Printer Friendly

The relationship between articulation time and memory performance in verbal and visuospatial tasks.

When people are asked to remember a set of items in order for a very short time there is a limit to the number of items they can recall. For verbal material, the amount which can be recalled has been related to both the length of the words used and the rate at which subjects can speak. If memory sets are composed of words that take short or long times to articulate, the amount which can be remembered depends on how long the items take to say (Baddeley, Thomson & Buchanan, 1975). Subjects using languages in which the items to be recalled are more slowly articulated show reduced spans, both between languages if they are bilingual (Ellis & Hennelly, 1980) and across different groups of subjects using different languages (Naveh-Benjamin & Ayres, 1986). Memory span within one language is affected by the length of the words being used as well as by how quickly subjects can speak (Hulme, Thomson, Muir & Lawrence, 1984). This relationship between how long items take to say and the number that can be recalled applies to both words and non-words (Hulme, Maughan & Brown, 1991). There is therefore considerable support for the view that there is a time base to the number of verbal items which can be recalled from immediate memory, and that both word length and articulation rate predict memory performance.

It has been argued (Baddeley, 1986) that verbal material which is to be recalled is held in a short-term phonological store from which it decays, unless rehearsed. Rehearsal takes time, and memory span for verbal material such as digits is affected by the speed at which rehearsal takes place. Rehearsal is not by actual articulation (Baddeley & Wilson, 1985), but by a covert process which is related to articulation and which shares the same temporal limitations. It is the rate of rehearsal which links articulation rate to the number of items which can be recalled. Baddeley's (1986) working memory model also suggests that non-verbal spatial material can be rehearsed and that rehearsal involves implicit motor processes such as those involved when an eye or a hand is moved to a target. If memory items are spatial and not verbal it seems unlikely that memory performance would be affected by limitations related to the articulation of verbal material.

In a previous paper (Smyth & Scholey, 1992), we investigated the recall of spatial memory items in a span paradigm. The task used was based on the Corsi blocks task (Milner, 1971) in which a sequence of blocks is tapped in order by an investigator and the subject is required to repeat the sequence by tapping the same blocks in the same order. We were interested in whether spatial span could be predicted by movement time of hand or eye, as a covert rehearsal mechanism for spatial information might involve a response based process related to actual movement time. While we did not find any evidence of a link between movement time and spatial span, and this is supported by subsequent work (Smyth & Scholey, 1994), we did find that spatial span was related to articulation rate in the groups of adults tested.

This finding was rather surprising and not easy to interpret. The relationship between articulation rate and memory span in studies of verbal memory span is traditionally explained in terms of the time taken to rehearse verbal items (Baddeley, 1986). This explanation cannot be used for spatial span, which is not maintained in the verbal subsystem, and is not affected by concurrent articulation (Smyth & Scholey, 1992). In spatial span the correct items have to be identified and they have to be recalled in order. In short-term verbal memory, information about serial order is thought to be maintained by a phonological system and an articulatory loop. Burgess & Hitch (1992) have pointed out that one of the major functions of this system is the preservation of order, even if accounts of how serial order is maintained in such a system do not explain order errors in verbal recall. If order in spatial span were maintained by verbal recoding then articulatory suppression should affect recall, just as it does digit span, but this is not the case.

Serial order effects in spatial memory have not been widely investigated, and as far as we know, no one has argued that one of the functions of a visuospatial working memory would be to maintain order. However, Jones Farrand, Stuart & Morris (1995) argue that there are similarities between spatial and verbal tasks when the items are known but order has to be recalled. They consider that serial order phenomena are important in understanding both spatial and verbal short-term memory and argue that dissociations between verbal and spatial memory which do not involve order in recall (e.g. Logie, Zucco & Baddeley, 1990) are not true dissociations between verbal and spatial processes, but are differences between memory tasks involving serial order and those which do not. They suggest that maintenance of serial order may share common features across memory tasks presented in different modalities.

The meaning of the relationship between articulation rate and verbal memory span has been questioned in a series of papers by Cowan and his colleagues. Cowan, Day, Saults, Keller, Johnson & Flores (1992) found that word length effects in verbal serial recall occurred because of the length of the words which were produced first in a list; that is, that response production was affected by word length. Cowan (1992) found that the time taken to produce the items in verbal span did not relate to span performance in the same way that articulation rate in separate repetition tasks does, but that the overall duration of recall was related to memory span. Overall duration is affected by the intervals between the spoken words and Cowan, following Sternberg, Wright, Knoll & Monsell (1980), has argued that phonological memory can be reactivated in the pauses between item production and that this reactivation is similar to the scanning of items in a memory set. In this account serial order is represented with a serial position tag and to produce items in the correct serial order it is necessary to scan them to find the correct tag before each item is produced. The longer the list, the longer the time taken to scan the set and therefore the longer the pauses between items at recall. Adding reactivation to scanning prevents decay, so faster scanners reactivate more items and are able to produce a longer response. This argument concerns verbal memory only and may only be applicable to the relationship between articulation rate and memory span. However, it may be the case that scanning speed affects both verbal and spatial span when serial position information is encoded and used to determine order at recall. Why articulation rate should be related to scanning speed when there is no verbal content is not clear, unless general processing rate limitations affect both tasks. However, it does seem at least possible, given the suggestions of Jones et al. (1995), that requiring serial order in recall may be in part responsible for the relationship between articulation rate and spatial span found by Smyth & Scholey (1992).

In the present study we explore the relationship between articulation rate and different types of visuospatial memory task. These tasks differ in several ways, the first of which is the need for the preservation of serial order. One of the tasks is the serial spatial span task described above, in which items are presented one at a time and recalled in the same order. However, spatial arrays can be presented and tested by recognition with no demands on order at all, making the test one of pattern recognition. If there is a relationship between ordered spatial recall and articulation rate based on scan rate for serial position information, then visuospatial tasks in which there is no serial order component at either presentation or recall should not be related to articulation rate. The task used by Logie et al. (1990) for measuring visuospatial memory span was based on the span test of visual memory for pattern (Wilson, Scott & Power, 1987), which involves the presentation of an array of black and white rectangles. These make up a meaningless pattern for which it is difficult to use a verbal description. If subjects are presented with the same pattern again, except for one item that is white instead of black, then the subject's decision about which item is missing can be made quickly, and does not involve order information in any way. This is called a span task in that the number of squares can be increased until error is produced consistently. Wilson et al. (1987) showed that adults could remember an average of 14.3 black squares in an array and have argued that this task provides an estimate of visual memory span which is larger than that which can be produced with recall procedures. Recall procedures for visuospatial span lead to recall of around five or six items (Orsini, Grossi, Capitani, Laiacona, Papagno & Vallar, 1987; Smyth & Scholey, 1992). Rapid decay from a visual memory system would mean that recall procedures underestimate the amount of visual material which can be held for a short time. There does not seem to be any reason why performance on such a task should be related to articulation, and while the task has similarities to the spatial span task it has few similarities to digit span as there is no requirement for serially ordered recall.

Visual pattern span, tested by recognition of which item is changed, has no serial order component. If it is ordering which mediates the relationship between recall and articulation rate, then visual pattern span should not be related to articulation rate. Spatial serial span and visual pattern span differ in presentation (sequential vs. simultaneous) and in responses (recall vs. recognition). Recognition memory for pattern span also encourages guessing and selective strategies in which only part of the array may be maintained and inspected to produce correct answers on a proportion of trials. If subjects were asked to respond to the pattern span by recalling which items of the array were black, and so had to produce them one at a time, strategic differences between pattern span and spatial span are not so important and output processes could interfere with the absolute amount of visual material which can be held, as Cowan et al. (1992) have suggested for verbal span.

In the third task used here, the recall version of the pattern span task, presentation is simultaneous and recall is required but there are no constraints on serial order. Thus, in the two versions of pattern span the same material is presented but the response required from the subject is either pointing to one location or pointing to a sequence of locations. In addition, there is a greater recall component when all the items have to be reproduced than there is when only the missing item has to be recognized. The fourth and final visuospatial task uses sequential presentation of a set of items with a recognition task at test. Two versions were possible, one asking for recognition of items and one asking for recognition of order over a given set of items. To keep the kinds of response made by subjects equivalent over the pattern and spatial span tasks, the recognition task used involved the sequential presentation of items, but memory performance was tested by presenting a simultaneous array with one item missing and asking the subjects to identify the missing item. Thus, of the four visuospatial memory tasks two involve order at presentation and two involve presentation of simultaneous arrays. One of each of these requires recall of a set of items at test while the remaining conditions required recognition of items at test. As we have said, these four tasks differ in many ways. However, they are all visuospatial memory tasks, but only one requires serial order at presentation to be reproduced at test.

The basic questions in this study, therefore, are concerned with the correlations between articulation rate and the memory tasks, and between the verbal and spatial memory tasks. If scanning for serial position information mediates the relationship between span tasks and articulation rate, then the correlation between articulation rate and pattern memory span should be small, and should be less than that found between articulation rate and both digit span and spatial recall span. Digit span and spatial recall span should also be more closely correlated than digit span and pattern memory. A final question is that of the relationship between spatial pattern memory and sequential spatial memory and whether there is a separate component of the latter which is attributable to the need for maintenance and retrieval of serial order information.



Subjects were 75 undergraduate students of the University of Lancaster, who volunteered and were paid for their services. Ages ranged from 19 to 36 years with a median of 21.


Seven tasks were used in all, but only six are reported here. These were: spatial recall span (sequential presentation with ordered recall), spatial recognition span (sequential presentation with recognition), visual pattern recall span (simultaneous presentation with recall), visual pattern recognition span (simultaneous presentation with recognition), digit span, and repeating digits as quickly as possible, which gives a measure of articulation time. The visual stimuli for the span tasks were presented, and the responses recorded, by a Macintosh IICX computer fitted with 21 in Aydin Ranger monitor and a Micro Touch touch screen. Digit span was also presented under computer control using digitized verbal materials prepared using Macrecorder. The Brooks matrix task was used as the last task in the sequence for each subject and is not reported here as it is a relatively impure measure of visuospatial processing.

Visual pattern span: Recall. As this was the first task presented, the demonstration of the task included familiarizing subjects with the touch screen and training on how to touch a sequence of targets on the screen. Following this, subjects were told that they would be shown a group of squares on the screen, half of which would be black and half white. The task was to remember the black squares. The array would then turn white and the subject's task was to touch those squares which had been black.

The task was based on that developed by Wilson (1991). Even numbers of rectangles, each 41 x 39 mm, were arranged so that they were touching and made a rectangular display, or an L-shaped display, starting with a 4 x 2 set of blocks, and ending with a 5 x 6 set, incrementing by two blocks every four trials. An example is shown in Fig. 1. Half of the blocks were black, half were white, and display sizes increased two blocks at a time from eight to 30 blocks. There were therefore from four to 15 black blocks in a set, with four trials at each set size giving a total of 48 trials per subject. The patterns displayed were the 'A' set from those developed by Wilson, which had been created to avoid simple groupings in the pattern. In each trial, a tone was followed by the presentation of the pattern of 2 s, the array was then replaced by an outline of empty blocks and the subject responded by touching half of the blocks shown, in any order.

Visual pattern span: Recognition. This task was similar to the previous one and used cells of the same size. The displays were created in the same way, but pilot testing indicated that the recognition test was easier, so the total number of items shown was increased, and testing began with six black blocks. The arrays used were the 'B' series developed by Wilson (1991), with the addition of four additional random patterns with 32, 34 and 36 blocks in the total display, up to 6 x 6 array. Subjects were told that the task was to remember the pattern and then to detect which block was missing when they were shown the same pattern again. On each trial a tone was presented, followed by the test pattern for 2s followed by a 3s interval during which the pattern was not present. The pattern then reappeared minus one item, and the subject was required to touch one of the white blocks to indicate the missing block. Practice was given with five black and five white blocks, and testing began with six black and six white blocks and continued to 18 black and 18 white blocks, with four trials at each set size.

Spatial span: Recall. An array of nine squares was used to present the spatial spans. This was based on the Corsi blocks task as shown in De Renzi & Nichelli (1975), and did not use a symmetrical array. The outline of the array is shown in Fig. 2. The blocks in each case were white rectangles outlined in black, which turned black when they were selected in the span sequence, and also turned black when they were touched by a subject. The blocks were 4.4 cm square and the maximum distance between the centres of any two blocks was 28 cm, while the minimum distance was 6 cm. During span testing a sequence of blocks turned black one at a time, at intervals of 1500 ms, with a tone signalling the beginning and end of the sequence. Subjects responded immediately by touching a sequence of blocks in the blank array on the screen. Subjects were instructed that the task was to remember the sequences and to repeat them by touching the appropriate squares in the correct order. Items were chosen at random for each trial with no repetitions within a trial. The task was demonstrated, practice was given on two sets of three items and testing began with four trials on four items and continued, increasing by one item every four trials, until four eight-item sets had been presented.

Spatial span: Recognition. This task used an array of blocks of the same size as those in the ordered recall task, but increased the set of blocks to 15. This was done to maintain a large set of unselected items and reduce attention to the unselected items as the set size increased. The maximum distance between blocks was 32 cm and the minimum distance was 6 cm. The blank display can be seen in Fig. 1. Each item in a sequence was presented for 1500 ms with a tone at the beginning and end of the sequence. After a 3s delay the blank array was filled with a simultaneous presentation of all the stimulus items minus one and the subject was asked to touch one square to indicate that it was missing from the array.

The missing square was chosen at random on each trial. Practice was given with three items and testing began at four items and continued to 10 items with four trials at each set size.

Digit span. Digit span was presented aloud using a recorded male voice. The digits were taken from the forward and backward span sets in the WISC, giving four lists at each set size from four to seven. Four random sequences of eight digits were used to extend the set size to the adult range. Each digit lasted for 750 ms and there was a 1 s interval between digits. On each trial, the subject touched a panel on the screen to indicate readiness, the panel disappeared, the digits were presented, the panel reappeared and the subjects recalled aloud. Accuracy of recall was scored by the experimenter. Practice was given at set size three, and subjects were told that the task was to recall the list of digits in the order in which it was presented.

Articulation time. Subjects were asked to repeat the digits 1, 2, 3, 4, 5 aloud as quickly as possible 15 times. The experimenter indicated when the repetition of digits was to stop. This was recorded and measured using a stop-watch. The measurement was carried out twice before the battery of memory tests and twice after the memory tests, making four measurements in all.

General procedure

Subjects were told that they would carry out a series of memory tests, but first they were asked to provide articulation time measures, and they were told that this would be repeated at the end. They were seated in front of the computer screen throughout the tests. Short breaks were provided between tasks. General familiarization with the computer and the touch screen was provided before testing on the pattern span task began. Accurate recall and recognition were emphasized throughout. All subjects carried out the tests in the order in which they are presented above. This means that simultaneous presentation of the visuospatial material takes place before sequential presentation so that subjects are not cued to a sequential scanning strategy.


The number of trials correct out of four was calculated for each subject for each set size in the five memory span tasks. This number was then divided by four and added to the number which was one below the starting set size in the test trials. This gives a span equivalent score for each subject, which is based on the number of trials correct throughout the testing sequence. Means for these scores can be found in Table 1. Articulation time was a total time to repeat five digits 15 times, averaged first over the two early trials and over the two late trials and then over all four. The means can be found in Table 2.

Table 2. Articulation time means measured before and after the
memory tasks, and the correlation between these

          Before          After                      Correlation
       memory tasks   memory tasks   Overall mean    before/after

Mean      13.19           12.39         12.79         r = .882

SD         1.75            1.52          1.59

Skewness was not significantly different from zero for any measure except for pattern recognition, which was significantly negatively skewed (p [less than! .01), and kurtosis was not significantly different from zero on any measure. Reflection and transformation of scores to remove the negative skew did not lead to any differences in the patterns of results reported below. However, the negative skew indicates that a ceiling effect may have been operating on the pattern recognition task and this is discussed further below.


The five span equivalents for the memory tasks and average articulation time over four trials were intercorrelated. The correlation matrix is shown in Table 3. All the spatial tasks are quite strongly correlated, but digit span is only weakly related to them. The four spatial memory tasks are highly interrelated. In particular, the correlations between pattern recall and pattern recognition, and pattern recall and spatial recall, are as large as the correlation of .71 found by Wilson (1991) between two successive tests with the A and B form of the pattern recall test. The spatial recognition task, on the other hand, is less well correlated with the other tasks, and its correlation with pattern recall is significantly less than the correlation of pattern recognition with pattern recall (t = 2.99, p [less than! .05). This is the only significant difference in correlations between the spatial tasks.

Rather surprisingly, articulation time correlates significantly with every memory measure. None of these correlations with articulation rate is significantly different from any other. This pattern suggests that retrieving serial order information is not the link to articulation rate, and this is supported by the finding that the correlation between the pattern recognition span (which has no order component) and articulation rate is not the smallest. In addition, digit span and spatial recall span are not significantly related, which would be expected if the constraints of scanning for serial order information affected both in the same way.

Regression analyses

If performance in the spatial recall span task is produced by a combination of factors including a visuospatial memory store and by scanning for serial position information, and articulation rate is also in some way linked to scanning speed, then unique variance in performance in the spatial span task should be predictable by articulation rate and by visuospatial memory. The hypotheses under investigation concern the relative contributions of visuospatial pattern memory and articulation rate to spatial recall span and digit span tasks. However, pattern recognition may be limited at the top of the range in this study as the span scores are negatively skewed, so both pattern recognition and pattern recall span are used as a set of predictor variables in hierarchical regression analyses, together with another set containing mean articulation time before the memory tasks and the mean articulation time after the memory tasks were carried out. These two sets are used hierarchically, with pattern memory first for spatial recall span and spatial recognition span, and articulation time first for digit span.

Spatial span: Recall. Multiple regression using both visual pattern span scores as a set was significant (F(2,72)= 43.925, p [less than] .01) and accounted for 55 per cent of the variance in spatial span. Adding the set of articulation time measures accounted for a further 1.7 per cent, which was not significant (F(2, 70) = 1.381). There is a second argument for spatial recall span which is that it shares variance with digit span, as both require sequential recall, even if articulation time is not causally related to span performance. However, when digit span was added after the pattern span measures only a further .02 per cent of variance was accounted for.

Spatial span: Recognition. The set of pattern tasks accounted for a significant 43.6 per cent of the variance in spatial recognition (F(2,72)=27.818, p [less than] .01), and articulation time accounted for a further 0.5 per cent of the variance, which was not significant (F [less than! 1).

Digit span. For digit span, articulation time is the main predictor and this was entered first, accounting for 15 per cent of the variance (F(2,72)= 6.368, p [less than] .01). The addition of pattern recall span accounted for a further 3.5 per cent of variance, which was not significant (F(2, 70) = 1.50, p [greater than] .05). Adding all other tasks increased the proportion of variance accounted for by a further 0.3 per cent. Digit span does not have a significant relationship with other variables in this study after articulation time is removed. This can be seen in the matrix of partial correlations between the memory measures after articulation has been removed shown in Table 4. If the order of entry in the regression analysis is reversed, however, all spatial memory tasks account for 8.2 per cent of the variance in digit span but articulation rate accounts for a significant further 10.6 per cent of variance (F(2, 68) = 4.38, p [less than] .05).
Table 4. Partial correlations between four visuospatial memory
and digit span, after articulation rate has been removed (N = 75)

                      1      2      3      4    5

1. Pattern recall     1

2. Pattern recog.   .711     1

3. Spatial recall   .680   .626     1

4. Spatial recog.   .549   .602   .562     1

5. Digit span       .187   .080   .123   .127   1

All visuospatial memory tasks

The analyses above suggest that articulation rate does not make a specific contribution to any spatial memory task, although it is correlated with all of them. Using another visuospatial task as a predictor in these analyses does not alter this picture. It is unlikely that any of these tasks has a particular relationship with articulation rate which causes all the correlations between articulation rate and spatial and pattern tasks. These four tasks do not seem to relate to one another on the basis of order. A factor analysis of the results of the four visuospatial memory tasks was carried out and two factors were selected, although the eigenvalue for the second was only 0.446. The spatial recognition task was the only task to load on the second factor, but as this is not a convincing solution given the low eigenvalue there is no clear evidence for a two-factor structure in performance of these tasks, and no evidence that ordered tasks are separable from non-ordered tasks. The correlation between the visuospatial tasks as a group and articulation time is not explained by any of the measures used here.


The four measures of visuospatial memory used here are highly intercorrelated, but there is no evidence that there are separate serial position and visuospatial processes operating. The correlation between spatial span and pattern recognition is as high as that reported elsewhere between two versions of the pattern recall task. As we have reported previously (Smyth & Scholey, 1992) there is a significant relationship between articulation time and spatial span score, but this relationship is also found for pattern span. The study was carried out to investigate the relationship between articulation rate and visuospatial short-term memory tasks that differ with respect to order information in presentation and recall. The relationship between the visuospatial memory tasks used here is very strong, whether they have serial position demands or not. Articulation time is significantly correlated with all the immediate memory tasks used, including pattern memory with recognition of a change in the pattern at test. However, articulation time does not explain any variance in spatial recall span over and above that explained by pattern memory. We have found no support for the hypothesis that articulation time predicts recall in spatial memory span because memory span tasks involve serial position, and no support for the view that spatial and pattern span tasks use visuospatial memory in different ways.

Further explanation of the differences between serial spatial span and pattern span will probably be more fruitful using interference techniques rather than by exploring the correlations between performance on the two types of task. However, the verbal working memory literature does emphasize the predictive nature of differences in articulation rate in determining memory span. The data reported here suggest that that relationship is just as great with other span tasks which share very few features with verbal memory span, but that verbal memory span and visuospatial memory performance are only weakly related. Only one correlation between digit span and visuospatial memory measures was significant in this study and that did not involve the spatial memory recall task. In Smyth & Scholey (1992) the correlation between digit span and spatial recall span also failed to reach significance, although there was a significant relationship between digit span and spatial span performance with articulatory suppression at presentation, which suggests that there are some relationships between visuospatial and verbal memory performance but that these are not reliable. Visuospatial memory and verbal memory may have relationships with articulation rate for different reasons, but it may also be the case that caution is required if articulation rate is taken to be a pure measure which explains verbal memory span performance directly.

It is possible that articulation rate is related in some way to general immediate memory ability, to general speed of processing or to intelligence. The finding in Smyth & Scholey (1992), that reading time did not correlate significantly with spatial span performance while the time to repeat digits did so, suggests that intelligence is not the whole answer. Stanovich, Cunningham & Feeman (1984) found a strong relationship between word naming times and scores on Raven's Progressive Matrices, which suggests that general intelligence would be more accurately represented by the speeded word reading task used in Smyth & Scholey (1992) than in the repeated articulation task.

While repeating the finding that articulation rate and verbal memory span are correlated in children, Cowan et al. (1994) have argued that the connection between articulation rate and verbal memory span is not causal. As the time available for covert rehearsal during immediate verbal span recall is much less than that required to repeat the memory set, it is unlikely that only covert rehearsal is involved in individual differences in memory span. Rather, they suggest that a combination of several relevant processes such as item identification, memory search, planning responses and/or covert rehearsal may all be accomplished more quickly by subjects with higher spans. They propose that the main causal variable in the relationship between articulation rate and verbal memory span performance may be the memory search rate, possibly because there is a relatively fixed ratio between the time it takes a subject to conduct a memory search and the minimal time it would take the same subject to pronounce the list. That is, a general speed factor is involved in both articulation rate and verbal memory span.

Scores on all the visuospatial tasks in the present study are highly intercorrelated and all correlate significantly with articulation rate, but this may mean that all the visuospatial tasks use the same visuospatial resources and are affected by general processing speed. All the memory tasks require scanning at input, covert rehearsal, pattern formation and other processes in which speed of processing may be a limiting factor. We have no measure of spatial processing independent of processing speed which could be used to show that articulation rate predicts unique variance in visuospatial memory performance over and above unspeeded visuospatial memory processes. No such tasks were included because our original question concerned scanning rate to determine serial order rather than processing speed per se. However, as the relationship between spatial memory performance and verbal memory performance is small it may be the case that the range of speeded processes which Cowan et al. propose can be separated into processes which are specific for dealing with language materials and processes which support memory performance on materials of different types. Our attempt to find evidence for a general memory scanning speed factor which accounted for serial position effects in both spatial and verbal recall, as suggested by Jones et al. (1995), has not been successful. However, general speed of processing can be used to account for the relationship between articulation rate and visuospatial memory performance. This adds some support to Cowan et al.'s view that articulation rate does not predict verbal memory span because it determines covert rehearsal time. If articulation rate is an indicator of speed of processing, even it is a weak indicator, then it would be expected to relate to performance in a range of tasks, although increased overlap with specifically verbal processes would lead to a stronger relationship with verbal memory performance.


Baddeley, A.D. (1986). Working Memory. Oxford: Oxford University 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.D. & Wilson, B. (1985). Phonological coding and short-term memory in patients without speech. Journal of Memory and Language, 24, 490-502.

Burgess, N. & Hitch, G. J. (1992). Toward a network model of the articulatory loop. Journal of Memory and Language, 31, 429-460.

Cowan, N. (1992). Verbal memory span and the timing of spoken recall. Journal of Memory and Language, 31, 668-684.

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.

Cowan, N., Keller, T. A., Hulme, C., Roodenrys, S., McDougall, S. & Rack, J. (1994). Verbal memory span in children: Speech timing clues to the mechanisms underlying age and word length effects. Journal of Memory and Languages, 33, 234-250.

De Renzi, E. & Nichelli, P. (1975). Verbal and non-verbal short memory impairment following hemispheric damage. Cortex, 11, 341-354.

Ellis, N. C. & Hennelly, R. A. (1980). A bi-lingual word length effect: Implications for intelligence testing and the relative ease of mental calculation. British Journal of Psychology, 71, 43-52.

Hulme, C., Maughan, S. & Brown, G. D. A. (1991). Memory for words and non-words: Evidence for a long-term memory contribution to short-term memory tasks. Journal of Memory and Language, 30, 685-701.

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.

Jones, D., Farrand, P., Stuart, G. & Morris, N. (1995). The functional equivalence of verbal and spatial information in serial short-term memory. Journal of Experimental Psychology: Learning, Memory and Cognition (in press).

Logie, R. H., Zucco, G. M. & Baddeley, A.D. (1990). Interference with visual short term memory. Acta Psychologica, 75, 55-74.

Milner, B. (1971). Interhemispheric differences in the location of psychological processes in man. British Medical Bulletin, 27, 272-277.

Naveh-Benjamin, M. & Ayres, T. J. (1986). Digit span, reading rate, and linguistic relativity. Quarterly Journal of Experimental Psychology, 38A, 739-751.

Orsini, A., Grossi, D., Capitani, E., Laiacona, M., Papagno, C. & Vallar, G. (1987). Verbal and spatial immediate memory span: Normative data from 1355 adults and 1112 children. Italian Journal of Neurological Science, 8, 539-548.

Smyth, M. M. & Scholey, K. A. (1992). Determining spatial memory span: The role of movement time and articulation rate. Quarterly Journal of Experimental Psychology, 45A, 479-501.

Smyth, M. M. & Scholey, K. A. (1994). Characteristics of spatial memory span: Is there an analogy to the word length effect, based on movement time? Quarterly Journal of Experimental Psychology, 47A, 91-117.

Stanovich, K. E., Cunningham, A. E. & Feeman, D. J. (1984). Intelligence, cognitive skills and early reading progress. Reading Research Quarterly, 19, 278-303.

Sternberg, S., Wright, C. E., Knoll, R. L. & Monsell, S. (1980). Motor programs in rapid speech: Additional evidence. In R. A. Cole (Ed.), Perception and Production of Fluent Speech. Hillsdale, NJ: Erlbaum.

Wilson, J. T. L. (1991). Visual short-term memory. Paper presented at a meeting on 'The evaluation of technology in the assessment and rehabilitation of brain damaged patients'. Bad Neuenahr, September.

Wilson, J. T. L., Scott, J. H. & Power, K. G. (1987). Developmental differences in the span of visual memory for patterns. British Journal of Developmental Psychology, 5, 249-255.
COPYRIGHT 1996 British Psychological Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Smyth, Mary M.; Scholey, Keith A.
Publication:British Journal of Psychology
Date:May 1, 1996
Previous Article:Exploring Applied Psychology: Origins and Critical Analyses.
Next Article:Mother tongue, language of schooling and bilingual digit span.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters