Skip to: Main content Side column

• Home • Contact Us

Journal of Adult Development© Springer Science+Business Media, LLC 201210.1007/s10804-012-9142-6

The Impact of Processing Speed Training on Working Memory in Old AdultsHuamao Peng1 , Jing Wen1, Dahua Wang1 and Yue Gao1

(1)Institute of Developmental Psychology, Beijing Normal University, No. 19 Xinjiekouwai Street, Beijing, 100875, China Huamao PengEmail: penghuamao@bnu.edu.cnPublished online: 15 February 2012AbstractCross-sectional and longitudinal studies have gotten different results as to whether processing speed can explain the aging of cognitive functions. Experimental analyses are needed to develop further evidence. To explore the relationship between speed and working memory in cognitive aging, processing speed intervention is conducted in the present study. Utilizing paper–pencil training, computer training, and a control group, 78 adults aged 58–83 years from Beijing participated in a speed intervention study. After 5 weeks of training,there were substantial training gains on two speed tests in two training groups, but no transfer effect on working memory was found. Discussion: The results fail to support the view that processing speed is a foundation of working memory for

aging. The aging of working memory may depend more on an executive attentioncomponent.KeywordsProcessing speed Working memory Training Cognitive aging

IntroductionThere has been a greater emphasis on the aging process, especially on cognitive aging as the proportion of older people in modern societies increase. Many cognitive aging psychology studies are not only concerned about the feature phenomenon caused by aging in cognitive tasks, but also about how to refine andintegrate, from the variety of views and theories currently available, to a more general cognitive aging theory, that is, the internal mechanism of age-related declines in cognitive abilities (Hershey et al. 1999).Processing speed has been regarded as a potential explanatory construct in cognitive development and aging (Birren 1974; Salthouse 1996). Birren (1974) first proposed that the decline of speed was the reason for the decline of general cognitive abilities. Evidence shows that a large number of individuals experience the slowing of processing speed (Salthouse 1985, 1990). Subsequent research has found that the aging of cognitive abilities can be attributed to the slowing of processing speed (Salthouse 1993; Schaie 1989, 1994). However, the place of processing speed within an explanatory account of cognitive aging is ambiguous and disputed. On the one hand, processing speed may be viewed as just one aspect of cognitive functioning that deteriorates along with a number of other aspects with which it is correlated and with which it might share etiology (Finkelet al. 2005; Salthouse 2004). On the other hand, processing speed is sometimes regarded as a more “fundamental property of the nervous system” (Madden 2001, p. 288) that serves a diverse range of other mental functions. When processing speed is slower, other mental functions are less well implemented. The latter view can often be supported in cross-sectional studies (Finkel et al. 2007; Salthouse 1996, 2001; Salthouse and Ferrer-Caja 2003), but not in longitudinal studies (Zimprich 2002; Zimprich and Martin 2002).Some of the evidence that support a processing speed theory comes from cross-sectional studies. The age-related variances of many cognitive abilities depend on their common variances with processing speed. Adjusting for individual differences in measures of processing speed often eliminates a majority of the age-related variation in more complex mental abilities (Zimprich and Martin 2002). The statistical control of processing speed can reduce the correlation between age and free recall (Bryan and Luszcz 1996), clue recall (Park et al. 1996), reasoning (Salthouse et al. 1998), and other cognitive abilities. A meta-analysis made by Verhaeghen and Salthouse (1997) showed that 70% of the variation in reasoning, spatial memory, and episodic memory can be explained bythe variance of processing speed.However, evidence from longitudinal studies is inconsistent. Only limited longitudinal research exists on the processing speed theory of cognitive aging. Longitudinal data show that the mediating role of processing speed in cognitive aging is less substantial. For example, Zimprich and Martin (2002) found that with the same group of subjects, the digit symbol test could account for about

85% of age-related variance with other tests in the Wechsler Adult Intelligence Scale in cross-sectional analysis but only 4% in a longitudinal analysis. A cross-lagged panel analysis of reaction time and cognitive ability over thirteen years showed that there was a significant association between verbal and numerical reasoning scores at age 56 and speed of processing at age 69 but not vice versa (Deary et al. 2009).Although the intervention effect of processing speed in older adults could transfer in some cognitive abilities and daily function (Ball et al. 2007; Vance et al. 2007), on the one hand, these transfer effects were less significant than those in cross-sectional studies, on the other hand, covariation may come from other common sources. The inconsistent results in cross-sectional, longitudinal, and intervention studies illuminate the notion that there is still an open question concerning whether processing speed is a powerful explanation of the aging of other cognitive abilities.Working memory is another important processing resource which is hypothesized to account for age differences in cognition. Processing speed is often considered a more basic resource than working memory. Many cross-sectional research studies showed that processing speed could account for more age-related variance in cognitive functions, and the age differences in cognition due to working memory were reduced after controlled speed (Salthouse 1995; Park 2000). We found a hierarchical relation of sensory functions, processing speed, working memory, and primary mental ability (inductive reasoning, spatial orientation, number skills, verbal ability, word fluency), using Structural Equation Modeling, which supported the before-mentioned views (Shen et al. 2003; Peng et al. 2004, 2006). However, this evidence comes from cross-sectionaldata. Experimental analyses are needed to develop further evidence.There is still a problem as to whether age-related changes in processing speed may be considered as a partial explanation for working memory changes. Longitudinal data are very important, but longitudinal design is time-consuming and the data results rely on statistical techniques in part. Intervention studies can attain data from causal inference. If working memory can be affected by processing speed, we can hypothesize that transferring effects may be observed in a processing speed intervention. Therefore, in the present study, we trained older adults’ processing speed to examine the relationship between processing speed and working memory.

Methods

DesignThe present study was a 2 × 3-mixed design with testing occasions as the within variable and type of training as the between variable. Prior to the start of training, all participants were assessed on a battery of tests regarding processing speed and working memory. Following the assessment, participants were randomly assigned to either the training or the control group. Approximately 7 weeks after the pre-training assessment, the participants were again assessed on the cognitive battery.

ParticipantsSeventy-eight adults aged 58–83 years (15 men, 63 women) were recruited from three communities in Beijing to participate in the study. Twenty-seven participants practiced a paper–pencil figure matching task and 25 participants had a figure comparison practice on the computer across 5 weeks for about 45 min–1 h per week, respectively. Twenty-three elders participated as an age-matched, education-matched, and no-contact control group. There was no significant difference in age and education level among the three groups. See Table 1.

Table 1 Descriptive statistic of participants’ age and education level (mean ± SD)

Control group

Paper–pencil training group

Computer training group

F

(1,75) p

Age 71.16 ± 6.33 67.22 ± 4.58 69.65 ± 6.43 3.05 .053

Education 9.20 ± 5.20 9.67 ± 3.45 9.35 ± 4.03 .08 .921

Measures

Processing Speed

Digit Comparison

For the digit comparison task, participants compared paired strings of digits (containing 3–9 digits per string) and determined whether the strings were the same or different. The digit strings were presented in the center of the computerscreen. Participants were asked to press response keys as quickly as possible. There were 7 practice trials and 84 test trials which included 12 pairs of strings for each length. The different length strings appeared randomly. The “yes” and “no” reactions accounted for 50% of the total reactions, respectively. The reaction time and accuracy rate were recorded. Reaction time for correct trials only was used in the present study. Only those participants with fewer than 20% incorrect responses were used in analyses. No participant was excluded.

Pattern Comparison

For the Pattern Comparison task, the participants selected one pattern from 5 choices which was the same as the target pattern. The target and choice patterns were presented in the center of the computer screen, where the target pattern was above the choice patterns. Participants were asked to press response keys, from left to right, 1, 2, 3, 4, 5, as quickly as possible. There were 5 practice sets and 60 test sets. The response 1, 2, 3, 4, 5 had the same chance to be the right answer. The 60 test sets were presented in random order. The reaction time and accuracy rate were recorded. Reaction time for correct trials only was used in the present study. Only those participants with fewer than 20% incorrect responses were used in analyses. No participant was excluded.

Working Memory

Operation Span

The Operation Span task involved both processing and memory storage components: Participants solved simple arithmetic problems (basic addition and subtraction) while trying to remember the answer to each equation. Each equation was presented in the center of the computer screen. Participants were instructed to calculate in their mind and remember the answer of each question, holding the entire set of answer numbers in memory until instructed to recall them in the same serial order as the equations had been presented. The answer to each equation ranged from 0 to 9. Set size varied from 1 to 9 and gradually increased. There were 3 sets with the same size. If the participants could not calculate or recall correctly two of the three sets, the task finished. If the set sizewas n when a participant finished, the maximum number of equations the participant could both calculate and recall correctly was n − 1. N − 1 was the working memory span for this participant. The equations used in the practice and formal experiment were not the same.

Spatial Span

Two 4 × 4 grids on the center of the computer screen side by side, randomly distributed some “*” in each cell. The participants were instructed to judge whether the distributions of the “*” in the two grids were the same or different and to remember the distribution of the “*” in the left grid. After the judgment, the participants were instructed to locate the “*” in the left grid on a blank grid. The number of “*” in each grid increased from 2 to 9, and each number included three trials. If the subject could not judge or recall 2 of the 3 trials correctly, the test finished. If a participant finished the test on the number n, then the maximum number the participant could judge and recall correctly was n − 1, which was the spatial working memory span. The materials used in the practice and the formal trials were not the same.

TrainingFollowing completion of the pretest battery, the participants were randomly assigned to a no-contact control group (n = 25), a paper–pencil training group (n = 27), and a computer training group (n = 26). Control participants did not receive any training, but participated in both pre- and post-test assessment sessions. The two assessment sessions for the control group were scheduled suchthat the amount of time that elapsed between the sessions was similar to that of the training participants (approximately 7 weeks apart). Both training groups included several subgroups with 6–8 participants. Each participant had the processing speed practice across 5 weeks for about 45 min–1 h per week, respectively.

Paper–Pencil Figure Matching Task

For the paper–pencil figure matching task, participants were instructed to find the same figures in a 7 × 7 figure matrix, as many as possible in a limited time. Each figure matrix was presented on one page. In order to cause and maintain

interest of the elderly subjects, these figures were closely related to their daily life, such as transportation, fruits, vegetables, animals, and so on. In each matrix,some figures were paired the same. Three difficulty levels of the training materials, from low to high, were thus: 12 pairs same figures, 9 pairs same figures, 7 pairs same figures. The proportion of the matrixes under each difficulty level was the same in each training session. The first, second, and third sessions included 9 matrixes, respectively, and the fourth and fifth sessions included 15 matrixes, respectively. In the first two sessions, the limited time was 1.5 min, and in the other sessions, the limited time was 1 min.

Figure Comparison Task

The figure comparison task was a training task practiced on the computer. Participants were instructed to select the item matching the target figure from several choices as quickly as possible. The image background was white, and pixels were 160 × 120. Each figure was a combination of no more than 3 simple geometric graphics. Four difficulty levels of training material, from low to high, were used: (1) one target and one choice figure, (2) one target and two choice figures, (3) one target and three choice figures, (4) one target and four choice figures. Participants needed to press the corresponding key as quickly as possible.The difficulty of training tasks increased gradually as the performance of the tasks completed by the elderly advanced. On average, for each participant, the proportion of four difficulty levels in the whole training was: level 1 20%, level 2 40%, level 3 30% and level 4 10%.

ResultsThe pre- and post-test performances of two processing speed tasks of training and control groups were showed as Figs. 1 and 2.

Fig. 1 The pre- and post-test performances of digit comparison (ms)

Fig. 2 The pre- and post-test performances of pattern comparison (ms)As shown in Figs. 1 and 2, the reaction time of the three groups measured in the pretests were faster than that in the post-test, especially for the two training groups, and the standard deviation showed the same trend. Cohen’s d score was calculated to evaluate the effect size between the pretests and the post-test of the three groups. In the Digit Comparison task, EScontrol = .20, ESpaper–pencil

training = .28, EScomputer training = .71. In the Pattern Comparison task,

EScontrol = .08, ESpaper–pencil training = .18, EScomputer training = .69. The effect

size of the computer training group was the largest, and the second one was thatof the paper–pencil training group.Repeated measure analysis of variance was conducted in order to determine whether there were differences among the training and control groups at the pre- and post-test assessments. In the Digit Comparison task, the main effect of pre- and post-test was significant, F(1,73) = 28.44 (p < .001), and the interactioneffect was also significant, F(2,73) = 3.21 (p < .05). In the Pattern Comparison task, the main effect was significant, F(1,73) = 16.79 (p < .001), and the interaction effect was significant, F(2,73) = 5.45 (p < .01).There were significant differences between the pre-tests and post-tests in the three groups on two tasksof processing speed.To further examine the difference between the pretest and post-test of the three groups in the two processing speed tests, respectively, a Simple Effects analysis was conducted. There was no significant difference between the pretest and post-test of the Digit Comparison in the control group, F(1,73) = 2.83, p > .05, the same for that of the Pattern Comparison, F(1,73) = .40, p > .05. A significant increase from the pretest to post-test of the Digit Comparison was found in the paper–pencil training group (F(1,73) = 6.05, p < .05) and the computer training group (F(1,73) = 25.84, p < .001). While no significant difference between the pretest and post-test of Pattern Comparison was found in paper–pencil training group (F(1,73) = 2.10, p > .05), the computer training group improved significantly (F(1,73) = 25.34, p < .001). After the training sessions, the two training groups improved in processing speed, especially for the computer training group.Substantial training gains on the two speed task in the two training groups were found. The paper–pencil training group (Digit Comparison, Δ = 462 ms; Pattern Comparison, Δ = 267 ms) and the computer training group (Digit Comparison, Δ = 975 ms; Pattern Comparison, Δ = 942 ms) had more improvement than the control group (Digit Comparison, Δ = 426 ms; Pattern Comparison, Δ = 204 ms). Did these training gains transfer to working memory? In the intervention study, was it conclusive that processing speed can explain the aging of working memory?The correlations between processing speed and working memory measures in the pre- and post-test were calculated, respectively, to examine the relationship between both on the cross-sectional level.As showed in Tables 2 and 3, consistent with prior cross-sectional analysis results in documents, at the same time point, the correlations between processing speed and working memory were very significant.Table 2 Correlation between processing speed and working memory in the pretest

Spatial Span Operation Span

Digit comparison (RT) −.404** −.498**

Pattern comparison (RT) −.254* −.437**

Table 3 Correlation between processing speed and working memory in the post-test

Spatial Span Operation Span

Digit comparison (RT) −.321** −.384**

Pattern comparison (RT) −.257* −.423**

According to Figs. 3 and 4, the post-test performances of two working memory measures were better than that in the pretest in both training groups (Spatial Span, Δpaper–pencil training = .52, Δcomputer training = .23; Operation Span, Δpaper–

pencil training = .59, Δcomputer training = .40). For the control group, the post-test

performance of Spatial Span improved (Δcontrol = .48), but that of Operation

Span decreased (Δcontrol = −.30). Cohen’s d score was calculated to evaluate the

effect size between the pretests and the post-test of the three groups. In Spatial Span, EScontrol = .29, ESpaper-pencil training = .32, EScomputer training = .13. In

Operation Span, ESpaper-pencil training = .32, EScomputer training = .05. The effect

size of the control group could not be calculated due to the scores declining. The effect sizes of the three groups were not high enough, and particularly, the advantage that the computer training group got from the processing speed training did not show on the working memory measures.

Fig. 3 The pre- and post-test performances of Spatial Span

Fig. 4 The pre- and post-test performances of Operation SpanRepeated measures analysis of variance found that: The main effect of the pretest and post-test of the Spatial Span (F(1,75) = 5.65, p < .05) was significant, but the interaction effect (F(2,75) = .276, p > .05) was not. That is to say, in both the training group and the control group, the post-test scores of Spatial Span were significantly better than that in the pretest, but this kind of increase could not be attributed to training. The main effect of the Operation Span (F(1,75) = 1.12, p > .05) and the interaction effect (F(2,75) = 1.53, p > .05)were both not significant. The differences between the pretest and post-test of the three groups were not significant, and the differences between the training and control groups were not significant. It can be concluded that processing speed training could not transfer to working memory.

DiscussionThe primary goal of the present study was to evaluate the extent of transfer of speed of processing training to working memory as a way to examine the relationship between processing speed and working memory in cognition aging; that processing speed was a fundamental explanation for the aging of other cognitive abilities, especially working memory. Initial results revealed a training

effect upon two measures of processing speed but no transfer of the training effect to two measures of working memory.The training effect of processing speed intervention was consistent with the results obtained in prior studies (Ball et al. 2007). The intervening effect had twofeatures: (1) The improvement of Digit Comparison was superior than that of Pattern Comparison. (2) The training gains of the computer training group were more than those of paper–pencil training group. These might be attributed to theproperties of test tasks and training tasks. Compared with the Pattern Comparison, the Digit Comparison was more simple and closer to the definition of processing speed. The components of processing speed included target detection, identification, recognition, and location (Ball et al. 2002). It was generally assessed by tasks that involve items with simple mental content that would rarely or never result in incorrect solutions if there were no time pressure (Deary 2000; Salthouse 1996). The contents in the Digit Comparison included target detection, identification, and recognition: first detecting the digit string, identifying the two strings of digits, and finally recognition. In addition to these components, the Pattern Comparison also included target location, which made itmore difficult than the Digit Comparison. Furthermore, the subjects were more likely to make mistakes in the Pattern Comparison, as they just needed simple comparisons of digit strings in Digit Comparison, while the complexity, clarity, and similarity of the figures in the Pattern Comparison might affect their response speed. From the perspective of training tasks’ properties, the protocol used in computer training was easier than that used in paper–pencil training, leading to more improvement through simple and repeated practice. Also, there were more similarities between the components included in the computer training task and the indicator tests, especially the Pattern Comparison. In addition, through the computer training, the elderly subjects could be more familiar with the use of the computer, which was good for them in completing thetest.In many cross-sectional studies, processing speed partly explained the age-related variance of working memory. Similar results were found in children’s processing speed and working memory development researches. Some cross-sectional studies showed that age-related increase in fluid intelligence was mediated by developmental changes in processing speed and working memory inchildren, and the improvement in working memory was mediated by developmental changes in processing speed. With faster processing speed, more information could be kept in working memory (Fry and Hale 1996; Li et al. 2004; Laurence et al. 2007). Some evidence from longitudinal data was consistent. For example, Hitch and colleagues (Hitch et al. 2001) found that variation in workingmemory span between 1-year interval test waves was systematically related to changes in processing speed in children aged 9–11. This relationship between processing speed and working memory could be explained by the simultaneity mechanism of the processing speed theory. Salthouse (1996) proposed that processing speed influenced other cognitive abilities through two different mechanisms, the limited time mechanism and the simultaneity mechanism. The basis for the limited time mechanism was simply that the time to perform later operations was greatly restricted when a large proportion of the available time was occupied by the execution of early operations. This mechanism was

primarily relevant when there were external time limits or other restrictions on the time available for processing, such as the presence of concurrent demands on processing. The simultaneity mechanism was based on the idea that the products of early processing may be lost by the time that later processing was completed. To the extent that this was the case, relevant information may no longer be available when it is needed. Processing deficits could therefore emergebecause of discrepancies between the time course of loss of information. Salthouse (1996) also proposed that the functioning of the simultaneity mechanism might be reflected by the working memory capacity because workingmemory was sometimes conceptualized as consisting of information that was currently available for storage or processing. The amount of simultaneously available information might be indexed by measures of the working memory.According to simultaneity mechanism, we can hypothesize that faster processing speed will lead to a higher span because it reduces the time for representations of stimulus to lose activation (Hitch and Towse 1995). However, the training effect of processing speed did not transfer to working memory in this study, contrary to the cross-sectional results. Does that mean simultaneity mechanism is not right? It is caution to conclude that simultaneity mechanism could explain the effect of processing speed on the individual differences in working memory incognitive aging and development. As to whether working memory change can be partially explained by decreasing processing speed which increases with age, further evidences from longitudinal and intervention studies are needed. It is notdenied that processing speed plays an important role in working memory development, but the maintenance and improving of working memory might depend more on other factors.Working memory is a kind of resource limited memory system that is necessary for concurrent storage and manipulation of information (Baddeley 1992). The impact of the simultaneity mechanism on working memory lies in the “processing,” because the role of the simultaneity mechanism was that the slower processing speed might extend the processing time as Salthouse (1996) postulated, instead of interference or decay of processing products in memory storage. Then, in the “processing” part of working memory, are there other affecting factors? Engel and his colleagues (Engle et al. 1999; Kane and Engle 2002) analyzed working memory from the perspective of process analysis and proposed the concept of “executive attention.” The executive attention was the component participating in a complex working memory span task which was not influenced by short-term memory span. The executive attention acted as the active maintenance of goals and inhibited the inaccurate reaction trend affected by context through maintaining the goals. Accordingly, the failure of executive attention could be seen as “goal neglect” (Kane and Engle 2003). In the processing course of working memory, faster processing speed could reduce the operation time to protect the memory from decay. However, the inhibition ability is also an important factor of working memory. Hasher and Zacks defined inhibition as the process of preventing irrelevant information to enter or remain in the working memory. The decline in inhibition will result in the activation, maintenance, and retrieval of irrelevant information, thereby affecting processing of relevant information. Researchers had found that when conductingcognitive tasks involving working memory, compared with young people, the

elderly were more susceptible to the irrelevant but significant interference information and found it more difficult to focus on the task-related information (Bell et al. 2008; Meijer et al. 2006). Therefore, during the processing, whether participants could inhibit the interference of irrelevant information and maintain the right goal information might impact on the accuracy of processing directly and thus influence the maximum span of working memory. Through training, the participants’ processing speed improved, but the executive attention did not. In this case, although the elderly subjects showed faster processing speed in working memory tests, they could not inhibit the irrelevant information well during processing. Therefore, the elders’ working memory could not be improved. It might be considered that the factors influencing scores of working memory were not only processing speed but also the ability of coordinating processing and storing, as well as the ability of inhibiting irrelevant information. Therefore, in order to improve the elderly subjects’ working memory, it was not only necessary to improve their processing speed, but also to train their ability tocoordinate and inhibit. Overall, the impact of processing speed on an aging working memory was not as large as that in cross-sectional studies. Both of themmight show a kind of covariation relationship in the aging process, and the maintenance and improving of working memory might depend more on other factors.AcknowledgmentsThis work was supported by National Scientific Funds (grant number 31000466),Program for Humanity and Social Science Research sponsored by Ministry of Education (07JCXLX003), and Program for Changjiang Scholars and Innovative Research Team in University (PCSIRT IRT0710).ReferencesBaddeley, A. D. (1992). Working memory. Science, 255, 556.PubMed CrossRefBall, K., Berch, D. B., Helmers, K. F., Jobe, J. B., Leveck, M. D., Marsiske, M., et al. (2002). Effects of cognitive training interventions with older adults: A randomized controlled trial. Journal of the American Medical Association, 13, 2271–2281.CrossRefBall, K., Edwards, J. D., & Ross, L. A. (2007). The impact of speed of processing training on cognitive and everyday functions. The Journals of Gerontology, 62B, 19–31.CrossRefBell, R., Buchner, A., & Mund, I. (2008). Age-related differences in irrelevant-speech effects. Psychology and Aging, 23(2), 377–391.PubMed CrossRefBirren, J. E. (1974). Translations in gerontology: From lab to life. Psychophysiology and speed of response. American Psychologist, 29, 808–815.PubMed CrossRefBryan, J., & Luszcz, M. A. (1996). Speed of information processing as a mediator between age and free-recall performance. Psychology and Aging, 11(1), 3–9.PubMed CrossRefDeary, I. J. (2000). Looking down on human intelligence. Cambridge, UK: Cambridge University Press.CrossRefDeary, I. J., Allerhand, M., & Der, M. (2009). Smarter in middle age, faster in old age: A cross-lagged panel analysis of reaction time and cognitive ability over 13 years in the west of Scotland Twenty-07 Study. Psychology and Aging, 24, 40–47.PubMed CrossRef

Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. A. (1999). Working memory, short-term memory, and general fluid intelligence: A latent-variable approach. Journal of Experimental Psychology: General, 128, 309–331.CrossRefFinkel, D., Reynolds, C. A., McArdle, J. J., & Pedersen, N. L. (2005). The longitudinal relationship between processing speed and cognitive ability: Geneticand environmental influences. Behavior Genetics, 35, 535–549.PubMed CrossRefFinkel, D., Reynolds, C. A., McArdle, J. J., & Pedersen, N. L. (2007). Age changes in processing speed as a leading indicator of cognitive aging. Psychology and Aging, 22, 558–568.PubMed CrossRefFry, A. F., & Hale, S. (1996). Processing speed, working memory, and fluid intelligence: Evidence for a developmental cascade. Psychological Science, 7, 237–241.CrossRefHershey, D. A., Boyd, M. L., Coutant, K. M., & Turner, K. (1999). Cognitive aging psychology: Significant advances, challenges, and training issues. Educational Gerontology, 25, 349–364.CrossRefHitch, G. J., & Towse, J. N. (1995). Working memory: What develops? In F. E. Weinert & W. Schneider (Eds.), Memory performance and competencies: Issues in growth and development (pp. 3–21). Mahwah, NJ: Erlbaum.Hitch, G., Towse, J. N., & Hutton, U. (2001). What limits children’s working memory span? Theoretical accounts and applications for scholastic development.Journal of Experimental Psychology: General, 130(2), 184–198.CrossRefKane, M. J., & Engle, R. W. (2002). The role of prefrontal cortex in working-memory capacity, executive attention and general fluid intelligence: An individual-differences perspective. Psychonomic Bulletin & Review, 9, 637–671.CrossRefKane, M. J., & Engle, R. W. (2003). Working-memory capacity and the control of attention: The contributions of goal neglect, response competition, and task set to Stroop interference. Journal of Experimental Psychology: General, 132(1), 47–70.CrossRefLaurence, B. L., Weismer, S. W., Miller, C. A., Francis, D. J., et al. (2007). Speed ofprocessing, working memory, and language impairment in children. Journal of Speech, Language, and Hearing Research, 50(2), 408–428.CrossRefLi, D., Liu, C., Chen, T., & Li, G. (2004). The role of processing speed and workingmemory in life span cognitive development. Journal of Nanjing Normal University(Social Science), 1, 81–87.Madden, D. J. (2001). Speed and timing of behavioral processes. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the psychology of aging (5th ed.). San Diego, CA: Academic Press.Meijer, W. A., de Groot, R. H. M., Van Boxtel, M. P. J., Van Gerven, P. W. M., & Jolles, J. (2006). Verbal learning and aging: Combined effects of irrelevant speech, inerstimulus interval and education. Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 25(1), 364–376.Park, D. C. (2000). The basic mechanisms accounting for age related decline in cognitive function. In: Cognitive aging. New Jersey: Erlbaum.Park, D. C., Smith, A. D., Lautenschlager, G., Earles, J. L., Frieske, D., Zwahr, M., et al. (1996). Mediators of long-term memory performance across the life span. Psychology and Aging, 11, 621–637.PubMed CrossRefPeng, H., Shen, J., & Wang, D. (2004). The roles of working memory capacity and

processing speed in inductive reasoning aging. Psychological Science, 27, 536–539.Peng, H., Shen, J., & Wang, D. (2006). Interrelations of visual function, processing speed and working memory in cognitive aging. Chinese Journal of Gerontology, 26(1), 1–3.Salthouse, T. A. (1985). Speed of behavior and its implications for cognition. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the psychology of aging (2nd ed., pp. 400–426). New York: Van Nostrand Reinhold.Salthouse, T. A. (1990). Cognitive competence and expertise in aging. In J. E. Birren & K. W. Schaie (Eds.), Handbook of the psychology of aging (3rd ed., pp. 310–319). San Diego, CA: Academic Press.Salthouse, T. A. (1993). Speed mediation of adult age differences in cognition. Developmental Psychology, 29, 722–738.CrossRefSalthouse, T. A. (1995). Aging, inhibition, working memory, and speed. The Journals of Gerontology, 50, 297–312.CrossRefSalthouse, T. A. (1996). The processing-speed theory of adult age differences in cognition. Psychological Review, 103, 403–428.PubMed CrossRefSalthouse, T. A. (2001). Structural models of the relations between age and measures of cognitive functioning. Intelligence, 29, 93–115.CrossRefSalthouse, T. A. (2004). Localizing age-related individual differences in a hierarchical structure. Intelligence, 32, 541–561.CrossRefSalthouse, T. A., & Ferrer-Caja, E. (2003). What needs to be explained to account for age-related effects on multiple cognitive variables? Psychology and Aging, 18,91–110.PubMed CrossRefSalthouse, T. A., Fristoe, N., McGuthry, K. E., & Hambrick, D. Z. (1998). Relation of task switching to speed, age, and fluid intelligence. Psychology and Aging, 13(3), 445–461.PubMed CrossRefSchaie, K. W. (1989). Perceptual speed in adulthood: Cross-section and longitudinal studies. Psychology and Aging, 4, 443–453.PubMed CrossRefSchaie, K. W. (1994). Intellectual development in adulthood: The Seattle Longitudinal study. New York: Cambridge University Press.Shen, J. L., Wang, D., Peng, H. M., & Tang, D. (2003). The effects of mediators on the aging of primary mental abilities. Acta Psychological Sinica, 6, 802–809.Vance, D., Dawson, J., Wadley, V., Edwards, J., Roenker, D., Rizzo, M., et al. (2007). The accelerate study: The longitudinal effect of speed of processing training on cognitive performance of older adults. Rehabilitation Psychology, 52, 89–96.CrossRefVerhaeghen, P., & Salthouse, T. A. (1997). Meta-analyses of age-cognition relations in adulthood: Estimates of linear and nonlinear age effects and structural models. Psychological Bulletin, 122(3), 231–249.PubMed CrossRefZimprich, D. (2002). Cross-sectionally and longitudinally balanced effects of processing speed on intellectual abilities. Experimental Aging Research, 28, 231–251.PubMed CrossRefZimprich, D., & Martin, M. (2002). Can longitudinal changes in processing speed explain longitudinal age changes in fluid intelligence? Psychology and Aging, 17, 690–695.PubMed CrossRef

Over 8.5 million scientific documents at your fingertips

Our Content• Journals • Books • Book Series • Protocols • Reference Works

Other Sites• Springer.com • SpringerProtocols • SpringerMaterials

Help & Contacts• Contact Us • Feedback Community • Impressum

© Springer International Publishing AG, Part of Springer Science+Business Media Privacy Policy, Disclaimer, General Terms & Conditions Not logged in Pitesti University (2000454052) ANELIS Plus Consortium (3000146411) 1246 Open Romania (3000208166) 194.102.70.234

Object 1

FEEDBACK