International Journal of

Science and Engineering Investigations

vol. 1, issue 7, August 2012 ISSN: 2251-8843

Memory and Higher Cognition

AbstractWorking memory (WM) can be defined as the system which actively holds information in the mind to do tasks in spite of the distraction. Contrary, short-term memory (STM) is a system that represents the capacity for the active storing of information without distraction. There has been accumulating evidence that these types of memory are related to higher cognition (HC). The aim of this study was to verify the relationship between HC and memory (visual STM and WM, auditory STM and WM). 59 primary school children were tested by intelligence test, mathematical tasks (HC) and memory subtests. We have shown that visual but not auditory memory is a significant predictor of higher cognition. The relevance of these results are discussed. Keywordshigher cognition, long-term memory, shortterm memory, working memory I. INTRODUCTION

Cowan et al. 6, 8 considered STM as a simple storage whereas WM as a storage with attention component. There is also opposite assumption, which believes that WM is a part of STM 9. Thirdly, there is also evidence, that WM and STM are isolated functions or functions with only a small overlap 10,11. Thus, it is difficult to define WM. Cowan 12 reported that there is no consensus of WM definition in the community of psychologists. WM can be described as a system of holding small amount of information in mind that is easy retrievable 12 or a system that operates via dynamic interaction between memory and executive attention processes 7. WM is a function which makes possible to maintain task goals in the face of interference 13, 14. The most important word in WM definitions are retrievable and interference. WM shares the first word with STM whereas the second word is characteristic only for WM. WM, for us, is the system which actively holds information in the mind to do tasks in spite of the distraction. In the contrary, STM is a system that represents the capacity for active storing of information without any distraction. This hypothesis wants to show that WM and STM are partially overlaid but concurrently unique. The basal differences between WM, STM and LTM were shown. However, which of them is responsible for the connection with higher cognition is unknown. Many studies highlighted significant correlation between memory and higher cognition 1521. There is strong evidence that WM correlates with higher cognition more than STM. For example, Daneman and Carpenter 22 found that reading span, which is traditionally considered as a measurement of WM, correlated more highly with several measures of fluid intelligence than did simple word span, which measures STM. In this and other similar studies there is an implicit assumption of separation between WM and STM. In contrast, other studies consider connection between these types of memory. Other discussed type of memory related to higher cognition is LTM. Mogle and coworkers tested the participants in processing speed, STM, WM, LTM and uid intelligence. It was shown that LTM component was the strongest predictor of uid intelligence. In contrast, WM did not predict

THERE is increasing evidence of the relationship between memory and higher cognition. However it is not still clear what is the basis of this relationship. The nature and the types of memory have been hotly debated for a long time. Usually, the memory is divided into two main structures. The long-term memory (LTM) can store vast amount of information for a longer time whereas the short-term memory (STM) represents active maintain of some little pieces of information for a shorter time 1. In psychology, this distinction has a long history dated back to William James 2. Despite it, some researchers demonstrated existence of a one-store model. But neuropsychological studies confirmed mostly the distinction between STM and LTM 3. Even more complicated situation concerns the relationship between STM and working memory (WM). Usually three possibilities are assumed. Firstly, WM cannot be removed from the construct of STM. WM and STM share overlapping neurocognitive areas and both of them measure the same ability, i.e.WM and STM are more or less the same constructs 46.Secondly, the relationship between WM and STM can be explained in terms of superiority and inferiority. Some studies have shown, that STM is a subset of WM 3, 7, 8

54

signicantly the variability in uid intelligence after accounting for the variance in uid intelligence associated with the LTM 19. But there is also evidence, that the LTM processes are not the only ones that cause the relationship between memory and higher cognition. Unsworth and his colleges 21 also considered STM processes to be important in the relationship between memory and higher cognition. They have shown that individual differences in working memory capacity (WMC) and subsequently in fluid intelligence are caused by two memory processes the active maintenance (STM) and the cue-dependent search (LTM). They came out of the idea that STM capacity is approximately four items 23. When more than four items are presented, the items currently within STM are displaced and consequently must be recalled from LTM by means of the cue-dependent search process 21. Thus, this group of authors considered the STM processes important in relationship between memory and higher cognition. These conclusions were confirmed by Shelton and coworkers 18 who investigated the relationship between processing speed, WM, LTM, STM, and fluid intelligence. They demonstrated that all constructs are significantly correlated with fluid intelligence. But only WM processes were significant in unique variance of fluid intelligence. These findings suggest that the combination of maintenance (STM) and retrieval processes (LTM) presented in WM tests makes them special in their prediction of higher cognition. It is possible therefore to conclude that the uniqueness of WM resides in the special position between STM and LTM (Fig. 1).

Fig. 2. Working memory model of Baddeley (2000) 25

Fig. 1 Simplified memory system

Both STM and WM are considered to be comprised by visual and auditory components. The most cited scheme is the scheme of Baddeley 24. They developed multi-component model of WM, which was based on Atkinson and Shiffrin modal model 26. This original Baddeley model was extended several times and its last version is given in Fig. 2.

Baddeleys model contents two stores auditory (phonological loop) and visual (visuospatial sketchpad). The newest part of this model is the episodic buffer 24. Two stores actively maintain information and the central executive is responsible for the focus attention to relevant information and for the coordination of cognitive processes which occur simultaneously. The fourth part, the episodic buffer, was added because of some facts which could not be explained by original model (for example, how verbal and spatial types could be combined, how abstract and other modality information could be remembered). Thus, the role of the new component is holding representations that integrate phonological, visual, and spatial information and information from other modalities. Concerning the differences between visual and auditory memory Baddeleys model advocates the differences between these memory types. But there is one trouble. In many visual memory tasks the participants are asked to remember list of some visual stimulus. This stimulus can be coded be visuospatial sketchpad but also by phonologoical loop. For example, the participants are asked to remember a series of pictures which are presented in onesecond intervals. Participants get visual stimuli but also can get auditory ones (because of silent or laud repeating). In contrast to Baddeleys model, Cowan 12 supposed that phonological and visuospatil stores are over-specified. He cited the experiment of Conrad (1964) which demonstrated the acustic confusion in the case of remembering printing letters 27. Newer studies offered similar results 28. The acustic confusion effect was shown not only in case of remembering English written words but also in remembering logographic Japanese Kanji. Phonologically similar words caused more confusion than visually similar words in both languages. Based on these findings, Cowan 8 presented a new model of WM. In his model (Fig. 3), the visual and the phonological stores are considered instances of the temporary activation of LTM. The idea of the separated stores is replaced by the idea called the focus of attention.

International Journal of Science and Engineering Investigations, Volume 1, Issue 7, August 2012 ISSN: 2251-8843 www.IJSEI.com

55 Paper ID: 10712-09

Fig. 3 Working memory model of Cowan (1988) 8

In summary, there is a relative agreement that memory is connected with higher cognition. Some inconsistencies are in finding of memory types which are responsible for this relationship. So, the aim of the present paper was to identify the nature of the relationship among memory and intelligence and to specify the link between visual and auditory stores.

subtest was not ordinary associated with WM, but according to WM definition presented above was convenient. The distraction can be in other modality. In the case of Bead Memory the distraction is created by searching for appropriate bead in a box. Visual STM was measured by Memory for Objects from Stanford-Binet Intelligence Scale 30. The participant is presented with sequences of single pictures and subsequently is challenged to show pictures in serial order. Auditory WM was measured by backward Digit Span from Stanford-Binet Intelligence Scale 31. Auditory STM was measured by forward Digit Span from Stanford-Binet Intelligence Scale 30. LTM was measured by Memory for Names from Woodcock-Johnson Inteligence Test 31. It is a test with supervised learning. Participants are presented with pictures of aliens and with their unusual names and ask to memorize them. Subsequently they are asked to assign aliens name to the picture. C. Analyses The normal distribution was not proved according to Shapiro-Wilk W test of normality (not shown) and therefore nonparametric methods were used. The differences between male and female were evaluated using Mann-Whitney U test.. The correlations between memory tests and higher cognition were examined by Spearman's rank correlation coefficient. A value of P < 0.05 was considered significant. Statistical analyses were performed with Statistica 6.1 software (StatSoft, Tulsa, OK). III. RESULTS

II.

METHODS

A. Participants Sixty children were recruited from primary schools (22 females, 38 males; mean age = 11.2 0.87). There were no significant between-group age differences and between-group gender differences. Similarly no gender differences were found in intelligence and mathematical tasks (Fig. 4). B. Measures Participants completed an intelligence test and mathematical tasks as well as five tasks measuring each of the following cognitive constructs: visual WM, visual STM, auditory WM, auditory STM and LTM. Fluid intelligence was measured by The Raven's Standard Progressive Matrices (SPM) 29. Mathematical ability was measured by didactic tasks according to childrens classes. Visual WM was measured by Bead Memory from Stanford-Binet Intelligence Scale 30. The participant was presented with pictures of a bead design and asked to replicate it from memory. This

The goal of the analyses was to examine the relationships among the LTM, visual WM and STM, auditory WM and STM, and fluid intelligence. The descriptive statistics are shown in Table 1 (end of paper). Only one variable revealed a single value greater than 3.5 standard deviation above or below the mean of the respective variable. The results were not changed if this value was replaced by the mean 3.5

Fig. 4 The results of intelligence test (standard progressive matrices, SPM) and mathematical tasks in females (F) and males (M).

International Journal of Science and Engineering Investigations, Volume 1, Issue 7, August 2012 ISSN: 2251-8843 www.IJSEI.com

56 Paper ID: 10712-09

standard deviation. Thus, the raw values were used in the following analyses. The values of skewness and kurtosis of all variables reached generally accepted values. However most of the values were not normally distributed (Shapiro-Wilk W test), therefore nonparametric methods were used for further analyses. To identify the relationship among the studied variables correlation analysis was performed. The results are summarized in Table 2. The results show that fluid intelligence (SPM) significantly correlates with visual type of memory (Bead Memory and Memory for Objects) and with LTM (Memory for Names), however there was not find any correlation between fluid intelligence and auditory memory (both type of Digit Span). The strongest correlation with fluid intelligence was found in case of Bead Memory (visual WM), followed by Memory for Objects (visual STM) and Memory for Names (LTM). Weak correlation was found in case of backward Digit Span (auditory WM) and none in case of forward Digit Span (auditory STM). In most cases of short memory types (WM, STM) the strongest correlation was found between memory of the same modality (Beat Memory and Memory for Objects, forward and backward Digit Span) and weaker between memory of the same type (backward Digit Span and Beat Memory). The strongest correlation of LTM (Memory for Names) was found with backward Digit Span. In addition, there was also strong correlation between LTM and mathematical ability. To verify the relationship between WM and STM and between visual and auditory memory the indexes of WM (counted as a sum of Bead Memory score and forward Digit Span), STM (counted as a sum of Memory for Names and backward Digit Span), visual memory (counted as a sum of Bead Memory and Memory for Name) and auditory memory (counted as a sum of forward and backward Digit Span) were introduced and correlations between these indexes and LTM, mathematical ability and SPM were calculated, respectively (Table 3). All indexes except index of auditory memory correlated significantly with SPM. In contrast, the mathematical ability correlated significantly with all memory types. Correlation between WM and STM was higher than correlation between visual memory and auditory memory. Based on the correlation data we have proposed a memoryintelligence model (Fig. 5) that demonstrates unique position of WM in the link between memory and intelligence. WM correlated with all variables more strongly than correlated STM vs, LTM, STM vs. SPM, and LTM vs. SPM. IV. DISCUSSION The aim of the current study was to investigate the relationship between memory and higher cognition in context of different memory types. Our data show that STM and WM are overlapping constructs as published by others 3, 7, 8, support previous ndings that memory is associated with

Fig. 5 The model of memory-intelligence.

higher cognition and demonstrate a stronger correlation between WM and intelligence than between STM and intelligence 21, 23. The memory-intelligence model (Fig. 5) implicates that WM is predominantly responsible for relationship between memory and intelligence and our data justify WM cognitive training for improvement of intelligence 32, 33. The major novel result of this study is demonstration that visual but not auditory memory is a significant predictor of uid intelligence. The question is therefore whether visual and auditory memory can be separated and their contributions to higher cognition are completely different. As shown in Table 3 WM and STM correlate more strongly than visual and auditory memory. However, in respect with previous studies 12, 27, 28 we assume that division memory into visual and auditory branch need not be self-evident. These presumptions were confirmed by our own experience with individual testing. To remember pictures many children assisted themselves by loudly repeating the visual stimuli. It is possible therefore, that some visual memory tests do not measure only visual memory. In case of auditory tests the situation is not inverse. Although the participants can also assist themselves by visual imagination it is impossible for them to create a real stimulus. It seems therefore that visual memory might be superior and that the relationship between visual and auditory memory might be analogous to the relationship between WM and STM 3, 7, 8. The relationship between visual and auditory memory could be depicted as shown in Fig. 6.

Fig. 6 Simplified scheme of visual WM (WMv), visual STM (STMv), auditory WM (WMa), and auditory STM (STMa).

International Journal of Science and Engineering Investigations, Volume 1, Issue 7, August 2012 ISSN: 2251-8843 www.IJSEI.com

57 Paper ID: 10712-09

Significant relationship between LTM (Memory for Names) and backward Digit Span can be explained by Cowans idea that stores are instances of the temporary activation of LTM 23, 32. However the correlation between forward Digit Span and LTM was absent in our study. This might reflect the fact that backward Digit Span is more dependent on LTM. In respect to Cowans magical number 4 (capacity of short term memory) 23 our children had to store two items into LTM in forward Digit Span on average (mean 5,7 items 4 items in STM/WM, 2 items in LTM). In contrast, backward Digit Span requires to store all items in LTM because WMC has to be able to invert the number (mean 3,9 items 4 items in LTM). Thus the data are in line with the argument WM is special in the relation with higher cognition due to its responsibility for interaction between the active maintenance of items in STM and the controlled search in LTM 23. The relationship between fluid intelligence and mathematical ability is generally accepted. Nevertheless in our study mathematical ability correlated significantly also with all memory indexes and all memory tasks with the exception of forward Digit Span. In contrast, fluid intelligence did not correlate with index of auditory memory. These findings implicate that there is something unique in the relationship between memory and mathematical ability; something what cannot be explain only by the relationship between mathematical ability and intelligence. In summary, the present study confirmed the relationship between memory and intelligence, particularly the responsibility of WM for relationship between memory and higher cognition. We have shown that visual but not auditory memory is a significant predictor of intelligence. This finding may not mean the distinction between phonological and visuospatial store but the superiority of visual memory. However, further studies will be required to prove this superiority. ACKNOWLEDGMENT This study was supported by grant agency of the Charles University (no. 298811).

6 N. Unsworth, R.W. Engle, On the Division of Short-Term and Working Memory: An Examination of Simple and Complex Span and Their Relation to Higher Order Abilities, Psychological Bulletin, vol. 133, 10381066, 2007. 7 N. Cowan, Attention and memory: An integrated framework. Oxford: Oxford University Press, 1995. 8 N. Cowan, Evolving conceptions of memory storage, selective attention, and their mutual constraints within the human information processing system. Psychological Bulletin, vol. 104, pp. 163191, 1988. 9 J. G. Seamon and D. T. Kenrick, Psychology. Englewood Cliffs, NJ: Prentice-Hall, 1994. 10 M. J. Kane, D. Z. Hambrick, S. W. Tuholski, O. Wilhelm, T. W. Payne, and R. W. Engle, The generality of working memory capacity: A latent-variable approach to verbal and visuo-spatial mem-ory span and reasoning, Journal of Experimental Psychology: General , vol. 133, 189217, 2004. 11 H. L. Swanson and D. Luxenberg, Short-Term Memory and Working Memory in Children with Blindness: Support for a DomainGeneral or Domain Specific System? Child Neuropsychology, vol. 15, pp. 280 294, 2009. 12 N.Cowan, Multiple Concurrent Thoughts: The Meaning and Developmental Neuropsychology of Working Memory, Developmental Neuropsychology, vol. 35, pp. 447474, 2010. 13 A. R. A. Conway, N. Cowan, and M. F. Bunting, M. F, The cocktail party phenomenon revisited: The importance of working memory capacity, Psychonomic Bulletin & Review, vol. 8, pp. 331335, 2001. 14 M. J. Kane and R. W. Engle, R. W, WM capacity, proactive interference, and divided attention: Limits on long-term memory Retrieval, Journal of Experimental Psychology: Learning, Memory, and Cognition, vol. 26, pp. 336358, 2000. 15 P. L. Ackerman, M. E. Beier, and M. O. Boyle, Individual differences in working memory within a nomological network of cognitive and perceptual speed abilities, Journal of E.xperimenlal Psychology General, vol. 131, pp. 567589, 2002. 16 R. Colom, I. Rebollo, A. Palacios, M. Juan-Espinosa, and P. C. Kyllonen, Working memory is (almost) perfectly predicted by g, Intelligence, vol. 32, pp. 277296, 2004. 17 R. Colom, V. Rubio, P. C. Shih, and J. Santacreu, Fluid intelligence, working memory and executive functioning, Psicothema, vol. 8, pp. 816821, 2006. 18 J. Shelton, E. Elliott, R. Matthews, B. Hill, W. Gouvier, The relationships of working memory, secondary memory, and general fluid intelligence: Working memory is special, Journal of Experimental Psychology: Learning, Memory, and Cognition, vol. 36, pp. 813820, 2010. 19 J. Mogle, B. Lovett, R. Stawski, and M. Sliwinski, What's so special about working memory? An examination of the relationships among working memory, secondary memory, and fluid intelligence, Psychological Science, vol. 19, pp. 1071-1077, 2008. 20 N. Unsworth and R. Engle, Simple and complex memory spans and their relation to fluid abilities: Evidence from list-length effects, Journal of Memory and Language January, vol. 54, pp. 68-80, 2006. 21 N. Unsworth and R. Engle, The nature of individual differences in working memory capacity: Active maintenance in primary memory and controlled search from secondary memory, Psychological Review, vol. 114, pp: 104-132, 2007. 22 M. Daneman, and P. A. Carpenter, Individual differences in working memory and reading, Journal of Verbal Learning and Verbal Behavior, vol. 19, pp. 450466, 1980. 23 N. Cowan, The magical number 4 in short-term memory: A reconsideration of mental storage capacity, Behavioral and Brain Sciences, vol. 24, 87114, 2001. 24 A. D. Baddeley and G. J. Hitch, Working memory. In G. H. Bower (Ed.), The psychology of learning and motivation: Advances in research and theory. New York, NY: Academic Press, 1974. 25 A. Baddeley, The episodic buffer: A new component of working

REFERENCES
1 A. F. Healy and D. S. McNamara, Verbal learning and memory: Does the modal model still work?, Annual Review of Psychology, vol. 47, pp. 143172, 1996. 2 W. James, The principles of psychology. London: Macmillan, 1890. 3 D. Talmi, C. L. Grady, Y. Goshen-Gottstein, and M. Moscovitch, Neuroimaging the Serial Position Curve: A Test of Single-Store Versus Dual-Store Models, Psychological Science, vol. 16, pp. 716723, 2005. 4 J. R. Anderson, Cognitive psychology and its implications. New York: Freeman, 1990. 5 R. Colom, P. Chun Shih, C. Flores-Mendoza, and M. A. Quiroga, The real relationship between short-term memory and working memory, Memory, vol. 14, pp. 804813, 2006.

International Journal of Science and Engineering Investigations, Volume 1, Issue 7, August 2012 ISSN: 2251-8843 www.IJSEI.com

58 Paper ID: 10712-09

memory? Trends in Cognitive Sciences, vol. 4, pp. 417 423, 2000. 26 R. C. Atkinson, and R. M. Shiffrin, Human memory: A proposed system and its control processes. In K. W. Spence and J. T. Spence (Eds.), The psychology of learning and motivation: Advances in research and theory, New York: Academic Press, 1968. 27 R. Conrad, Acoustic confusion in immediate memory, British Journal of Psychology, vol. 55, pp. 7584, 1964. 28 M. Flaherty, and A. Moran, Acoustic and visual confusions in immediate memory in Japanese and English speakers, Psychologia: An International Journal Of Psychology In The Orient, vol. 42, pp. 80-88, 1999. 29 J. C. Raven, J. E. Raven and J. H. Court, Progressive matrices. Oxford: Oxford Psychologists Press, 1998.

30 R. L. Thorndike, E. P. Hagen, and M. Sattler, Stanford-Binet Intelligence Scale: Fourth Edition. Chicago: Riverside, 1986. 31 R. W. Woodcock, K. S. McGrew, and N. Mater, Woodcock-Johnson III. Itasca. IL: Riverside Publishing, 2001. 32 T. Klingberg, H. Forssberg, and H. Westerberg, Training of working memory in children with ADHD, Journal of Clinical and Experimental Neuropsychology, vol. 24, pp. 781-791, 2002. 33 S. M. Jaeggi, M. Buschkuehl, J. Jonides, and W. J. Perrig, Improving fluid intelligence with training on working memory, Proceedings of the National Academy of USA, vol. 105, pp. 68296833. 2008.

TABLE I DESCRIPTIVE STATISTICS FOR ALL MEASURES Task SPM Mathematical tasks Bead Memory Digit Span forw. Digit Span Backw. Memory for Names Memory for Objects Mean 36.85 56.29 23.51 5.73 3.92 52.69 6.86 Minimum 13.00 17.39 10.00 3.00 2.00 24.00 4.00 Maximum 51.00 100.00 35.00 9.00 9.00 71.00 9.00 SD 8.53 21.91 4.45 1.58 1.49 12.65 1.21 Skew -0.71 0.17 -0.36 0.41 0.80 -0.53 0.09 Kurtosis 0.56 -0.59 1.22 -0.20 1.06 -0.60 -0.40

N = 59 for all measures. Raw scores were used for all tests except mathematical tasks. Mathematical scores were percentage of fulfilment of the test because of the different tests for different age (100% get children with higher raw scores in every age group). TABLE 2 CORRELATION MATRIX AMONG ALL COGNITIVE TASKS 1 2 3 4 5 1. SPM 2. Mathematical tasks 3. Bead Memory 4. Digit Span Forw. 5. Digit Span Backw. 6. Memory for Names 7. Memory for Objects

1.00 .45 .54 .05 .13 .34 .42

1.00 .42 .13 .39 .45 .51

1.00 .18 .40 .34 .42

1.00 .46 .19 .01 1.00 .49 .22 1.00 .42 1.00

p < .05.

TABLE 3 CORRELATIONS AMONG ALL MEMORY INDEXES AND LTM, SPM AND MATHEMATICAL TASKS 1 1. WM 2. STM 3. Visual Memory 4. Auditory memory 5. LTM 6. SPM 7. Mathematical tasks

1,00 .50 .91 .54 .45 .50 .45

1,00 .53 .76 .39 .30 .43

1,00 .32 .40 .59 .51

1,00 .41 .12 .31

1,00 .34 .45

1,00 .45 1,00

p < .05.

International Journal of Science and Engineering Investigations, Volume 1, Issue 7, August 2012 ISSN: 2251-8843 www.IJSEI.com

59 Paper ID: 10712-09