. (1992). The role of verbal output time in the effects of word-length on immediate memory. Journal of Memory and Language, 31, 1–17. Dahle, A.J., & McCollister, F.P. (1986). Hearing and otologic disorders in children with Down syndrome. American Journal of Mental Deficiency, 90, 636–642. Das, J.P. (2002). A better look at intelligence. Current Directions in Psychological Science, 11, 28–33. Dempster, F.N. (1981). Memory span: Sources of individual and developmental differences. Psychological Bulletin, 89, 63–100. Dodd, B. (1975). Recognition and reproduction of words by Down’s syndrome and non-Down’s syndrome retarded children. American Journal of Mental Deficiency, 80, 306–311. Dodd, B., & Thompson, L. (2001). Speech disorder in children with Down’s syndrome. Journal of Intellectual Disability Research, 45, 308–316. Dosher, B.A., & Ma, J.J. (1998). Output loss or rehearsal loop? Output-time versus pronunciationtime limits in immediate recall for forgetting-matched materials. Journal of Experimental Psychology: Learning, Memory and Cognition, 24, 316–335. Dunn, L.M., Dunn, L.M., Whetton, C., & Burley, J. (1997). British Picture Vocabulary Scale (2nd edn). Windsor, UK: NFER-Nelson. Fowler, A.E. (1990). Language abilities in children with Down syndrome. In D. Cicchetti & M. Beeghly (Eds.), Children with Down syndrome: The developmental perspective (pp. 302–328). New York: Cambridge University Press. Gathercole, S.E., & Baddeley, A.D. (1990). Phonological memory deficits in language disordered children: Is there a causal connection? Journal of Memory and Language, 29, 336–360. Gathercole, S.E., & Martin, A.J. (1996). Interactive processes in phonological memory. In S.E. Gathercole (Ed.), Models of short-term memory (pp. 73–100). Hove, England: Psychology Press. Gathercole, S.E., Pickering, S.J., Hall, M., & Peaker, S.M. (2001). Dissociable lexical and phonological influences on serial recognition and serial recall. Quarterly Journal of Experimental Psychology, 54A, 1–30. Gelman, R., & Cohen, M. (1988). Qualitative differences in the way Down syndrome and normal children solve

Short-term memory in Down syndrome

a novel counting problem. In L. Nadel (Ed.), The psychobiology of Down syndrome. Cambridge, MA: MIT Press. Gibson, D. (1978). Down’s syndrome: The psychology of mongolism. Cambridge: Cambridge University Press. Gupta, P., & MacWhinney, B. (1997). Vocabulary acquisition and verbal short-term memory: Computational and neural bases. Brain and Language, 59, 267–333. Healy, A.F. (1974). Separating item from order information in short-term memory. Journal of Verbal Learning and Verbal Behavior, 13, 644–655. Hitch, G.J., & McAuley, E. (1991). Working memory in children with specific arithmetical learning difficulties. British Journal of Psychology, 82, 375–386. Hulme, C., & Mackenzie, S. (1992). Working memory and severe learning difficulties. Hove, UK: Lawrence Erlbaum Associates. Hulme, C., Maughan, S., & Brown, G.D.A. (1991). Memory for familiar and unfamiliar words – evidence for a long-term-memory contribution to short-termmemory span. Journal of Memory and Language, 30, 685–701. Hulme, C., & Roodenrys, S. (1995). Practitioner review – verbal working-memory development and its disorders. Journal of Child Psychology and Psychiatry, 36, 373–398. Jarrold, C. (2001). Applying the working memory model to the study of atypical development. In J. Andrade (Ed.), Working memory in perspective (pp. 126–150). Hove, UK: Psychology Press. Jarrold, C., & Baddeley, A.D. (1997). Short-term memory for verbal and visuo-spatial information in Down’s syndrome. Cognitive Neuropsychiatry, 2, 101– 122. Jarrold, C., Baddeley, A.D., & Hewes, A.K. (1999a). Genetically dissociated components of working memory: Evidence from Down’s and Williams syndrome. Neuropsychologia, 37, 637–651. Jarrold, C., Baddeley, A.D., & Hewes, A.K. (2000). Verbal short-term memory deficits in Down syndrome: A consequence of problems in rehearsal? Journal of Child Psychology and Psychiatry, 41, 233–244. Jarrold, C., Baddeley, A.D., Hewes, A.K., Leeke, T., & Phillips, C. (2004). What links verbal short-term memory performance and vocabulary level? Evidence of changing relationships among individuals with learning disability. Journal of Memory and Language, 50, 134–148. Jarrold, C., Baddeley, A.D., & Phillips, C. (1999b). Down syndrome and the phonological loop: The evidence for, and importance of, a specific verbal short-term memory deficit. Down Syndrome Research and Practice, 6, 61–75. Jarrold, C., Baddeley, A.D., & Phillips, C.E. (2002). Verbal short-term memory in Down syndrome: A problem of memory, audition, or speech? Journal of Speech, Language, and Hearing Research, 45, 531– 544. Jarrold, C., & Brock, J. (2004). To match or not to match? Methodological issues in autism-related research. Journal of Autism and Developmental Disorders, 34, 81–86.

315

Jones, D., Farrand, P., Stuart, G., & Morris, N. (1995). Functional equivalence of verbal and spatial information in serial short-term memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 1008–1018. Kay-Raining Bird, E., & Chapman, R.S. (1994). Sequential recall in individuals with Down syndrome. Journal of Speech and Hearing Research, 37, 1369– 1380. Laws, G. (1998). The use of nonword repetition as a test of phonological memory in children with Down syndrome. Journal of Child Psychology and Psychiatry, 39, 1119–1130. Laws, G. (2002). Working memory in children and adolescents with Down syndrome: Evidence from a colour memory experiment. Journal of Child Psychology and Psychiatry, 43, 353–364. Laws, G., MacDonald, J., & Buckley, S. (1996). The effects of a short training in the use of a rehearsal strategy on memory for words and pictures in children with Down syndrome. Down’s Syndrome: Research and Practice, 4, 70–78. Limongi, S.C.O., Carvallo, R.M.M., & Souza, E.R. (2000). Auditory processing and language in Down syndrome. Journal of Medical Speech-Language Pathology, 8, 27–34. Logie, R.H., Della Salla, S., Laiacona, M., Chambers, P., & Wynn, V. (1996). Group aggregates and individual reliability: The case of verbal short-term memory. Memory and Cognition, 24, 305–321. Mackenzie, S., & Hulme, C. (1987). Memory span development in Down’s syndrome, severely subnormal and normal subjects. Cognitive Neuropsychology, 4, 303–319. Marcell, M.M., & Armstrong, V. (1982). Auditory and visual sequential memory of Down syndrome and nonretarded children. American Journal of Mental Deficiency, 87, 86–95. Marcell, M.M., & Cohen, S. (1992). Hearing abilities of Down syndrome and other mentally handicapped adolescents. Research in Developmental Disabilities, 15, 533–551. Marcell, M.M., Harvey, C.F., & Cothran, L.P. (1988). An attempt to improve auditory short-term memory in Down’s syndrome individuals through reducing distractors. Research in Developmental Disabilities, 9, 405–417. Marcell, M.M., & Weeks, S.L. (1988). Short-term memory difficulties and Down’s syndrome. Journal of Mental Deficiency Research, 32, 153–162. Martin, N., & Saffran, E.M. (1997). Language and auditory-verbal short-term memory impairments: Evidence for common underlying processes. Cognitive Neuropsychology, 14, 641–682. McDade, H.L., & Adler, S. (1980). Down syndrome and short-term memory impairment: A storage or retrieval deficit? American Journal of Mental Deficiency, 84, 561–567. Milner, B. (1971). Interhemispheric differences in the localisation of psychological processes in man. British Medical Bulletin, 27, 272–277. Numminen, H., Service, E., Ahonen, T., & Ruoppila, I. (2001). Working memory and everyday cognition in adult persons with Down’s syndrome. Journal of Intellectual Disability Research, 45, 157–168.

316

Jon Brock and Christopher Jarrold

Nye, J., Fluck, M., & Buckley, S. (2001). Counting and cardinal understanding in children with Down syndrome and typically developing children. Down’s Syndrome: Research and Practice, 7, 68–78. Pennington, B.F., Moon, J., Edgin, J., Stedron, J., & Nadel, L. (2003). The neuropsychology of Down syndrome: Evidence for hippocampal dysfunction. Child Development, 74, 75–93. Purser, H.R.M., & Jarrold, C. (2004, January). Modality dissociations of serial position in a probed recall task: Evidence from Down syndrome and typical development. Poster presented at the Experimental Psychology Society London Meeting. Raven, J., Raven, J.C., & Court, J.H. (1998). Coloured progressive matrices. Oxford: Oxford University Press. Rosin, M.M., Swift, E., Bless, D., & Vetter, D.K. (1988). Communication profiles of adolescents with Down syndrome. Journal of Childhood Communication Disorders, 12, 49–63. Saint-Aubin, J., & Poirier, M. (1999). The influence of long-term memory factors on immediate serial recall: An item and order analysis. International Journal of Psychology, 34, 347–352. Seung, H.-K., & Chapman, R. (2000). Digit span in individuals with Down syndrome and in typically developing children: Temporal aspects. Journal of Speech, Language, and Hearing Research, 43, 609– 620. Snart, F., O’Grady, M., & Das, J.P. (1982). Cognitive processing by subgroups of moderately mentally retarded children. American Journal of Mental Deficiency, 86, 465–472. Steele, J., & Stratford, B. (1995). Present and future possibilities for the UK population with Down’s syndrome. American Journal on Mental Retardation, 99, 664–682. Strauss, M.E. (2001). Demonstrating specific cognitive deficits: A psychometric perspective. Journal of Abnormal Psychology, 110, 6–14. Tehan, G., & Lalor, D.M. (2000). Individual differences in memory span: The contribution of rehearsal access to lexical memory, and output speed. Quarterly Journal of Experimental Psychology, 53A, 1012–1038.

Thorn, A.S.C., & Gathercole, S.E. (2001). Language differences in verbal short-term memory do not exclusively originate in the process of subvocal rehearsal. Psychonomic Bulletin and Review, 8, 357– 354. Treisman, A.M., & Gelade, G. (1980). A feature-integration theory of attention. Cognitive Psychology, 12, 97– 136. Turner, J.E., Henry, L.A., & Smith, P.T. (2000). The development of the use of long-term knowledge to assist short-term recall. Quarterly Journal of Experimental Psychology, 53A, 457–478. Vallar, G., & Baddeley, A.D. (1984). Phonological shortterm store, phonological processing and sentence comprehension: A neuropsychological case study. Cognitive Neuropsychology, 1, 121–141. Varnhagen, C.K., Das, J.P., & Varnhagen, S. (1987). Auditory and visual memory span: Cognitive processing by TMR individuals with Down syndrome or other etiologies. American Journal of Mental Deficiency, 91, 398–405. Vicari, S., Marotta, L., & Carlesimo, G.A. (2004). Verbal short-term memory in Down syndrome: An articulatory loop deficit. Journal of Intellectual Disability Research, 48, 80–92. Vousden, J.I., Brown, G.D.A., & Harley, T.A. (2000). Serial control of phonology in speech production: A hierarchical model. Cognitive Psychology, 41, 101– 175. Wang, P.P., & Bellugi, U. (1994). Evidence from 2 genetic syndromes for a dissociation between verbal and visual-spatial short-term-memory. Journal of Clinical and Experimental Neuropsychology, 16, 317–322. Welsh, T.N., & Elliot, D. (2001). The processing speed of visual and verbal movement by adults with and without Down syndrome. Applied Physical Activity Quarterly, 18, 156–167. Wright, D.B. (1997). Understanding statistics: An introduction for the social sciences. London: Sage. Manuscript accepted 11 February 2004