Cognitive processes in children's reading and attention: the role of working memory, divided attention, and response inhibition.
It has also been argued that children with attention difficulties experience problems performing two tasks simultaneously, even if those tasks utilize distinct sense modalities. Theoretical models of attention based upon single capacity models (e.g. Broadbent, 1958; Posner & Rafal, 1987), which assume that dual tasks call upon general attention/executive resources, as well as multiple-resources theories (e.g. Baddeley, Della Sala, Gray, Papagno, & Spinier, 1997; Baddeley & Hitch, 1994; Della Sala & Logie, 2001; Neisser, 1967), which assume that resources are relatively specific to some cognitive/perceptual processes, both predict dual-task deficits in children with attention difficulties. While not all research studies have reported clear evidence of dual-task deficits in children with ADHD diagnoses, at least some empirical evidence supports the existence of deficit in this domain (e.g. Adams & Snowling, 2001; Dige & Wik, 2005; Karatekin, 2004; Li, Lin, Chang, & Hung, 2004).
Baddeley and Hitch's well-known working memory (WM) architecture has also been used more directly as a potentially useful framework for discussing individual differences in attention, including developmental disorders of attention. In most models of WM, the central executive is generally viewed as being akin to a central supervisory attention processor that operates to schedule competing action plans (e.g. Baddeley, 2004). A good deal of experimental work on normal adults supports the plausibility of this general assertion (e.g. Conway & Engle, 1994; Cowan, 1988, 1997; Kane & Engle, 2000). Evidence from typical adult functioning also suggests that the central executive system is related to the control of attention through habituation or inhibition processes (e.g. Conway & Engle, 1994; Cowan, 1988; Kane & Engel, 2000). Conway and Engle, for example, report that WM differences were only evident in probed recall tasks that lead to interference or response competition. Conway and Engle also note that the capacity to inhibit which serves to limit entry to WM is likely to be relevant to developmental WM and attention problems. Consistent with this general view, problems on cognitive tasks designed to assess Central executive functioning has also been reported among children with clinical attention deficits (e.g. Barnett et al., 2001; Kempton et al., 1999; Nigg, Blaskey, Huang-Pollock, & Rappley, 2002; Schachar, 1991). Some of these studies have also implicated visuospatial WM deficits in ADHD.
Recently, this developmental pattern has been somewhat complicated by evidence suggesting that all three of the cognitive deficits described above: namely, dual-task processing, response inhibition and WM deficits, may all be related to reading disability (RD) rather than, or as well as, attention problems. Purvis and Tannock (2000) contrasted children experiencing either ADHD or RD alone, or RD and ADHD with typical controls. They found that both ADHD + RD and RD samples of children showed deficits in speeded processing tasks but not on a continuous performance task used to measure response inhibition. Children with RD did show poorer performance than non-RD children on this measure. This pattern led Purvis and Tannock to conclude that inhibition tasks may not be unique cognitive markers for ADHD (see also Adams & Snowling, 2001; Dige & Wik, 2005; Karatekin, 2004; Li et al., 2004). Purvis and Tannock suggest that children with ADHD may have a more pervasive inhibitory deficit shown in behavioural impulsivity, whereas RD children may experience inhibition problems more specifically with rapidly presented material.
Some reading researchers have suggested that dual-task performance might also distinguish good and poor readers (e.g. Nicolson & Fawcett, 1990, 1995, 2000; Nicolson, Fawcett, & Dean, 2001; Yap & Van Der Leij, 1994). In these experimental studies, developmental dyslexics were asked to carry out two tasks such as balancing on a beam and counting backwards, or playing a computer game while also concurrently responding to on-line signals. Children completed each of these tasks first separately, and then simultaneously. Dual-task performance deficits have been reported on a range of well-learned tasks for developmental dyslexics compared with regular reading controls. As the tasks used in these studies do not require use of reading, spelling or phonological skills, this pattern is therefore consistent with a general dual-task deficit in developmental dyslexia. This work has itself been criticized on methodological grounds (Savage, 2004). Failures to replicate this dual-task deficit in children with dyslexia have also been reported (e.g. Wimmer, Mayringer, & Lander, 1998; Wimmer, Mayringer, & Raberger, 1999).
The specificity of associations between developmental attention difficulties and WM component systems such as the central executive and visuospatial processing has also been questioned. Swanson (1993), for example, suggests that problems at all three postulated levels of WM architecture characterize children with a wide range of developmental problems. Others have reported that there are no clear associations between attention problems and visuospatial processing (Kerns, McInerny, & Wilde, 2001; Sonuga-Barke, Dalen, Daley, & Remington, 2002). Similarly, the reliability of deficits in central executive processing in attention disorders has also been questioned by other research. Roodenrys et al. (2001) contrasted 16 children with ADHD plus RD and 16 children with pure RD and non-disabled controls. Both the ADHD/RD and the RD groups showed phonological loop deficits compared with controls. Deficits were also evident for both groups on tasks tapping central executive function. August and Garfinkel, (1990) and Halperin, Gittelman, Klein, and Rudel (1984) have failed to find differences between ADHD and co-occurring ADHD plus poor reader groups sampled from clinics on attention measures. Finally, Pennington et al. (1993) contrasted four groups of children drawn from schools, parent networks and clinics. Children with pure ADHD (N = 16) and reading problems (N = 15; hereafter called RD) were contrasted with a group with both ADHD and RD (N = 16) and controls (N = 23). The pure RD and ADHD + RD groups both showed phonological processing deficits. Pennington et al. also report that the co-morbid ADHD and RD group did not show executive function deficits compared with the pure ADHD group who did show deficits. Pennington et al. argue on the basis of this associational data that this pattern suggests that a primary reading problem might frequently lead to a secondary attention difficulty that reflects reading problems in conjunction with environmental risk, rather than a ccentral executive deficit. The broader environmental risk factors identified by Pennington et al. include lower maternal education, mother-only households and greater family psychiatric difficulty. Pennington et al. term this general notion the phenocopy hypothesis.
In summary then, existing studies provide at best mixed support for clear cognitive profiles of dual-task speeded response inhibition and WM deficits underpinning developmental attention problems. Furthermore, there is some evidence that a number of the purported deficits in developmental attention disorders may be most closely related to cognitive processes in reading acquisition (e.g. Nicolson & Fawcett, 1990, 2000; Pennington et al., 1993; Purvis & Tannock, 2000).
Caution is currently also necessary in interpreting all deficit patterns because of a number of important methodological limitations in many of the existing studies described above. Firstly, existing studies have often involved comparisons of relatively small cells of between 10 and 25 ADHD and poor-reading children. Arguably, more reliable patterns among WM and other theoretically important attention processing measures are likely to emerge only from somewhat larger studies.
Secondly, many studies have also been based exclusively on children with multiple clinical problems that are more likely to come to the attention of services. Given genetic evidence that, for example, the behavioural traits associated with ADHD may be distributed in a continuum across the population (Gillis, Gilger, Pennington, & DeFries 1992; Levy, Hay, McStephen, Wood, & Waldman 1997), the literature may benefit from the use of use of complimentary non-clinical populations as clinical populations may bring with them issues such as referral bias (Adams & Snowling, 2001; Pennington et al., 1993).
Thirdly, the very mixed patterns of association reported between variation in reading and attention on the one hand and cognitive processes on the other hand may well reflect problems with specific statistical and methodological models used. One reason for the mixed patterns of effects reported to date may be that multiple statistical contrasts are undertaken on related cognitive measures. As a result, findings may be unreliable. A further methodological issue in this literature is that cognitive constructs are operationalized in a range of ways across studies, potentially adding further to measurement error and unreliability. A superior statistical approach may be to explore a more limited number of latent variables derived from preliminary factor analyses of variables selected on principled grounds (WM tasks, dual-task attention tasks, response inhibition tasks). Such latent variables can be considered 'error-free' measures of variance (e.g. Wagner, Torgeson, & Rashotte, 1994) and thus provide stronger test of associations.
Finally, the strongest case for specificity may be made where there is control for systematic variation in IQ. There is considerable current debate in the normal adult literature concerning the independence of WM and IQ constructs (e.g. Ackerman, Beier, & Boyle, 2005; Kane, Hambrick, & Conway, 2005; Oberauer, Schulze, Wilhelm, & Suss, 2005) although there appears to be some agreement that they represent somewhat separate constructs (Beier & Ackerman, 2005). Thus, strong evidence of specificity of association between WM components and behaviourally defined attention problems would be provided if the association remained significant even after IQ was partialled out. Finally, the inclusion of additional controls for chronological age (CA) would also add to confidence in the specificity of the association between cognitive processes and behavioural attention ability.
The current study
In the current study, we set out to examine whether variation in reading or attention problems were each specifically associated with processing variability in WM and attention subtasks. To this end, we took advantage of a large-scale study that had collected teacher ratings of ADHD-type difficulties in a sample of 1,811 children aged between 6 and 11 years attending mainstream schools in central England (the details of this study are reported in Cornish et al., 2005). From this initial population, boys who were rated by their teachers as having a high (top 10%) or low (bottom 10%) frequency of ADHD characteristics were administered a range of attentional subtests from the Test of Everyday Attention for Children (TEA-Ch) battery (Manly, Robertson, Anderson, & Nimmo-Smith, 1999) alongside WM subtests from the Working Memory Test Battery for Children (WMTB-C) battery (Pickering & Gathercole, 2001) and measures of reading ability--the Neale analysis of reading ability second edition (NARA; Neale, 1997).
Sample ascertainment and selection
In the present study, we screened an epidemiological sample of 6- to 11-year-old children from central England (UK) using teacher ratings on the SWAN scale (Strengths and weaknesses of ADHD-symptoms and normal behaviour scale; Swanson, McSyephenen, Hay, & Levy, 2001; see below). Children were recruited to this study by contacting principals and teachers from all primary schools (elementary schools) in the county of Nottinghamshire, UK. Of the 402 schools listed, a total of 157 agreed to participate. Each school was then asked to complete the SWAN questionnaire for one class in a given year group (across year Groups 2 to 6, CA 6- to 11-year-olds). This yielded teacher questionnaires on 1,776 children. Ninety-two questionnaires were excluded from further analysis due to missing or incomplete responses. There were complete questionnaires on 872 boys and 812 girls. SWAN summary (total) scores were normally distributed: boys: mean = 4.7, SD = 23.1, skewness, s = -0.11, kurtosis, k = -0.30; girls: mean = -9.9, SD = 20.4, skewness, s = -0.24; kurtosis, k = -0.30. The aim of this study was to assess children at the extremes of a 'normal' ADHD continuum. Because the distribution of boys' scores SWAN was shifted to the right relative to that of girls (such that 86% of the top 10% SWAN ratings were for boys), only boys were included in the subsequent parts of this study.
Boys were eligible for inclusion in the second stage of the study if they scored either above the 90th percentile or below the 10th percentile on the sex-specific distribution of boy's scores on the inattentive or hyperactivity/impulsivity subscales of the SWAN. All schools participating in the second stage wrote to parents inviting their child to participate in the study. Consent was obtained for 126 boys (range: 6-11 years; mean 9;5). This resulted in 58 boys who were rated by teachers above the 90th percentile for inattentive and/or hyperactivity/impulsivity subscale items on the SWAN questionnaire (age range: 6-11 years; mean age: 8;6) and 68 boys who were rated by teachers as below the 10th percentile for inattentive and/or hyperactivity/impulsivity subscale items on the SWAN questionnaire (age range: 6-11 years; mean age: 9;5). None of the children were receiving stimulant medication (e.g. methylphenidate or dexamfetamine). Of the sample, 98% (N = 124) were Caucasian (indigenous white British). One child was African Caribbean, and one child was Cypriot (Mediterranean Caucasian).
The SWAN ADHD Scale (Swanson et al., 2001)
The SWAN scale is based on the 18 ADHD symptoms listed in the DSM-IV manual. Scoring for each item goes from a low level of problems (3, 2, 1) through average (0) to high level (-1, -2, - 3, - 4). Children's scores ranged from a minimum of -27 to a maximum of 27 for each subscale. This scale allows for a normal distribution of the data and avoids the potential psychometric flaws that are associated with skewed distributions. Moreover, this scale allows identification of extremes at opposite ends of the distribution.
All participants were tested individually on the Wechsler abbreviated scale of intelligence (WASI; Wechsler, 1999). This test provides a composite IQ score based on four subtests tapping both verbal and performance domains. Table 1 shows a summary of mean CA and IQ across the SWAN ADHD score groups.
Attention and working memory tasks
A battery of recently published standardized neuropsychological tasks specifically designed for use with children was selected to assess the different cognitive aspects of dual attention, response inhibition and WM (phonological, visual-spatial and central executive memory): The TEA-Ch (Manly et al., 1999) and the WMTB-C (Pickering & Gathercole, 2001).
Dual attention (TEA-Ch)--Sky Search Task
(1) The sky search dual task required participants to perform two tasks simultaneously. In one task, participants kept count of the number of sounds presented intermittently throughout the trial on a tape. The participant was first reminded about the scoring sounds using a practice trial. In the sky search task, a participant is asked to find as many target 'spaceships' as possible on a sheet also containing distractor 'spaceships'. The sky search sheet was then placed in front of the participant and both tasks began when a voice on the tape said '54321 ... start'. Timing also began on 'start' and stopped when the participant ticked the finish box. The sky search dual-task score was calculated by taking the proportion of counting games correct (games correct divided by games attempted) and dividing this by the time per target, to give a weighted-time-per-target score. The time per target score from the first sky search task was then subtracted from the weighted-time-per-target score to give the sky search dual-task score.
Dual attention (TEA-Ch)--Score Dual Task
(2) The score dual task also required participants to perform two tasks simultaneously. In one task, participants kept count of the number of sounds presented on a tape. In the second task, they listened to a news report in order to identify an animal name mentioned in a report. Participants were given two practice trials before the main task. There was a total of 10 trials. After each trial, the participant was asked, 'What was the animal and how many sounds did you count?' Participants scored 1 point per trial for each correct animal identified and 1 point per trial for each correct number of scoring sounds.
Response Inhibition (TEA-Ch)--Same World-Opposite World Task
Impairments in the capacity to inhibit prepotent responses have been argued to be central to ADHD (Barkley, 1999). The opposite world task from the TEA-Ch was derived from similar measures of verbal inhibition described by Gerstadt, Hong, and Diamond (1994) and Passler, Isaac, and Hynd (1985). In each of the four trials of the test, children were shown a stimulus array of 24 boxes, each containing either the digit 1 or 2. Following practice, in the first part of the test (same world), the examiner pointed to each digit in the array in turn whilst the child named that digit aloud. This condition is used to lend the verbal responses to the digits some prepotency within this particular test context. In the second part, opposite world, the child was asked to say the opposite for each digit (i.e. 'one' for 2 and 'two' for 1). The remaining two sections of the test repeat the same- and opposite world conditions. Each trial was timed. The standardized instructions are for the examiner to move on to the next digit in the arrays only when the correct response has been given. Errors therefore contribute to the overall time score. The final measure is the total time taken to complete the two opposite world trials. The sensitivity of this measure to ADHD is reported in Manly, Anderson, Nimmo-Smith, Turner, Watson, and Robertson (2001) and is correlated with ratings of ADHD (Adams & Snowling, 2001).
Working memory measures
WM measures were taken from the WMTB-C (Pickering & Gathercole). The test battery is based upon an extensive literature concerning the triarchic structure of WM that consists of a central executive and devoted visual and verbal 'slave' systems (Baddeley, 1986; Gathercole & Baddeley, 1993). The test itself has high construct validity for the central executive and phonological loop measures and external validity as a predictor of vocabulary, literacy and arithmetic tests (e.g. Gathercole & Pickering, 2000).
Phonological memory tasks (WMTB-C)
(1) The digit recall task required participants to recall a series of digits. Participants first completed a practice trial where they were asked to recall one-, two- and three-digit sequences. Digit sequences were read out by the experimenter at a rate of about one digit per second. The main task began at the greatest span at which the participant was successful in the practice trial. A maximum of six trials per span were presented and participants had to complete four of six correct trials in order to continue on to a higher span. The maximum span tested was nine. Total number of correct trials was recorded.
(2) The word list matching task required participants to judge whether a word list was presented in the same or different order to the original presentation, for example, 'dog plug' is presented and followed by 'plug dog'. Participants first completed a practice trial where word lists containing two, three and four words were read out. For the main task, details of start point, presentation rate and progression on to next span are as described for the digit span task. The maximum span tested was eight. The number of correct trials was recorded.
(3) The word list recall task was administered in exactly the same way as digit recall, except words rather than digits were used. The maximum span tested was seven.
(4) The non-word list recall task was administered in exactly the same way as digit recall, except pseudo-words (e.g. meek) rather than digits were used. The maximum span tested was six.
Visual spatial memory tasks (WMTB-C)
(1) The visual patterns test required participants to reproduce a checkerboard pattern that was presented to them for 3 seconds, on to a blank grid of the same size and shape. The patterns consist of a grid with half the squares filled in that progress in size from the smallest 2 x 2 matrix (with 2 filled squares) to the largest, a 5 x 6 matrix (with 15 squares filled). Three different patterns were presented at each level of complexity and testing continued until the participant failed to recall correctly any of the three patterns on the same level. Participants were first given a practice trial before the main task. There were no time restrictions and feedback was given after each trial. Visual span was recorded from the level of complexity of the largest grid with at least one of the three patterns correctly recalled.
(2) The Corsi blocks task required participants to copy the exact sequence of block 'tapping' demonstrated by the experimenter. A board with nine blocks attached was used. Participants first completed a practice trial in which they copied the experimenter by tapping two blocks in the same order and then three blocks. For the main task, details of start point, presentation rate and progression on to next span are as described for the digit span task. The maximum span tested was nine. The number of correct trials was recorded.
(3) The mazes task required participants to reproduce a maze solution that was presented to them on to an empty maze of the same shape and size. Participants first completed four practice trials of varying complexity (Span 2 [i.e. a maze with two walls] to Span 4 (i.e. maze with four walls). For the main task, details of start point and progression on to next span are as described for the digit span task. The maze was presented for the length of time it took the participant to trace the route with their finger. The maximum span tested was seven. The number of correct trials was recorded.
Central executive memory tasks (WMTB-C)
(1) The listening span task required participants to listen to a series of sentences, judge whether they are true or false and recall the final word from each sentence, for example, 'fishes have long hair' (correct response is 'false, hair'). Participants first completed a practice trial where they were asked to judge/recall one- and two-sentence sequences. The main task began with one-sentence sequences for all participants. Details of progression on to the next span are as described for the digit span task. The maximum span tested was six. The number of correct trials was recorded.
(2) The counting recall task requires participants to count the number of dots per card on a series of cards and recall the number of dots per card. Participants first completed a practice trial where they were asked to recall one-, two- and three-card sequences. Each card is presented at a rate of about one card per 5 seconds. For the main task, details of start point and progression on to next span are as described for the digit span task. The maximum span tested was seven. The number of correct trials was recorded.
(3) The backwards digit recall task required participants to recall a series of digits in the reverse order to which they were presented, for example, '2 8 5' would be recalled as '5 8 2'. The task was administered in exactly the same way as the digit recall task.
The NARA (Neale 1997) was used to measure reading. This test has been used extensively to assess reading rate, accuracy and comprehension (e.g. Nation & Snowling, 1997; Savage & Frederickson, 2005; Stothard & Hulme, 1992; Stuart, 2004). Children are asked to read aloud a series of graded fiction and non-fiction narratives as speedily and as accurately as possible. Children are also told in advance that they will be asked questions about the narratives they have read. The test provides an age-standardized score measure of reading accuracy (calculated by the number of reading errors to a discontinuation point of 14 errors in a single narrative), rate (based on words read per minute in each narrative) and comprehension (based on the accuracy of subsequent responses to questions about the passage).
The NARA reading comprehension scale, but not the reading accuracy scale, is known to be strongly associated with measures of listening comprehension, and this relationship remains even after variation in word reading accuracy is first controlled (Nation & Snowling, 1997; Savage, 2001), showing that it is a valid test of comprehension as well as word recognition. It also produces a measure of reading rate that is specifically related to other fluency measures such as rapid naming and which loads alongside rapid naming as a distinct factor separate from reading accuracy (Savage & Frederickson 2005), attesting to the separable components of reading accessed by the test. Form A was selected as Form B may show a gender skew (Stothard & Hulme, 1992).
Following regional and local ethical approval, all participants were tested in the school environment, and testing followed the same standardized routine on each occasion. Feedback was not given on individual performance. Teachers and parents completed the Conner's questionnaire and they were returned via post to the researcher.
Standardized scores were used for all analyses. Cluster standard scores for the three components of the WM architecture (central executive, visuospatial scratchpad, phonological loop) were calculated as directed in the WMTB-C manual (Gathercole & Pickering, 2000). At this stage, three boys were excluded from the final analysis, as they were unavailable for the reading assessments. Data distributions were screened for normality and found to be adequate for skew (s < 1), and kurtosis (k = ns). Means and standard deviations of all cognitive measures used in the study are presented in Table 2.
The aims of this study were to explore the relationship between error-free latent variables representing attention and WM processes and to examine the specificity of the association of these latent variables with reading ability and attention problems. Statistical analyses are therefore presented in three parts. Simple associations between measured variables were first explored. These are presented in the first section of the results. These data were subjected to factor analysis to create latent variables. These analyses are presented in the second section of the results. The final aim of the paper was to explore the specificity of association between measured reading and attention. Latent variables were used as independent variables in hierarchical regression analyses that first control for IQ and attention (when reading ability is the dependent variable) or reading (when attention is the dependent variable), respectively. The third section of the results depicts these analyses.
Prior to the main analyses, preliminary analyses were first run to explore the simple associations between the complete set of variables of interest. The first phase of these analyses explored variables associated with attention. As this variable is dichotomous, univariate ANOVA was used to explore patterns. ANOVA revealed that the high- and low-attention group children differed on all measures of reading (reading accuracy, F(1, 122) = 112.01, p < .001; reading rate, F(1, 122) = 78.55, p < .001); reading comprehension, F(1, 122) = 109.10, p < .001). For the sake of economy of space, data from combined measures of the three WM measures only are presented. The groups differed on WM subprocesses: phonological loop, F(1, 122) = 112.01, p < .001; visuospatial scratchpad, F(1, 122) = 78.55, p < .001; central executive, F(1, 122) = 109.10, p < .001, with poorer performance characterizing the poor attention group in all cases. Unsurprisingly, children with good and poor attention differed on all attention measures: Score dual task F(1, 122) = 77.94, p < .001); sky search dual-task measure F(1, 122) = 19.39,p < .001; response inhibition: same world task F(1, 122) = 31.66, p < .001); inhibition: opposite world task F(1, 122) = 30.57, p < .001). After adjustment for IQ, significant differences between high- and low-attention children on nearly all measures was still evident. The one exception to this pattern was the phonological loop measure of WM, that was no longer significant F(1, 122) = 2.75 ns.
The interrelation between all other non-dichotomous variables was explored through Pearson product moment correlation coefficients. The correlations are reported below in Table 3. Only composite scores for the WM measures are reported here to aid accessibility of tables. These reveal that reading accuracy, rate and comprehension are, unsurprisingly, strongly associated with each other. Reading accuracy, rate and comprehension was also strongly correlated with all three subcomponents of the WM test battery, as well as with IQ. Of the attention task measures, dual-task performance and the inhibition measures are generally the strongest correlates of reading accuracy, rate and comprehension, although effects are reduced for the speeded dual task over the accuracy-based dual task. Modest but significant patterns of association are evident for the WM components in relation to the inhibition and dual-task measures.
Partial correlations between these variables after adjustment for IQ are depicted in the lower corner of Table 3. These reveal a somewhat similar pattern. Reading rate accuracy and comprehension are generally strongly correlated with dual-task performance and the inhibition measures. All three reading tasks are associated with central executive processing in WM and of the three reading tasks, reading accuracy is most strongly associated with phonological loop processing. Visuospatial scratchpad measures no longer correlate with reading performance. Finally, both the phonological loop and visuospatial measures correlate with central executive functioning but not with each other, as predicted by the WM architecture proposed by Baddeley (1986).
Factor analyses of candidate cognitive variables
Data reduction was undertaken to identify the underlying variance in the large body of WM response inhibition and attention tasks. Participants' scores on the seven cognitive processing variables that our review had identified as being related to variation in literacy and attention in previous research were therefore subjected to principal components analysis (PCA). Initial assessment of the suitability of the data for factor analysis was carried out. There are a large number of individual variables. However, as the sample size was fairly large, the ratio of cases to variables of 9:1 was adequate (Tabachnick & Fidell, 2001). Inspection of the correlation matrix has already identified several correlations above .3. The Kaiser-Meyer-Olkin value at .87 exceeded the recommended value of .6, and the Bartlett's Test of Sphericity was statistically significant, supporting the factorability of the correlation matrix.
PCA revealed the presence of three components with eigenvalues exceeding 1, explaining 43.89%, 11.08% and 7.93% of the variance, respectively. When oblique rotation was requested, transformation matrix correlations were found to be substantially reduced over orthogonal rotation. Because of this, orthogonal rotation was selected and is reported in Table 4, where loadings of variables on factors, communalities and percentages of variance are shown. To aid interpretability, only loadings higher than .4 are shown. It can be seen from Table 4 that tasks purportedly measuring verbal memory, visual-spatial memory and central executive functions indeed emerged as separate factors distinct from each other. Attention measures from the TEACh loaded in the main with the WM battery central executive measures, and not at all with the phonological memory tasks. The score dual task but not the sky search task loaded strongly with the visual-spatial processing measures of the WM test battery. We therefore chose to label these latent variable factors after the familiar WM model components they represent but labelled Component 3 dual task/visuospatial WM.
The reading subscale scores were all highly correlated and loaded as a single factor in factor analyses. A latent variable of reading ability was created and used in all subsequent analyses involving reading ability.
Regression analyses predicting attention and reading from candidate cognitive variables
The primary aim of this paper was to explore the associations between latent variables of WM, dual-task performance, response inhibition on the one hand and reading and attention on the other. These relationships were explored using hierarchical linear regression (HLM) models. As previous analyses had revealed that good and poor attention groups also differed on both age and IQ, these variables were always entered as extraneous variables in the first two steps of analyses. For the prediction of normally distributed reading dependent variables (DVs), standard hierarchical multiple regression analyses were used, as these require a normally distributed DV, predicted from either normally distributed or dichotomous IVs. As the attention measure was dichotomous, binomial logistic regression was used in these cases. This approach can be utilized when dichotomous and normally distributed IVs are used to predict a bivariate dichotomous DV. Data were carefully screened for evidence of co-linearity. Conservative but conventional tolerance measures were used to evaluate co-linearity (Tabachnik & Fidell, 2001). No evidence of significant co-linearity was found.
Predictors of attention group
The first set of analyses explored the specific predictive validity of the three latent variables identified from the PCA analysis of WM, dual-task and response inhibition above as predictors of attention group membership. The aim of these analyses was to explore whether there was a unique association between cognitive measures and behavioural categorization of attention even after controls for possible extraneous variables, including literacy. In these analyses, CA and IQ were entered first, as high- and low-attention groups differed on these measures. Reading ability was entered at Step 3 prior to entering central executive and visuospatial scratchpad latent variable measures at Stage 4 of the regression analyses. The results are depicted in Table 5. These analyses showed that the dual-task/visual-spatial latent variable, but not the central executive latent variable, was a significant predictor of attention-group membership at Step 4 of the analyses, even after controls for reading ability. These analyses also confirmed that the verbal WM factor did not predict attention group membership.
Predictors of reading ability
The second phase of these analyses sought to address the predictors of reading ability from the latent variables after parallel control for extraneous factors (age, IQ) and attention group membership, in order to establish the nature of relationships between cognitive latent variables and reading ability. Analyses here explored the specific predictive validity of the three latent variables of WM and TEA-Ch tasks as predictors of reading ability. A series of parallel regressions were run to those reported above but where attention now acted as a final independent variable and reading was the dependent variable. As before, CA and IQ were entered as extraneous variables at Steps 1 and 2 of regression analyses. Attention group membership was entered at Step 3 of the analysis after CA and IQ. Finally, reading ability was entered at Step 4. These analyses are depicted in Table 6. These analyses reveal that after controls for age, IQ and attention group membership that the verbal WM latent variable was a significant unique predictor of reading. After additional controls for attention group membership at Step 3 of the analyses, there was a significant association between the central executive latent variable and reading ability. Analyses also revealed that after controls for age, IQ and attention group membership, there was no association between reading ability and the dual-task/visual memory latent variable.
The primary aim of this study was to explore the relationship between error-free latent variables that represent clusters of cognitive attention and WM processes and their association with behavioural reading or attention problems. In order to explore whether specific cognitive-behavioural associations were evident, regression analyses were run that sought to predict reading and attention after first controlling for extraneous variables (CA, IQ), as well reading ability (attention analysis) and reading (attention analysis).
Previous research suggested that of the three WM structures (visuospatial scratch pad, verbal WM and central executive) only the verbal WM measures will predict reading ability after extraneous variables such as IQ and proximal factors such as attention are controlled because it measures a specific deficit in phonological processing (e.g. Morton & Frith, 1995). Some previous research also suggested that visual-spatial, attention tasks and central executive measures of WM should similarly be specifically associated with variation in attention after controls for IQ, and for variation in reading ability (e.g. Barnett et al., 2001; Castellanos & Tannock, 2002; Kempton et al., 1999).
Preliminary factor analyses indicated that three latent variables can be extracted from the variability among individual selected TEA-Ch tasks and the WM subtasks. These three latent factors corresponded reasonably closely to the global WM structure proposed by Baddeley (e.g. Baddeley, 1986; Gathercole & Baddeley, 1993) wherein a central executive is served by visual and verbal 'slave' systems. However, one purported measure of central executive processing--listening span also loaded with visual-spatial processing. Furthermore, listening span, counting span and backwards digits loaded with verbal memory tasks, suggesting that 'central' processing factors loaded with visual and verbal 'slave' systems. The present data thus provide only modest support for the existence of an independent central executive in the WM architecture. The response inhibition and dual tasks loaded strongly alongside the majority of WM central executive measures. The score dual task also loaded significantly with the visual-spatial tasks of WM. The verbal memory measures were independent of all other measures in the factor analysis.
The results of regression analyses that controlled extraneous factors such as age, IQ and attention difficulty showed that reading ability was uniquely predicted by the verbal WM measure. The 'dual-task/visual memory' latent variable was not a unique predictor of outcome in reading. In contrast, this construct was a significant predictor of attention group membership after parallel controls for age, IQ and reading ability.
The evidence here thus suggests some specificity of association between distinct cognitive processes and developmental behavioural abilities, with reading ability being closely associated with core phonological processing tasks and attention difficulties closely associated with visual-spatial components of WM (Barnett et al., 2001; Castellanos & Tannock, 2002; Kempton et al., 1999; Morton & Frith, 1995). The present results also support the utility of using the WM architecture to identify specific associations in clinic and school samples (Adams & Snowling, 2001; Roodenrys et al, 2001). Finally, the data reported here using an individual differences approach to exploring reading and attention skills in the classroom are also complimentary to the factorial studies reported by Pennington et al. (1993).
The 'central executive' latent variable measure was a strong unique predictor of reading rather than attention group membership in the present analyses after controls for extraneous variables. This latent variable included WM central executive and TEA-Ch behavioural inhibition measures. How might such results be understood? Behavioural inhibition and central executive processing have both often been seen as key components--even candidate endophenotypes for a shared genetic basis of attention problems (Castellanos & Tannock, 2002; Cornish et al., 2005). The response inhibition tasks used here, like most of those used in the literature to date, require speeded response. It may be that these sorts of inhibition task are less closely associated with attention difficulties than response inhibition tasks requiring high levels of accuracy (e.g. Wilding, 2003).
Perhaps the most direct reason for the close association between these variables and reading might be that the present reading measure was a broader and more comprehensive measure of literacy skills than that used in many studies of ADHD and reading. The NARA reading test required children to integrate simultaneous reading rate, accuracy and comprehension demands. The 'reading' latent variable reflects this combination of skills in fluency, and text-level analysis. It is perhaps unsurprising that this measure loads most closely with central processing tasks in the WM battery, and with response inhibition tasks. Indeed, the link between text comprehension and the central executive WM is a well-replicated finding in the literature (e.g. Leather & Henry, 1994; Gathercole & Baddeley, 1993; Stothard & Hulme, 1992). Evidence of such a link from 'error-free' latent variable analysis is, however, new to this literature, as is the finding that after controls for reading, there is no strong association between attention difficulties and central executive tasks. Nevertheless, a role for central processes is suggested in attention problems, as tasks that are often assumed to require 'central processing' such as listening span also loaded with the visuospatial memory/dual-task factor that uniquely predicted attention group membership.
In the present study, response inhibition tasks loaded with central executive measures. This combined factor predicted reading but not attention group membership. This finding may also be consistent with recent findings by Tannock and colleagues. Rucklidge and Tannock (2002) have reported from clinic-based samples of ADHD children and RD children that rapid naming deficits were evident for both ADHD and RD samples. Their regression analyses indicated that response inhibition measures were poor predictors of ADHD across the sample. Our findings from a school-based sample are therefore consistent with those reported by both Rucklidge and Tannock and Purvis and Tannock (2000) in finding a close link between reading and speeded inhibition measures, but a less strong link between inhibition measures and attention problems.
It is also possible that normal variation in reading, and particularly in more complex reading tasks required of the age-appropriate reader, such as the integration of accuracy and text comprehension with reading fluency may be the primary source of at least some inhibition effects in ADHD. Clearly, at this point, such a link is rather speculative. Potentially, however, our reported link between reading problems and certain inhibition tasks might provide one specific cognitive mechanism underpinning the 'phenocopy hypothesis' advanced by Pennington et al. (1993) to explain the apparently secondary nature of attention problems in a significant number of children with co-occurring ADHD and RD diagnoses. The nature and implications of this finding are therefore potentially important and should be the focus of future study.
In summary, the results of the present latent variable regression-based approach suggests that there are indeed reasonably specific domains of deficit in attention and poor reading, with visual WM and dual-task processing problems specifically associated with attention difficulties, and with phonological loop difficulties closely and specifically associated with variation in reading. This sort of finding of specificity cannot be the result of measurement error, as latent variable analyses are 'error-free'. These results using a robust statistical and sampling design should also be readily replicated. There was also some evidence of shared processing as tasks such as 'listening span' from the WM test battery loaded on all factors in analysis that predicted both reading and attention.
The finding of specificity of cognitive processing patterns in attention tasks and reading is important for several reasons: as well as clarifying the cognitive basis of reading and attention problems, ultimately, the clearer the designs of studies providing behavioural evidence of specificity, the greater the chance of identifying separable genetic components of developmental disorders (e.g. Stanovich, 1993). It seems likely that this will also hold true for attempts to identify the neurological bases of reading and attention problems. A final important implication of this work is practical in nature.
Practical implications of these findings
The first practical implication of the present results is that they inform assessment for intervention. Results validate both the WM test battery and the selected tasks from the TEA-Ch, and their potential utility in separating cognitive function variability associated with reading and children with significant teacher-rated attention problems. Such measures can thus be used for assessment and specifically guided interventions. Of the TEA-Ch measures, it would appear that the score dual-task components are the reliable unique predictors of high and low attention, but may also be influenced by variation in reading. Caution in using these measures alone in diagnosis and assessment is therefore probably warranted. Indeed, more generally, it seems wise on the basis of present findings to routinely consider the contribution of reading ability (rate, accuracy and comprehension) when assessing possible attention problems in school-age children.
More generally still, we would argue that the separability of cognitive function here also has implications for clinical intervention and wider educational planning. ADHD and reading are not, it seems, behaviourally homogenous. Separate cognitive patterns of strength and weakness are quite evident, once appropriate controls for common overlapping incidence are taken. It is therefore likely that very different sorts of intervention are needed in each case to remediate and support distinct and separable underlying cognitive deficits. For children experiencing both primary reading and primary attention problems, both difficulties will have to be supported to ensure educational success.
Ackerman, P. L., & Boyle, M. E., & Boyle, M. O. (2005). Working memory and intelligence--the same or different constructs? Psychological Bulletin, 131, 30-60.
Adams, J. W., & Snowling, M. J. (2001). Executive function and reading impairments in children reported by their teachers as hyperactive. British Journal of Developmental Psychology, 19, 293-306.
August, G., & Garfinkel, B. D. (1990). Comorbidity of ADHD and reading disability among clinic referred children. Journal of Abnormal Child Psychology, 18, 29-45.
Baddeley, A. D. (1986). Modularity, mass-action and memory. Quarterly Journal of Experimental Psychology: A. Human Experimental Psychology, 38, 527-533.
Baddeley, A. D. (2004). Working memory: Looking forward and back. Nature Reviews, Neuroscience, 4, 829-839.
Baddeley, A. D., Della Sala, S., Gray, C., Papagno, C., & Spinier, H. (1997). Testing central executive functioning using a paper and pencil test. In P. M. A. Rabbitt (Ed.), Methodology of frontal and executive function (pp. 61-80). Hove: Psychology Press.
Baddeley, A. D., & Hitch, G. J. (1994). Developments in the concept of working memory. Neuropsychology, 8, 485-493.
Barkley, R. A. (1997). Behavioral inhibition, sustained attention, and executive functions: Constructing a unifying theory of ADHD. Psychological Bulletin, 121, 65-94.
Barkley, R. A. (1999). Response inhibition in attention-deficit hyperactivity disorder. Mental retardation and developmental disabilities research reviews, 5, 177-184.
Barnett, R., Maruff, P., Vance, A., Luk, E. S., Costin, J., Wood, C., & Pantelis, C. (2001). Abnormal executive function in attention deficit hyperactivity disorder: The effect of stimulant medication and age on spatial working memory. Psychological Medicine, 31, 1107-1115.
Beier, M. E., & Ackerman, P. L. (2005). Working memory and intelligence: Different constructs. Reply to Oberauer et al. (2005) and Kane et al. (2005). Psychological Bulletin, 131, 72-75.
Broadbent, D. E. (1958). Perception and communication. London: Pergamon Press.
Castellanos, F. X., & Tannock, R. (2002). Neuroscience of attention-deficit/hyperactivity disorder: The search for endophenotypes. Nature Reviews: Neuroscience, 3, 617-628.
Conway, A. R. A., & Engle, R. W. (1994). Working memory and retrieval: A resource-dependent inhibition model. Journal of Experimental Psychology: General, 123, 354-373.
Cornish, K. M., Manly, T., Savage., R., Swanson, J., Grant, C., Morisano, D., Butler, D., Cross, G., Sutherland, L., & Hollis, C. P. (2005). Association of the dopamine transporter (DAT1) 10/10-repeat genotype with ADHD-symptoms and response inhibition in a general population sample. Molecular Psychiatry, 10, 686-698.
Cowan, N. (1988). Evolving conceptions of memory storage, selective attention, and their mutual constraints within the human information processing system. Psychological Bulletin, 104, 163-191.
Cowan, N. (1997). Attention and memory: An integrated framework. New York: Oxford University Press.
Della Sala, S., & Logie, R. H. (2001). Theoretical and practical implications of dual task performance in Alzheimer's Disease. Brain, 124, 1479-1481.
Dige, N., & Wik, G. (2005). Adult attention deficit hyperactivity disorder identified by neuropsychological testing. International Journal of Neuroscience, 115, 169-183.
Gathercole, S., & Baddeley, A. D. (1993). Working memory and language. Hillsdale, NJ: Erlbaum.
Gathercole, S. E., & Pickering, S. J. (2000). Working memory deficits in children with low achievements in the national curriculum at 7 years of age. British Journal of Educational Psychology, 70, 177-194.
Gerstadt, C. L., Hong, Y. J., & Diamond, A. (1994). The relationship between cognition and action: Performance of children 2-7 years old on a Stroop-like day-night test. Cognition, 53, 129-153.
Gillis, J. J., Gilger, J. W., Pennington, B. F., & DeFries, J. C. (1992). Attention deficit disorder in reading-disabled twins: Evidence for a genetic etiology. Journal of Abnormal Child Psychology, 20, 303-315.
Halperin, J., Gittelman, R., Klein, D., & Rudel, R. (1984). Reading-disabled hyperactive children: A distinct sub-group of attention deficit disorder with hyperactivity. Journal of Abnormal Child Psychology, 12, 1-14.
Kane, M. J., & Engle, R. W (2000). Working memory capacity, pro-active interference, and divided attention: Limits on long-term memory retrieval. Journal of Experimental Psychology: Learning, Memory and Cognition, 26, 336-358.
Kane, M. J., Hambrick, D. Z., & Conway, A. R. A. (2005). Working memory capacity and fluid intelligence are strongly related constructs: Comments on Ackerman, Beier & Boyle (2005). Psychological Bulletin, 131, 66-71.
Karatekin, C. (2004). A test of the integrity of the components of Baddeley's model of working memory in attention-deficit/hyperactivity disorder (ADHD). Journal of Child Psychology and Psychiatry, 45, 912-926.
Kempton, S., Vance, A., Maruff, P., Luk, E., Costin, J., & Pantelis, C. (1999). Executive function and attention deficit hyperactivity disorder: Stimulant medication and better executive function performance in children. Psychological Medicine, 29, 527-538.
Kerns, K. A., McInerney, R. J., & Wilde, N. J. (2001). Time reproduction, working memory, and behavioral inhibition in children with ADHD. Child Neuropsychology, 7, 21-31.
Leather, C. V., & Henry, L. A. (1994). Working memory span and phonological awareness tasks as predictors of early reading ability. Journal of Experimental Child Psychology, 58, 88-111.
Levy, F., Hay, D. A., McStephen, M., Wood, C., & Waldman, I. (1997). Attention-deficit hyperactivity disorder: A category or a continuum? Genetic analysis of a large-scale twin study. Journal of the American Academy of Child Adolescent Psychiatry, 36, 737-744.
Li, C. S. R., Lin, W. H., Chang, H. L., & Hung, Y. W. (2004). Psychophysical measures of attention deficit in children with attention-deficit/hyperactivity disorder. Journal of Abnormal Psychology, 113, 228-236.
Manly, T., Anderson, V., Nimmo-Smith, I., Turner, A., Watson, P. C., & Robertson, I. H. (2001). The differential assessment of children's attention: The Test of Everyday Attention for Children (TEA-Ch), normative sample and ADHD performance. Journal of Child Psychology and Psychiatry, 42, 1065-1081.
Manly, T., Robertson, I. H., Anderson, V., & Nimmo-Smith, I. (1999). Test of Everyday Attention for Children (TEA-Ch). Bury St Edmunds: Thames Valley Test.
Morton, J., & Frith, U. (1995). Causal modelling: A structural approach to developmental psychopathology. In D. Cicchetti & D. J. Cohen (Eds.), Manual of developmental psychopathology (Vol. 1, pp. 357-390). New York: Wiley.
Nation, K., & Snowling, M. J. (1997). Assessing reading difficulties: The validity and utility of current measures of reading skill. British Journal of Educational Psychology, 67, 359-370.
Neale, M. D. (1997). Neale Analysis of Reading Ability (Second revised British edition). Windsor: NFER-Nelson Education.
Neisser, U. (1976). Cognition and reality: Principles and implications of cognitive psychology. San Francisco: Freeman.
Nicolson, R. I., & Fawcett, A. J. (1990). Automaticity: A new framework for dyslexia research. Cognition, 35, 159-182.
Nicolson, R. I., & Fawcett, A. J. (1995). Dyslexia is more than a phonological disability. Dyslexia: An International Journal of Research and Practice, 1, 19-37.
Nicolson, R. I., & Fawcett, A. J. (2000). Long term learning in dyslexic children. European Journal of Cognitive Psychology, 12, 357-393.
Nicolson, R. I., Fawcett, A. J., & Dean, P. (2001). Developmental dyslexia: The cerebellar deficit hypothesis. Trends in Neurosciences., 24, 506-514.
Nigg, J. T., Blaskey, L. G., Huang-Pollock, C. L., & Rappley, M. D. (2002). Neuropsychological executive functions and DSM-IV ADHD subtypes. Journal of the American Academy of Child and Adolescent Psychiatry, 41, 59-66.
Nigg, J. T., Hinshaw, S. P., Carte, E. T., & Treuting, J. L. (1998). Neuropsychological correlates of childhood attention-deficit/hyperactiviy disorder: Explainable by comorbid disruptive behavior or reading problems? Journal of Abnormal Psychology, 107, 468-480.
Oberauer, K., Schulze, R., Wilhelm, O., & Suss, H. M. (2005). Working memory and intelligence--their correlation and their relation: Comments on Ackerman, Beier, and Boyle (2005). Psychological Bulletin, 131, 61-65.
Passler, M. A., Isaac, W., & Hynd, G. W. (1985). Neuropsychological development of behaviour attributed to frontal lobe functioning in children. Developmental Neuropsychology, 1, 349-370.
Pennington, B. F., Groisser, D., & Welch, M. C. (1993). Contrasting cognitive deficits in attention deficit hyperactivity disorder versus reading disability. Developmental Psychology, 29, 511-523.
Pickering, S. J., & Gathercole, S. E. (2001). The Working Memory Test Battery for Children. London: The Psychological Corporation.
Posner, M. I., & Rafal, R. D. (1987). Cognitive theories of attention and the rehabilitation of attentional deficits. In M. J. Meier, A. L. Benton & L. Diller (Eds.), Neuropsychological rehabilitation (pp. 182-201). Edinburgh: Churchill Livingstone.
Purvis, K. L., & Tannock, R. (2000). Phonological processing, not inhibition, differentiates ADHD and reading disability. Journal of the American Academy of Child and Adolescent Psychiatry, 39, 485-494.
Roodenrys, S., Koloski, N., & Grainger, J. (2001). Working memory function in attention deficit hyperactivity disordered and reading disabled children. British Journal of Developmental Psychology, 19, 325-337.
Rucklidge, J. J., & Tannock, R. (2002). Neuropsychological profiles of adolescents with ADHD: Effects of reading difficulties and gender. Journal of Child Psychology and Psychiatry, 43, 988-1003.
Savage, R. S. (2001). The simple view of reading: Some evidence and possible implications. Educational Psychology in Practice, 17(1), 17-33.
Savage, R. S. (2004). Motor skills, automaticity and dyslexia: A Review of the research literature. Reading and Writing: An Interdisciplinary Journal, 17, 301-324.
Savage, R. S., & Frederickson, N. (2005). Further evidence of a link between rapid naming and text reading rate. Brain and Language, 93, 152-159.
Schachar, R. (1991). Childhood hyperactivity. Journal of Child Psychology and Psychiatry, 32, 155-191.
Sonuga-Barke, E. J., Dalen, L., Daley, D., & Remington, B. (2002). Are planning, working memory, and inhibition associated with individual differences in pre-school ADHD symptoms? Developmental Neuropsychology, 21, 255-272.
Sonuga-Barke, E. J. S., Taylor, E., & Hepinstall, E. (1992). Hyperactivity and delay aversion II. The effect of self versus externally imposed stimulus presentation periods on memory. Journal of Child Psychology and Psychiatry, 33, 399-409.
Stanovich, K. E. (1993). A model for studies of reading disability. Developmental Review, 13, 225-245.
Stothard, S. E., & Hulme, C. (1992). Reading comprehension difficulties in children: The role of language comprehension and working memory skills. Reading and Writing, 4, 245-256.
Stuart, M. (2004). Getting ready for reading: A follow-up study of inner city second language learners at the end of Key Stage 1. British Journal of Educational Psychology, 74, 15-36.
Swanson, H. L. (1993). Working memory in learning disability subgroups. Journal of Experimental Child Psychology, 56, 87-114.
Swanson, J. M., McStephen, M., Hay, D., & Levy, F. (2001). The potential of the SWAN rating scale in the genetic analysis of ADHD. Poster session presented at the annual meeting of the International Society for Research in Child and Adolescent Psychiatry, Vancouver, BC, Canada.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). New York: Allyn and Bacon.
Wagner, R. K., Torgeson, J. K., & Rashotte, C. A. (1994). Development of reading related phonological processing abilities: New evidence of bi-directional causality from a latent variable longitudinal study. Developmental Psychology, 30(1), 73-87.
Wechsler, D. (1999). Wechsler Abbreviated Scale of Intelligence. London: The Psychological Corporation.
Wilding, J., Cornish, K. M., & Munir, F. (2002). Further delineation of the executive deficit in males with fragile-X syndrome. Neuropsychologia, 40, 1343-1349.
Wilding, J. M. (2003). Attentional difficulties in children: Weakness in executive function or problems in coping with difficult tasks? British Journal of Psychology, 94, 427-436.
Wimmer, H., Mayringer, H., & Lander, L. (1998). Poor reading: A deficit in skill automatization or a phonological deficit? Scientific Studies of Reading, 2, 321-340.
Wimmer, H., Mayringer, H., & Raberger, T. (1999). Reading and dual-task balancing: Evidence against the automatization deficit explanation of developmental dyslexia. Journal of Learning Disabilities, 32, 473-478.
Yap, R. L., & Van Der Leij, A. (1994). Testing the automatisation deficit hypothesis of dyslexia via a dual task paradigm. Journal of Learning Disabilities, 27, 660-665.
Received 23 November 2004; revised version received 3 October 2005
Robert Savage (1*), Kim Cornish (1,2), Tom Manly (3) and Chris Hollis (4)
(1) Department of Educational and Counselling Psychology, McGill University, Canada
(2) Departments of Neurology and Neurosurgery, McGill University, Canada
(3) UK Medical Research Council Cognition and Brain Sciences Unit, Cambridge, UK
(4) Department of Psychiatry, University of Nottingham, UK
* Correspondence should be addressed to Robert Savage, Department of Educational and Counselling Psychology, Faculty of Education, McGill University, 3700 McTavish Street, Montreal, Quebec H3A IY2, Canada (e-mail: email@example.com).
Table 1. Means and standard deviations of age, IQ and reading characteristics by group Poor attention/activity Good attention/activity Age (months) 103.60 (16.74) 114.22 (16.88) IQ 88.14 (14.00) 108.60 (14.38) Reading accuracy 82.96 (12.38) 105.91 (12.50) Reading rate 87.57 (14.42) 107.47 (11.35) Reading comp 84.41 (12.95) 107.66 (12.61) Note. Standard deviations are shown in brackets. Key: IQ, Wechsler abbreviated scale of intelligence standard score; Reading accuracy, NARA reading accuracy standard score; Reading rate, NARA reading rate standard score; Reading comp, NARA reading comprehension standard score. Table 2. Means and standard deviations of age, IQ, reading and cognitive characteristics by group Mean SD Read Acc 95.23 17.20 ReadCom 96.89 17.46 Read Rat 98.33 16.26 IQ 99.18 17.47 WMBph 105.13 21.66 WMB ex 90.37 23.76 WMB spat 92.01 17.45 Dualtime 6.22 3.86 Dual acc 8.60 4.08 Samewld 8.78 3.16 Oppwld 8.41 3.55 Key: Read Acc, Neale analysis of reading ability reading accuracy standard score; ReadCom, Neale analysis of reading ability reading comprehension standard score; Read Rat, Neale analysis of reading ability reading rate standard score; IQ, Wechsler abbreviated scale of Intelligence standard score; WMBph, Working Memory Test Battery for Children phonological loop standard score; WMB ex, Working Memory Test Battery for Children central executive standard score; WMB spat, Working Memory Test Battery for Children visuospatial scratch pad standard score; Dualtime, Sky search dual-task age-scaled score; Dual acc, Score dual-task age-scaled score; Samewld, Response inhibition: same world task age-scaled score; Oppwld, Response inhibition: opposite world task age-scaled score. Table 3. Inter-correlations between literacy skill, IQ, working memory and attention measures 1 2 3 4 5 6 Read Acc .96** .88** .75** .62** .67** ReadCom .89** .85** .78** .61** .67** Read Rat .78** .72** .65** .50** .59** IQ .62** .69** WMBph .31** .27* .18 .65** WMB ex .33** .32** .29** .41** WMB spat .17 .21 .15 .21 .49** Dualtime .32** .26* .24* .18 .22 Dual acc .34** .31** .31** .07 .30** Samewld .41** .42** .44 .24* .21 Oppwld .42** .38** .43 .15 .33** 7 8 9 10 11 Read Acc .52** .48** .63** .62** .59** ReadCom .55** .44** .65** .62** .57** Read Rat .46** .42** .59** .62** .59** IQ .57** .39** .62** .53** .49** WMBph .50** .36** .42** .49** .41** WMB ex .68** .43** .60** .49** .54** WMB spat .40** .55** .46** .47** Dualtime .24* .51** .26** .38** Dual acc .32** .36* .57** .60** Samewld .22 .09 .37** .80** Oppwld .26* .25* .45** .73** Note. *p < .05. **p < .01. Note. Simple correlations are presented in the upper triangle, partial correlations adjusted for WASI IQ in the lower triangle. Key: Read Acc, Neale analysis of reading ability reading accuracy standard score; ReadCom Neale analysis of reading ability reading comprehension standard score; Read Rat, Neale analysis of reading ability reading rate standard score; IQ, Wechsler abbreviated scale of Intelligence standard score; WMBph, Working Memory Test Battery for Children phonological loop standard score; WMB ex, Working Memory Test Battery for Children central executive standard score; WMB spat, Working Memory Test Battery for Children visuospatial scratch pad standard score; Dualtime, Sky search dual-task age-scaled score; Dual acc, Score dual-task age-scaled score; Samewld, Response inhibition: Same world task age-scaled score; Oppwld, Response inhibition: opposite world task age-scaled score. Table 4. Factor loadings and communalities ([h.sup.2]) for principal components analysis and varimax rotation on cognitive processing measures Component Component 3 1 Verbal Component Dual-task/ working 2 Central Visuospatial Measure memory Executive working memory [h.sup.2] Sky search dual task .58 .42 Sky search score dual .60 .52 .65 task Same world .84 .80 Opposite world .90 .85 WMB digit recall .86 .75 WMB word matching .65 .52 WMB word recall .79 .68 WMB non-word recall .72 .56 WMB Corsi .57 .41 WMB mazes .79 .72 WMB visual patterns .77 .66 WMB Listening span .46 .57 .55 WMB counting span .46 .41 .54 WMB backward digit .49 .58 .70 recall Percent of Variance 43.89% 11.08% 7.93% Note. Only loadings above .4 are displayed. Table 5. Regression analyses exploring working memory and TEA-Ch component tasks as predictors of high-versus low-attention group membership 1. Dependent variable: Attention group membership Step IV Exp(B) [R.sup.2] change 1 Chronological age 1.03 .10*** 2 IQ 1.13 .47*** 3 Reading ability 6.56 .11*** 4 Verbal working memory 1.13 .00 4 Central executive 1.28 .01 4 Dual task/visuospatial memory 1.04 .03* Note. *p < .05. ***p < .001. Table 6. Regression analyses exploring working memory and Tea-Ch tasks as predictors of NARA reading 1. Dependent variable: Reading Step IV [beta] [R.sup.2] change 1 Chronological age 0.10 .01 2 IQ 0.78 .59*** 3 Attention group 0.35 .06*** 4 Verbal working memory 0.22 .02* 4 Central executive 0.25 .05*** 4 Dual task/visuospatial memory -0.09 .00 Note. *p < .05. ***p < .001.
|Printer friendly Cite/link Email Feedback|
|Author:||Savage, Robert; Cornish, Kim; Manly, Tom; Hollis, Chris|
|Publication:||British Journal of Psychology|
|Date:||Aug 1, 2006|
|Previous Article:||Does hunger influence judgments of female physical attractiveness?|
|Next Article:||Driving aggression in forensic and non-forensic populations: relationships to self-reported levels of aggression, anger and impulsivity.|